Wikipedia talk:Did you know/Archive 61

Archive 55

←

Archive 59

Archive 60

Archive 61

Archive 62

Archive 63

→

Archive 65

Incorrect info in new hook for Wide Awake in Europe

The rewritten hook for the U2 EP Wide Awake in Europe (queue link) has some mistakes. The way its written now means that the release was selected to be released on Black Friday to draw customers to record shops. That is not true. The release was especially created for a program to bring people into record shops for the holiday season, beginning on Black Friday. I don't know who rewrote the sentence, but it was rewritten with a new meaning that doesn't make much sense. The sentence says that the EP "was selected as part of a plan to draw customers to independent record shops". I don't really like the word "plan", it makes it sound like it's some sort of scheme. The word "event" or "program" works a lot better (which is what I originally had). Also, the EP wasn't selected for this program, it was created specifically for it.

How about:

"...that U2's latest live EP was created as part of a program to draw customers to independent record shops during the 2010 holiday shopping season?"

–Dream out loud (talk) 18:57, 8 November 2010 (UTC)

Well, I moved it to the prep queue, so I'll take the blame for that one. Unfortunately I'm lacking an admin bit so I can't fix it; hopefully someone else can? 28bytes (talk) 19:44, 8 November 2010 (UTC)

I edited the hook as requested. Thanks! --Orlady (talk) 20:36, 8 November 2010 (UTC)

Nice hook, by the way, IMHO. (with a wink to those who are reading the discussion above!) Physchim62 (talk) 23:01, 8 November 2010 (UTC)

Thanks, I'm glad you like it. I meant to add a link to "record shops" to the hook above (as it was linked in the original hook). Could that be added in before it comes out of the queue? Thanks. –Dream out loud (talk) 02:25, 10 November 2010 (UTC)

DYK as a possible part the sourcing of BLP drive

Since there are discussions of which groups of articles should be eligible for DYK (new, 5x expanded, GA etc) I would like to add another proposal. As you may know there are discussions in various spots [1][2] about the problem of unsourced BLPs, and the desire to eliminate the backlog of 20K articles by the 10th Anniversary of English Wikipedia. My suggestion is that, until the backlog is gone, newly and completely (top to toe) sourced BLPs be made eligible for DYK. Not new articles, but older articles that were unsourced as of some recent date, and which have now been fully and reliably cited. No minimum length requirements, but an interesting hook would obviously need to be part of the deal. Special care would need to be taken here at DYK with the checking the sources, tone etc, before putting an article on the front page, but that kind of checking could only be helpful for improving quality control overall. --Slp1 (talk) 00:15, 9 November 2010 (UTC)

I would support that. Seems like a reasonable way to help support folks who are resolving a (hopefully) short-term problem. I'd prefer some minimum length for the article, though, even if not the usual 1500 characters. 28bytes (talk) 00:32, 9 November 2010 (UTC)

I'd support that as long as there can be an interesting hook found from the article. I also agree with 28bytes that there are some really short biographies out there, so there needs to be a length limit in practice. I suggest saying that they can be nominated as "new" (i.e. minimum 1500 characters) without being a five-fold expansion, but that otherwise they have to pass the same requirements as any new BLP on DYK. Physchim62 (talk) 00:45, 9 November 2010 (UTC)

I'd accept Physchim62's suggestion, with the caveat that previously unsourced BLPs may be considered "new" articles for DYK purposes as of the date (a) source(s) is/are added. In other words, they need to be nominated in a timely manner following the addition of sources. And of course, must meet all other DYK requirements. cmadler (talk) 12:54, 9 November 2010 (UTC)

I've been adding references to unsourced BLPs as one of my contributions to the 'pedia. The above suggestion made me consider some of the articles that I have worked on recently, so I invite some thoughts on these:

• Oxana Yablonskaya - I added refs on 30 October, the length passes DYK comfortably, but the article would need some more refs. Thing is, I'm having trouble with a hook.
• Matthew Scannapieco - an American former Mayor and now convicted criminal. Someone has added refs that would need fixing, but is there a hook to be had here that isn't going to raise BLP issues as 'negative'?
• Robert Olmstead - this would need some expansion, but that could be done. But, are we going to accept the suggestion if the referencing was done recently enough (2 November, in this case)?

Thanks. EdChem (talk) 13:40, 9 November 2010 (UTC)

I have added a nomination at Template talk:Did you know#Robert Olmstead to test this proposal. All comments invited. EdChem (talk) 06:24, 10 November 2010 (UTC)

Let's decide

Ok, setting aside the following questions still being discussed:

• How to determine interesting vs. mundane
• Implementing "some" change to the article requirements (maybe not just new articles)
• Implementing "some change" to the nomination/selection process.
• How to deal with concerns of "immediate gratification".

I think that there are a couple things which could be implemented immediately, now.

• 1.) Frequency - change from 4 times a day to 3, 2, or 1
• 2.) Process of review - change to requiring 2 reviewers, at least for now.

Can we agree on at least this, and make it happen now? - jc37 02:31, 9 November 2010 (UTC)

Having read through much of the above discussion, I don't think there's consensus for either 1 or 2. The only proposal with any consensus so far is to reduce the workload by implementing monthly limits on submissions, but an exact numerical limit is still being debated on.--hkr Laozi speak 02:38, 9 November 2010 (UTC)

I think there's consensus on reducing the frequency of updates. Physchim62 (talk) 02:43, 9 November 2010 (UTC)

Very borderline. And (for most of the commentators) as a result of reduced output and workload, not as a cause. I wouldn't be opposed to it, but the consensus on that one is shaky, especially without implementing anything else first.--hkr Laozi speak 02:51, 9 November 2010 (UTC)

Some people are more than happy to argue round in circles – that any reduction in frequency needs a process to go with it, but that any proposed process would lead to more work under the current system. It's really quite dangerous to argue for the staus quo in that way, as it risks a consensus forming that the frequency of DYK updates should be zero. Physchim62 (talk) 02:57, 9 November 2010 (UTC)

Also, the proposal to suggest that mundane hooks not be approved could be introduced tomorrow, as it is already in DYK rules and procedures (just not used enough). Physchim62 (talk) 02:57, 9 November 2010 (UTC)

I don't see a consensus for changing the frequency of the updates. (I see a couple of folks suggesting it, but no actual consensus in the larger discussion for doing it.) There does appear to be a consensus for capping monthly submissions. - —Preceding unsigned comment added by Dravecky (talk • contribs) 02:58, 9 November 2010 (UTC)

I support a reduction in updates (and therefore number of hooks per day) if there is a process to do that which is more than the current rules and procedures, where (I think) a single reviewer can reject a hook because they are bored by science/sports/popculture/or botany/etc. First Light (talk) 03:11, 9 November 2010 (UTC)

Current reviewers feel they're not even able to insist on a hook change if the article itself fits the technical requirements: at least one has said so above. Again, this is simple arguing around in circles: people are complaining against a cut in frequency because they say there is no process for refusing submissions, and then refusing any effective process for refusing submissions, all in order to keep the cosy status quo. Physchim62 (talk) 03:21, 9 November 2010 (UTC)

I support Johnbod's specific suggestion for a process that would help guide the rejection of submissions. See my explanation above.[3] I also realize this discussion is all over the place, and too long to read everything. First Light (talk) 04:27, 9 November 2010 (UTC)

Could you please point me to the section containing the discussion/consensus for capping monthly submissions? It is proving rather difficult to find, much less follow, individual discussions on a page that is consistently 300–400 KB in length. Thank you, -- Black Falcon ^(talk) 04:13, 9 November 2010 (UTC)

Wikipedia talk:Did you know#A simple proposal. There seems to be some agreement that there needs to be a cap, but not on what that cap should be.--hkr Laozi speak 04:31, 9 November 2010 (UTC)

I'm afraid I don't see it. First Light's proposal involved limiting the number of hooks (i.e., reducing the frequency of updates), not a cap on the number of nominations per month. I recall reading a comment in opposition to a cap (something along the lines of "... we should not limit or discourage editors who can and choose to write many decent articles in a short period of time"; I paraphrase, of course), but I can't find it now... (sigh) -- Black Falcon ^(talk) 04:53, 9 November 2010 (UTC)

Whoops, I meant the section above that one, Wikipedia talk:Did you know#Dispelling a myth about DYK. The proposal begins a few paragraphs below your comment, with various monthly caps suggested.--hkr Laozi speak 04:58, 9 November 2010 (UTC)

Ah, thanks. : ) -- Black Falcon ^(talk) 05:05, 9 November 2010 (UTC)

I think the discussion above has gotten too convoluted to discern a consensus for changing the current process. I suggest a separate RfC page and perhaps a WP:CENT listing in order to get clearer consensus before making changes. Having said that, I don't think the changes proposed above solve the root cause of the problem, and that if there is going to be change, much more radical change is desirable. –Grondemar 05:00, 9 November 2010 (UTC)

The much more radical changes will (due to their very nature) have a difficult time at achieving consensus. Due to the sheer amount of opposition, radical changes like switching to Good Articles or completely scrapping DYK (which has been proposed on this talk) will never happen. The much more conservative proposals, like an article cap, are likelier to be implemented. This doesn't mean that radical proposals are wrong, or don't merit discussion, however.--hkr Laozi speak 05:24, 9 November 2010 (UTC)

Consensus so far

Here's the basic gist of the consensus so far on having a monthly self-nom cap for DYK articles: Cbl62 suggested a cap of 10, 15, or 20. Cmadler supports 15 a month, with a limit of 2 per day. Lampman supported a cap, but did not name a number. Dravecky supports a 15 hook/20 article cap, or a 12 hook/15 article cap. Mikenorton agrees with Dravecky, 12 hook/15 article cap. I (hkr) am for any cap between 15 and 20. Physchim62 supports a cap, didn't name a number, but is against a high cap, which is implied as anything in the double digits. First Light is for a cap of 4 per month. Gatoclass is hesitant about implementing a cap, pointing out that it may contribute to problems with quality. MaterialScientist is firmly against a cap, saying that the emphasis should be on reviewing, not on limitations. Victuallers supports a cap, but preferably the lowest. Orlady supports a limit, but is against a rigid cap. Johnbod supports a single figure cap. jc37 opposes a cap, due to potential problems with socking. Hansandler commented in the section, but not on the cap. This is not a vote, just a brief overview of the suggestions so far. No one's supported a cap on nominations, just on self-noms.--hkr Laozi speak 05:17, 9 November 2010 (UTC)

I personally would oppose a cap, simply because I don't see how instituting an individual nomination cap fixes the root problem. –Grondemar 05:21, 9 November 2010 (UTC)

And I just added my opposition too, for basically the same reason as Materialscientist. -- Black Falcon ^(talk) 05:23, 9 November 2010 (UTC)

Well, the idea would be that a cap would reduce the workload, which (along with a reduction in DYK frequency) could give reviewers more time to review each article, thus fixing the problems of quality. Whether or not this can work in practice is debatable.--hkr Laozi speak 05:38, 9 November 2010 (UTC)

Not a vote, but so that makes the consensus (roughly around): 1/3 for a cap around the 10-20 range, 1/3 for a single figure cap, and 1/3 opposing any limitations. --hkr Laozi speak 05:33, 9 November 2010 (UTC)

And I oppose the cap due to it potentially fostering socking... - jc37 05:53, 9 November 2010 (UTC)

Yes, you're in there: "jc37 opposes a cap, due to potential problems with socking". ;) --hkr Laozi speak 06:11, 9 November 2010 (UTC)

Please add me to the list of those opposing any limitations at this stage. I think MaterialScientist's comments have been spot-on as usual. And I think before we start trying to work out a consensus on what changes we need to make, we ought to first work out a consensus on exactly what problem(s) we're intending to solve. Of all of the problems with DYK (real and otherwise), the one that concerns me the most is the issue of copyright violation and/or plagiarism, and yet that seems to have been largely dropped from the conversation. If someone believes there is a more pressing problem than improving the review process vis-a-vis copyright issues that the caps are intended to address, please chime in. As I said earlier, I would much rather have a boring hook than an interesting one that links to a plagiarized article. I strongly suggest we focus on that issue first and foremost. EdChem and others have made some excellent suggestions about automating the copyright-checking process and providing greater transparency for the purposes of tracking down problems, but those suggestions seem to have gotten lost in the shuffle as the conversation has shifted towards (in my view) much less pressing issues. 28bytes (talk) 06:37, 9 November 2010 (UTC)

I'm also opposed to a rigid cap. Far better to apply a throttle, as I suggested above, requiring experienced DYKers to make contributions in terms of reviewing or nominating newcomers' articles. Rd232 ^talk 09:38, 9 November 2010 (UTC)

No, it's all connected. To do more stringent reviews we need fewer hooks to review; everybody was agreed on that a few days ago, and it makes sense. Automated copyright checking is great, but will not need much discussion here, & I don't think there are specific issues that need agreeing on it just now. Johnbod (talk) 06:51, 9 November 2010 (UTC)

It may be connected, but putting hook caps in, then addressing systemic copyright issues is precisely backwards. Getting some copyright tools and guidelines in place first will naturally slow the hook output down anyway if it's burdensome, and if it's not burdensome and copyright checking becomes standard operating procedure with minimal impact on the hook flow, then there wasn't really any need for a cap, was there? 28bytes (talk) 07:19, 9 November 2010 (UTC)

Question: how would such a system be enforceable? Would we have a bot maintain a giant list of nominators? Ed ^{[talk] [majestic titan]} 06:32, 9 November 2010 (UTC)

Principally by the honor system. If reviewers notice a lot of noms from a user, the user's talk page can also be checked. Cbl62 (talk) 06:38, 9 November 2010 (UTC)

• I don't support requiring 2 reviewers, & I don't think many others did - let GA try that first, I say. Reducing frequency yes, also self-noms as noted in the summary just above; anything down to single figures. I'd also just point out that the "interesting" rating, because it doesn't have rules or mandate anything, could be tested for a period at any time without affecting other aspects of DYK. Let's see if the sky really does fall on our heads! Johnbod (talk) 06:15, 9 November 2010 (UTC)

I think I must point out again that merely reducing the frequency of updates does nothing to resolve any of the problems discussed - it won't do much to reduce overall workload and it certainly won't guarantee that articles get more scrutiny. The only thing that will do the latter is to have a more robust review process in place - more like the GA process, perhaps.

As somebody said above, I think the first thing we really need to do is work out which problem or problems we are trying to fix. Is it to crack down on plagiarism? Is it to provide better quality articles? More interesting hooks? What? The lack of focus of the discussion so far has been part of the problem I think. Gatoclass (talk) 08:12, 9 November 2010 (UTC)

To make any progress this discussion clearly needs to be restarted on a subpage. We had that, and for some reason it was undone, with predictable consequences. Rd232 ^talk 09:38, 9 November 2010 (UTC)

Proposal to increase basic size requirement

Okay, here's my take. The main issue I think is that reviewers need more time to check submissions for plagiarism/copyvio. The only practical way to do that is to cut down the number of reviews, which means either disqualifying many submissions or reducing their number. If we are going to disqualify/reduce the number, we may as well have a system in place for picking the best submissions.

On that score, the simplest way to pick the best submissions is by size, because the least qualified submissions are usually the short stubby ones. To put it another way, whether we go down the disqualification path or the higher bar for submissions path, it's mostly the short stubby articles which will go. I know that I personally do not want to have to try and pick and choose between 40 articles in a given day to try and work out which are the best. So I would rather simply raise the bar to reduce the number. So, let's say we raise the bar to a minimum of 3,000 characters instead of 1,500.

The objection to raising the minimum size of submissions is that users will try and "bloat" their articles out to 3,000 characters to try and get them through. This is certainly a potential problem, but I still think that raising the minimum size should reduce the overall number of submissions, as well as making it easier to decide which ones are quality submissions. The important thing is that raising the character requirement reduces the burden on reviewers, whereas trying to pick which are the "best" submissions from the existing pool of about 40 a day is not going to reduce the burden (as we've seen from the suggestions that two users review each hook etc.).

So I think the only practical way to reduce the burden on reviewers is to reduce the initial number of submissions, and the most straightforward way to do that is raise the character requirement. We could try the 3,000 level and if that doesn't work, or if it's too much, we can raise or lower it as appropriate. But once we have reduced the number of submissions, we can then put in place a more robust process for reviewing itself. In my opinion this is the most practical approach, and probably the only one. Certainly I think it deserves a try before we start thinking about more radical methods - the simplest approaches are usually the best. Gatoclass (talk) 10:58, 9 November 2010 (UTC)

I support this suggestion, especially the last line. --Gerda Arendt (talk) 11:06, 9 November 2010 (UTC)

Stray thoughts: Quality of hooks is hardly related to the article size. Copyvios are much easier to spot in stubs. Quality of articles generally does increase with the prose length, but is it really the major issue to be solved?(forgive my ignorance, I missed most of the discussion) Materialscientist (talk) 11:16, 9 November 2010 (UTC)

"Quality of hooks" is the least important issue, because it's too subjective an issue. What you find interesting and I find boring might be very different and vice versa. Quality of article is the important thing. I should also have added above that another advantage of bar raising v disqualification is that it prevents systemic bias creeping in with reviewers picking their favorite topics for promotion and discarding the topics they dislike. We will still be getting the same variety of topics with higher requirements, which I think is another important point. Gatoclass (talk) 11:21, 9 November 2010 (UTC)

We do have editors who write good, bulky and useful articles but don't care about shaping a hook and otherwise - stubs with a nice hook. Article quality is fine with me (as a DYK insider), but I thought the outsiders (who started all the discussion) blame lame hooks, not lame articles. Materialscientist (talk) 11:31, 9 November 2010 (UTC)

I think lame hooks are both a problem in themselves (because its the hooks that appear on the Main Page, not the articles) and also a symptom of a wider problem (that reviewers are simply not willing to refuse submissions, always dangerous from a quality control point of view whatever the QC system). Physchim62 (talk) 11:41, 9 November 2010 (UTC)

We can't make "hookiness" the primary criterion by which to select articles. If we were to go by the most interesting hooks to the most people, we would just pick hooks involving sex and violence, because they are the hooks that get the most hits. We are not a tabloid, we are an encyclopedia. An important function of DYK IMO has always been to showcase the variety of topics that are written about, and if some of those topics are less interesting to the average reader, that's a price that I think has to be paid. Gatoclass (talk) 11:59, 9 November 2010 (UTC)

Raising the character requirement does absolutely nothing to relieve the pressure on reviewers looking for copyvio or verifiability issues. Even if doubling the character requirement to 3000 led to the number of submissions dropping by half (a very dubious assumption), reviewers would still have exactly the same amount of text to study. This is the one proposal that's just about guaranteed to make DYK worse, not better. Physchim62 (talk) 11:41, 9 November 2010 (UTC)

I disagree. Doubling the character count does not mean doubling the number of sources to check. It should be much easier to check one 3k article for copyvio than two 1.5k articles. The many other advantages of raising the minimum requirement would also more than compensate for that. Gatoclass (talk) 11:49, 9 November 2010 (UTC)

(ec)I shouldn't say this as I nearly stopped reviewing :-P, but speaking of simple solutions - obliging a submitter to review an article per submission mathematically solves the problem. Our submitters are mostly experienced enough to review .. Materialscientist (talk) 11:50, 9 November 2010 (UTC)

But that's going to put an unequal burden on the most prolific contributors, and it's also going to mean a lot more noob reviews - neither of which is likely to improve the quality of reviews. Gatoclass (talk) 12:04, 9 November 2010 (UTC)

Extra review is always helpful. Prolific contributors are usually not bad reviewers, they just hesitate to make time to review. Materialscientist (talk) 12:09, 9 November 2010 (UTC)

Well look, if you want to try that, I guess we could give it a go. I'm not sure if it will be enough to resolve the problem though. Are Cbl and Alansohn going to really buckle down and thoroughly check every article they review for copyvios? Can we rely on every editor to do that? It seems a bit dubious to me. Gatoclass (talk) 12:16, 9 November 2010 (UTC)

More reviews will improve quality. I do not see a solution to the copyvio problem - if they don't get caught at FAC, we shouldn't expect them to be caught by a single reviewer at DYK. The post-factum "punishment" for the editor who created copyvio should be the main solution, IMHO. Materialscientist (talk) 12:43, 9 November 2010 (UTC)

Another, even more heretic thought: those who write need motivation to write, and this expands WP. Those who nominate other articles (no slight to them) help bringing articles, which are created anyway, to DYK attention. If we really have to cut submissions by any means available - block the nomination of other's articles. Materialscientist (talk) 12:02, 9 November 2010 (UTC)

I think that's a little too heretical :) We'd be likely to get complaints if we did that. Gatoclass (talk) 12:06, 9 November 2010 (UTC)

Aren't we getting them above? No radical solution will cause no complaints. Materialscientist (talk) 12:09, 9 November 2010 (UTC)

I mean, complaints from those outside the process. People would start complaining that DYK is just a club for a small cabal of editors to give each other awards. And I think that argument would have some legitimacy. Gatoclass (talk) 12:12, 9 November 2010 (UTC)

Let's just say that people already complain that DYK is just a club for a small cabal of editors to give each other awards! ;) But yes, it would significant raise the barriers to newcomers participating in DYK, which cannot be a Good Thing, and I don't think it would help very much with the submission numbers. Physchim62 (talk) 12:28, 9 November 2010 (UTC)

We can easily get numbers on that, I think the percentage is significant. Materialscientist (talk) 12:43, 9 November 2010 (UTC)

I think a lot would depend upon whether PFHLai is active or not :) Gatoclass (talk) 12:47, 9 November 2010 (UTC)

The dark side of this solution is that those nominators actually help with reviews, and it is only meant as a last resort for cutting down the submission number. Materialscientist (talk) 12:56, 9 November 2010 (UTC)

Physchim62 says "Raising the character requirement does absolutely nothing to relieve the pressure on reviewers looking for copyvio or verifiability issues." I could not agree more strongly. In fact, I would go further and suggest this would make the problem worse in the absence of automated copyvio tools, simply because there would be twice as much text to manually check! I also would like to point out, quite selfishly, that most of my nominations are between 1500 and 3000 characters and I take great care to avoid stepping into the gray area between summarizing and close paraphrasing. 28bytes (talk) 14:45, 9 November 2010 (UTC)

Well, obviously you have a considerable conflict of interest in commenting on this proposal - but your honesty is appreciated. I'm sure you are not the only contributor who may be alarmed at the notion of increasing the basic requirement. I might point out however that if the object is to reduce the total number of submissions, many of your articles would probably get scrubbed under a different system for being too insubstantial in any case. And BTW I already responded to Physchim's point that you quote above. Gatoclass (talk) 17:01, 9 November 2010 (UTC)

I, too, agree with Physchim62 in this case. Requiring more text does not address the key problem (which sparked all of this discussion) about the need to review articles for proper sourcing and improper paraphrasing. It would reduce the number of nominations to be reviewed, but would also increase the amount of text per nomination. I think that we ought to consider raising the character requirement, but immediately doubling it is too extreme (in my opinion, any increases should be incremental: i.e., 1500 → 1800 or 2000 → 2500 → 3000, and so on). -- Black Falcon ^(talk) 17:40, 9 November 2010 (UTC)

I agree that if the goal is to reduce output, character count is a reasonable way to do it, since it's an objective metric (relatively) resistant to gaming. I just don't see how it solves any practical problem we have. As MaterialScientist points out, it's a lot easier to spot a sentence or two of plagiarized text in a short article than a long one. 28bytes (talk) 17:34, 9 November 2010 (UTC)

I'm thinking that matsci's suggestion about tit-for-tat reviewing might have some merit. Certainly we could give it a try. I don't think it will be enough by itself to solve the problems, but it ought to reduce the burden of the regular reviewers, which would be a step in the right direction. Gatoclass (talk) 12:45, 9 November 2010 (UTC)

• I'd suggest that new (to DYK) editors get some leeway before they're required to start reviewing. I'd suggest the following rule: Editors who have submitted five previous DYK hooks are required to review at least one hook for each hook they nominate. cmadler (talk) 13:14, 9 November 2010 (UTC)

That sounds reasonable. Gatoclass (talk) 14:47, 9 November 2010 (UTC)

Not trying to be difficult here, but how do you define a review? Say that you immediately spot that a hook is way too long, put up a question mark and inform the nominator, does that count? Or you add missing alt text to several images, does that count? I'm just pointing out that there are a many hooks where a lot of work is done by reviewers, before it reaches the final checking and tick stage. Mikenorton (talk) 15:09, 9 November 2010 (UTC)

That occurred to me too. I would define a review, I think, as actually reviewing the article, and then placing one of the icons, (, , ) with reasons for the decision, beneath the hook at T:TDYK. Gatoclass (talk) 16:08, 9 November 2010 (UTC)

Statistics on DYK article size

I went through the last 100 articles in the archive, and came up with the following size statistics:

Characters	# under	Total under	Total over
2000	17	17	83
2500	15	32	68
3000	5	37	63
4000	9	46	54
5000	13	59	41
6000	9	68	32
7000	7	75	25
10000	8	83	17
20000	9	92	8
30000	7	99	1
50000	1	100	0

It should be said that the numbers above aren't directly transferable to a situation after size requirements have been raised. In many cases it will be a quick job to fluff up an article by a few hundred characters, so a significant portion of the 37 articles under 3k must be counted among the ones over 3k. I doubt that such a measure would reduce the nominations by as much as a third. If we wish to cut the number of hooks, say, in half, we would have to raise the minimum requirement to 5k.

On a different note (and here I'm talking as a repentant sinner myself) the quickest way to reduce numbers would be to put a cap on Norwegian topics. For some reason, Norwegian-proficient editors seem to be the worst culprits when it comes to pushing through sub-standard, stub-like articles, simply to rack up numbers. Discounting some highly admirable rail road-related articles by User:Arsenikk, there were six Norwegian-related articles with an average of less than 2k of readable prose, none of them with over 2.5k. This average was less than a third of the overall. I'm not naming any names, but I think a lot could be achieved if certain established editors realised that not everything they create has to go on the main page. Especially since, as several have mentioned above, DYK is supposed to work as an introduction for new editors.

Which, of course, brings us back to the issue of self-nom caps. Lampman (talk) 15:06, 9 November 2010 (UTC)

And boring hooks! But the table makes a good point. I do think forcing those determined to nominate to pad out their articles will itself make plagiarism/close paraphrase more likely. Johnbod (talk) 15:47, 9 November 2010 (UTC)

I've thought some more about the cap issue. It is not intended as an attack on any of the prolific DYK contributors. But what Lampan and Johnbod say is true. If you look at the most prolific contributors (myself included), there is sometimes a tendency to put too much forward at DYK. Despite my personal passion for football history, I realize that not every college football article is Main Page worthy. The same is true of Geschichte's Norwegian articles, Alansohn's New Jersey legislators, Billy Hathorn's Louisiana/Texas legislators, TonyTheTiger's sports teams, Daniel Case's NY historical sites, Sasata's mushrooms, etc. We each should be considering whether an article we create is really Main Page material. Otherwise, we risk backlash (as well as more radical changes that can gut the DYK process). Aside from self-control, I believe a per month cap on DYKs would be helpful. It serve multiple goals: (1) helps focus prolific editors on self-selecting only their best items for DYK, (2) increases the variety of hooks and reduces backlash against being inundated by particular topics, (3) increases space at DYK for other contributors, (4) reduces the number of overall submissions and allows reviewers to focus on a smaller number of hooks, (5) minimizes the "game" (or "Smarty" collecting) aspects of DYK and thus reduce the temptation to generate lower quality articles (or even to engage in copyvio). I could be wrong. A cap might serve to discourage great contributors like Alansohn and Daniel Case. I don't think it would, but if it created problems, we can adjust or get rid of it. I suggest we try it for a month and see how it goes. Cbl62 (talk) 16:31, 9 November 2010 (UTC)

I'm not at all persuaded a cap would be effective. The number of contributors who submit more than, say, 20 hooks a month would be pretty small - and DYK features over 1000 hooks a month. Even if you eliminated a couple of hundred hooks, every month, by this method, which seems unlikely, it would only be a 20% reduction. Gatoclass (talk) 16:46, 9 November 2010 (UTC)

While a 20% reduction would not solve all our problems (I don't think any single change will), it would certainly be significant. Even if it were just a 5% reduction, it would be a start. cmadler (talk) 18:37, 9 November 2010 (UTC)

I'm not clear on the meaning of figures in the table. Are these sizes as measured by DYKcheck, or just raw bytesize? If the latter, the table cannot be considered as reliable at all, since articles are frequently much larger in raw bytecount than they are in character count. Gatoclass (talk) 16:13, 9 November 2010 (UTC)

It's readable prose. Lampman (talk) 19:51, 9 November 2010 (UTC)

Second opinion requested

The nomination for Danny van Bergen checks out on date, length and hook source, but before I approve it I wanted to solicit a second opinion on the sourcing. All nine references are from sherdog.com. That site has a wiki article with references to the Washington Post discussing the site, so I'm not disputing its status as a reliable source, but I'm concerned that the article is essentially single-sourced.

According to rule D12, "Multiple sources are generally preferred, though more leeway may be given for more obscure topics." I'm leaning towards calling this an obscure topic and passing the hook; any objections or further comments? Thanks, 28bytes (talk) 14:33, 9 November 2010 (UTC)

• Sherdog verifiably provides mixed martial arts results / data to ESPN, so it's reliability isn't an issue
• Adding reference(s) from other MMA sources (like the M-1 Global site, or mmauniverse.com, or, ...) wouldn't be hard
• I'm guessing the nominator is Paralympiakos, with whom I have collaborated on MMA articles, so I consider myself CoI-barred from going near the nom in any decision-making capacity; feel free to weight my opinion / comment accordingly

EdChem (talk) 14:52, 9 November 2010 (UTC)

There is material at the M-1 Global site, but it's pretty sparse so I wouldn't say it really adds to the verifiability of the article. I'd say give it the leeway as an "obscure topic". Physchim62 (talk) 02:38, 10 November 2010 (UTC)

Thanks both for the input. I'll go ahead and approve. 28bytes (talk) 02:47, 10 November 2010 (UTC)

Reviewing

As far as I can see, the one issue where there is broad consensus is the desirability of improved reviewing. Consequently, I am thinking of writing a reviewing guide, something that would remind each of us of things to check. I think it could usefully be accompanied by a DYKrev template, which would serve several purposes:

1. In accountability terms, it means someone is taking responsibility for having checked the article - which will also help us to help newer reviewers if / when they miss things. It will also be useful for transparent record-keeping
2. In recognition terms, it will allow us to thank those who regularly contribute to DYK in this way - maybe reviewing medals? I have already created a DYK reviewer ribbon, for those that like that sort of thing
3. It will give us data on whether some regulars who might be reasonably expected to participate in reviewing aren't choosing to

I think maybe requiring a reviewer to include themself as a DYKrev, and record in their !vote that they checked technical aspects, readability, etc, would make sure these are all actively considered in a review. A few issues to consider:

• what are reasonable expectations of a reviewer, in terms of plagiarism / copy vio / close paraphrasing checking?
• should reviewers look for alternative hooks if the proposed one seems (to them) dull, or just note a view of 'dull hoook' and leave the review to another reviewer?
• what are reasonable expectations for article quality, in a non-technical sense?

Thoughts? EdChem (talk) 13:10, 9 November 2010 (UTC)

I think the primary responsability for finding a suitable hook must lie with the submitter. There is no obligation on DYK to review articles which come with unsuitable hooks: the DYK selection criteria say "since the fact [in the hook] is an extraordinary claim", so mundane hooks are obviously unsuitable for DYK. A reviewer may choose to help out in finding a better hook, especially for newbie submitters, but there shouldn't be any general obligation to do so. Physchim62 (talk) 13:25, 9 November 2010 (UTC)

Agreed, but of course if one sticks out they can suggest it. The English, allowing for international varieties, should be free of gross grammatical errors, which is quite often a problem with non-mother tongue nominators. A list of noms that need a quick copy edit in this respect, maybe on the talk page, would be a good idea. It seems harsh to reject them out of hand. There should be a reasonable level of wikification - this is usually ok at the moment I think - articles are rejected if they fail this. Johnbod (talk) 15:35, 9 November 2010 (UTC)

I agree; the article need not be well-written to the GA standard, but should be, as you put it, free from gross grammatical errors. cmadler (talk) 17:20, 10 November 2010 (UTC)

There is a general rule for Main Page content that it must be free of orange-level cleanup tags. Reading through T:TDYK, I see that reviewers are asking for yellow-level cleanup issues to be resolved as well, and that's probably a reasonable expectation for DYK as long as reviewers are aware of possible tag bombing. And Johnbod makes a good point that reviewers should make sure there are no glaring problems with grammar or vocabulary: FAC has passed, and TFA has posted onto the Main Page, an article containing the phrase "At the age of five, she left Armenia with her parents and her four-years older brother, Mikael, because…" as representing "the very best work on Wikipedia"! Physchim62 (talk) 03:10, 10 November 2010 (UTC)

It would be useful if we could state the requirements in terms of existing requirements. For example, if the requirements could be restated in terms of the requirements for C-class, B-class, and GA-class articles, it would save re-inventing the wheel on at least one level. Haus^Talk 03:33, 10 November 2010 (UTC)

SimplySiti as lead hook

• ... that lipsticks in the SimplySiti cosmetic line are named after hit singles and albums produced by the company's founder and Malaysian singer, Siti Nurhaliza (pictured)?
  

Other opinions requested. I saw the hook above in prep area 3 as the lead hook. The hook reads like an ad for a line of cosmetics, and giving it the prominence of the lead hook with a photo of the sponsor takes it even further. If we're going to let the hook run, I suggest it not be as a lead. Cbl62 (talk) 15:46, 9 November 2010 (UTC)

You are probably right that it's a little too promotional. On the other hand, that's a really cute pic. And we don't feature many Malaysian hooks in the lead spot. And these products are unlikely to be for sale in English-speaking countries. I'm a little undecided. Gatoclass (talk) 17:11, 9 November 2010 (UTC)

I agree that it shouldn't be used as the lead hook, and personally I'm unsure if it should be used at all. Sure it's a nice pic + reasonably interesting, but it looks as if we are advertising for them. I think we need to be more careful about hooks and articles which read like adverts in general at DYK. Before this all kicked off, we had a set of hooks with both 1366 Technologies and i (newspaper) in, which was really bad. SmartSE (talk) 22:15, 9 November 2010 (UTC)

Can something be done with the hook so that it looks less commercial? Since when I suggested the hook after creating the article, I have no attention of whatsoever to make it looks like an advert. By the way, the picture is taken during one of her SimplySiti's event and it took me forever to come with a good and clear picture. =_=" This is my first nomination though, I still have many things to be learned... :) SyFuel^Ignite_Burned 17:23, 10 November 2010 (UTC)

Anyone know what's happened to this? I can't find it in the queue or back at T:TDYK. SmartSE (talk) 23:43, 10 November 2010 (UTC)

I've returned it to T:TDYK yesterday. There is an ALT there already by Orlady, please comment. Materialscientist (talk) 23:48, 10 November 2010 (UTC)

Hitchhiker's Guide to DYK Reform: A summary of history, purpose, problems, and everything else

Alright, let's take a deep breath. How did this 400+ KB discussion on DYK reform start? What were the initial problems? What are the current problems? And what are the solutions?

Overview of problems: History, initial problems, and a chronology of the discussion

After a string of RfAs listing large DYK numbers, User:SandyGeorgia became concerned that "DYKs are a much bigger part of the reward culture, and those people are demanding their DYK prize, while turning out shoddy content and overwhelming your review ". She began to regularly watch DYK, and on October 28, identified that the article Malta Test Station (Wikipedia talk:Did you know/Archive 59#Copyright problem: Malta Test Station), was plagiarised. This brought up a large discussion over on ANI over plagiarism. SandyGeorgia then began to identify other articles that were either closely paraphrased (User:SandyGeorgia/sandbox#Identified DYK problems) or fail Wikipedia's notability and reliable souce guidelines (Wikipedia:Articles for deletion/Black Eyed Kids (3rd nomination)).

So the initial problem was (Wikipedia talk:Did you know/Archive 59#Copyright problem: Malta Test Station):

1. "DYKs are a much bigger part of the reward culture, and those people are demanding their DYK prize, while turning out shoddy content and overwhelming your review"

Which led to (User:SandyGeorgia/sandbox#Identified DYK problems):

1. There have been incidents of plagiarism and close paraphrasing, and resulting copyright violation problems.
2. Some articles accepted DYK for do not follow the guidelines of WP:N, WP:OR and WP:V.

And that led to the conclusion that (Wikipedia talk:Did you know#Changing DYK):

1. There aren't enough reviewers. The DYK workload is too high, resulting in a lack of scrutiny.
2. There is an emphasis on quantity over quality. The quality of DYK nominations is not consistent. There have been many poor quality DYKs without reliable sources.

And then brought up (with varying consensus) the complaints that:

1. Good articles are not featured on the main page, and yet they require much more effort than DYKs.
2. The time limit for DYKs is too restrictive.
3. Many hooks are mundane.
4. The size requirements for DYK are too low.
5. DYK should not focus purely on new articles.
6. DYK doesn't do anything specifically to help systemic bias issues
7. DYK articles need to have higher requirements for sources.
8. DYK has not done enough to highlight the new articles of new editors, consistenting of articles by only established ones.

Overview of every solutions proposed so far, their flaws and consensus

These are all (or at least most) of the solutions that have been proposed and discussed so far:

1. Slowing down the output rate.

   Flaw: Nearly unanimous consensus. But the question is: How?

2. Incorporate CorenBot to scan for plagiarism

   Flaw: None, other than technical issues. Corenbot does not check Google Books for plagiarism due to TOS issues. Unanimous consensus on this suggestion.

3. Encourage editors to manually use online plagiarism checkers.

   Flaw: None, other than technical issues. Free tools and tools on Wikipedia tend not to be as effective as something like Turnitin.com.

4. Require two reviewers per hook.

   Flaw: Will double the workload, and DYK lacks volunteers.

5. More transparent logs, better accountability.

   Flaw: Not a lot of opposition, but COI issues are rare and this solution doesn't really address any of the other problems.

6. Introduce Good Articles to DYK.

   Flaw: It prevents articles by new editors from being featured on the main page and restricts all the main page sections to insiders and established users.

7. Abandon the "new article" concept.

   Flaw: DYK would no longer represent the fact that Wikipedia is constantly growing. Might not address workload problems.

8. Abandon or increase the time limit criteria.

   Flaw: Does not address the workload problems. Time limit criteria should already be enough.

9. Reject boring hooks.

   Flaw: Appeal is subjective. How do we determine if a hook is interesting? And this suggestion may require multiple reviewers.

10. Increasing the character limit.

    Flaw: Character limit may not be a reflection of actual quality. And having to check for more content increases the workload.

11. Increasing the citation requirements.

    Flaw: Having to check for more sources may increase the workload.

12. Reducing DYK frequency.

    Flaw: Lacks clear consensus, more of a result of a solution than an actual solution.

13. Reqire article nominators to review articles.

    Flaw: Process may result in poor, half-hearted reviews.

14. Having a DYK lottery (either through Yzx's method or Roux's gradual approach).

    Flaw: Will disincentivise reviewers, who might end up reviewing articles that never get chosen.

15. Limit monthly or daily self-nominations

    Flaw: Does not address concerns with the review process. May encourage sockpuppetry.

16. Abandon DYK entirely.

    Flaw: Not going to happen.

17. Do nothing.

    Flaw: And wait until SandyGeorgia adds DYK back on to her watchlist?

Discuss.--hkr Laozi speak 17:58, 9 November 2010 (UTC)

Discussion

I think that's an excellent summary of the discussion so far, with the possible quibbles that I don't consider the "lower the output" recommendation to be unanimous, and I'm completely happy to have SandyGeorgia watching us. I'd rather her find a plagiarism problem than it go unfound. 28bytes (talk) 18:37, 9 November 2010 (UTC)

Thanks for the summary. Actually adding GAs will not increase the workload; as I understood it, the idea was that hooks should come pre-vetted from the GA project. It will reduce the number of regular DYKs, but as you say, there's (near-?) consensus on doing that anyway. Lampman (talk) 20:13, 9 November 2010 (UTC)

Not my personal opinion, just the summary of the opposition in Wikipedia_talk:Did_you_know#Discussion_.28Prop._3.29. However, in retrospect, Rjanag and Cbl32's rationales should have been included too. Correcting accordingly.--hkr Laozi speak 20:28, 9 November 2010 (UTC)

...and it may lead to socking... Oh wait, that's on there... You added that just for me, didn't you? : )

Anyway, I think the above is a decent summary of events so far.

One quibble: technically any change (even doing away with the process) is going to "increase the workload" on the short term. That comes hand in hand with change.

That said, if the DYK process was changed to resemble the format of WP:CFD (daily pages with noms for that day on a single page) I think you might have more helpers. I don't think that the review process need to be that different than the deletion discussion process. it's about achieving consensus. And on CfD (more than the other deletion discussion forums in my opinion) it's more about discussion, and ideas, than should we keep or not (due somewhat to the technical issues of changing categories being different than that of pages - which is part of why cfd stands for categories for discussion now). Oh and it also allows for automatic archiving : )

Anyway, comparisons aside, in my opinion, the workload issue is merely that the system design makes it "appear" that only a few reviewers are wanted. I think requiring more than one reviewer may prompt more participation, not less. This is about something which is to appear on the main page, after all. and isn't that one of the complaints? The immediate gratification? I would think that being a reviewer would be tapping into that sense of self-achievement as well.

And to put is another way: If the system is changed to be a bit more "user friendly" and "feel" a bit more inclusive, you might even convince me to come join in here helping too (though depending on one's perspective, that may not be considered a good thing, I dunno - lol) - jc37 21:36, 9 November 2010 (UTC)

And yes, before I get poked - I'm heading to CfD now... - jc37 21:36, 9 November 2010 (UTC)

That is an interesting proposal. Heck, why not go one step further, and have DYK mirror how AfD is formatted? I don't think (and correct me if I'm wrong) that there have been any major complaints about AfD; it's generally considered a process that works, and I would have no problem with DYK emulating it. The current format of DYK (being too cluttered) may be one of the reasons that users don't review as much as they nominate. This is a viable solution.--hkr Laozi speak 22:56, 9 November 2010 (UTC)

A single page per nom would be off putting, in my opinion. I won't speak for anyone else, but I find the AfD system to be one that leaves me feeling disincluded, and really makes it tough to follow all the discussions. But that's just my personal experience. - jc37 23:06, 9 November 2010 (UTC)

It could be transcluded, like how WP:FAC and WP:MfD handles submissions, so you'll still have all the discussions in the same place, unlike the AfD system.--hkr Laozi speak 23:25, 9 November 2010 (UTC)

While in toto, this idea seems like overkill, it would have a nice side-effect, the ability to easily create pages of hooks needing to be reviewed per-subject, along the lines of Wikipedia:WikiProject Deletion sorting/Transportation. Cheers. Haus^Talk 23:10, 9 November 2010 (UTC)

An excellent benefit if we're interested in encouraging reviewing. Making DYK structurally better in this way can only help everything. And it would also make a "vote on hooks" approach more feasible, if that was ever wanted. Rd232 ^talk 00:58, 10 November 2010 (UTC)

I don't think there needs to be a change in format to T:TDYK to encourage more collegiate reviewing. The collegiality aspect is more in the mindsets of the people involved, and it doesn't help if the people involved feel as if they're snowed under with stuff to do so can't take a few minutes to have a second look at what someone else has done. I'd also warn about the ITN experience of moving to separate pages for each day: the entire process ground to a halt, and the change was called off after four days! Don't underestimate the force of habit in human beings!! Physchim62 (talk) 01:57, 10 November 2010 (UTC)

One huge benefit of going to per-hook transclusion would be the ability to watch (and edit) individual discussions rather than having to pick through 200+ discussions to find the one you're looking for. The downside, of course, would be transcluding 200+ articles onto one page, or breaking it down into per-day pages. Personally, if we can go to per-hook transclusion, I'd strongly support it, and if per-day pages are needed to make it happen, I'd support that to. That would fix my #1 pet peeve about DYK. cmadler (talk) 21:38, 10 November 2010 (UTC)

I can't help quibbling at the disconnect between a consensus or near-consensus to reduce the number of hooks published per day, and a rejection of including any GA hooks, however few, into the process on the grounds that it would take away precious hooks! Rd232 ^talk 00:58, 10 November 2010 (UTC)

I don't see the disconnect there... For me, the disconnect is between the consensus to reduce the number of hooks approved and the greater reticence to limit the rotation of sets (say, to two per day). The slower rotation of sets is in many ways the payback to contributors for accepting harsher reviewing and (self-)limitations on nominations – the hooks that pass stay up for longer and are seen by more people! Physchim62 (talk) 01:45, 10 November 2010 (UTC)

change film studies in queue 2

Hello my name is Joseph and I am a student at Georgetown University. Currently I am part of the US public policy wikiproject in a Culture and politics class. My article Film studies is currently in queue 2 and I am requesting that it be changed to queue 4 so that the page will be live during the time our class meets. This would be greatly appreciated. ThanksJoko123nm (talk) 19:45, 10 November 2010 (UTC)

Done. - Dravecky (talk) 21:28, 10 November 2010 (UTC)

Bach cantata - again

This is a reminder (again, sorry for being boring) that the weekly Bach cantata in the Special occasions section is still unreviewed. Rlevse did the first and the last one (#22), hope someone is willing now, False world, I don't trust you BWV 52, --Gerda Arendt (talk) 12:30, 3 November 2010 (UTC)

I'll check it for you now. Bencherlite^Talk 12:40, 3 November 2010 (UTC)

Thank you! --Gerda Arendt (talk) 13:09, 3 November 2010 (UTC)

Next: O Ewigkeit, du Donnerwort, BWV 60, nom 4 November, following the new instructions not to nominate in the Special occasions. The one for the following is also ready, Du Friedefürst, Herr Jesu Christ, BWV 116, nom 9 November. --Gerda Arendt (talk) 22:22, 10 November 2010 (UTC)

Done and done. Bencherlite^Talk 14:35, 11 November 2010 (UTC)

Matsci's quid pro quo proposal

I'm thinking that whatever else we do, we could implement this idea now without too much fuss. To summarize the proposal, it will mean that everyone with more than 5 DYKs to their name will have to review somebody else's hook for every submission of their own. That alone should substantially increase the number of reviewers. We could also make it an obligation that such reviewers check the reviewed article for copyvio/plagiarism. I really don't think it's too much to ask to have someone do a thorough review of someone else's article for every submission of their own. Possibly we could extend the system to include nominators as well.

With such a system, the volunteer reviewers will only have to review the submissions of noobs with 5 articles or less, and the rest of the time they can spend double-checking other reviews. To keep track, article submitters will have to name which hook or hooks they reviewed for each hook of their own, so it could be easily monitored.

The only possible downside, as I said earlier, is that compulsory reviews might not be as rigorous as volunteer reviews. But on reflection, I don't think this is a huge risk as I think most of the committed DYKers are probably responsible enough to do a decent job. If we come across an individual who is clearly not, that can be dealt with as a separate issue.

In discussing this proposal, I'd appreciate it if users stuck to discussion of this proposal alone as I think too many of the previous threads have digressed into a discussion of multiple other proposals, which really hasn't been helpful. Thanks, Gatoclass (talk) 05:57, 10 November 2010 (UTC)

• Support, though I don't follow the 1st sentence of the 2nd para, or "Possibly we could extend the system to include nominators as well." - does that mean nominators of work by others? Johnbod (talk) 12:48, 10 November 2010 (UTC)

Yes. I think it would be appropriate to include them since they can also contribute substantially to the number of submissions. Gatoclass (talk) 13:01, 10 November 2010 (UTC)

• Support. I have followed the entire discussion (except the AN/I parts, which I couldn't find) since the beginning, but have not yet commented. There have been many well-considered and reasonable suggestions, but this has several advantages: it can be implemented more or less immediately; it properly formalises the quid pro quo arrangement by which parts of Wikipedia such as DYK operate; it should engage nominators in the project (I used to do reviews from time to time, but that fell by the wayside as I found more and more articles to write, research to do etc... for me, a requirement to review is exactly what I need); and I don't think Gato's possible downside (identified above) would actually prove to be much of a problem. Nominators familiar with DYK are likely to have a good idea of what a DYK-ready article should look like. I fully agree that such reviews should have a requirement to check for copyvio/plagiarism: these issues kicked off this discussion, after all. Hassocks5489 (tickets please!) 13:38, 10 November 2010 (UTC)

• Support - I strongly agree with the core proposal, that experienced DYK-ers submitting self-noms should review other noms on a 1:1 basis. I'm on the fence about applying the same standard to other-noms, because I think we should be encouraging experienced editors to seek out and nominate suitable articles by newer editors, and this raises a barrier, however small. There will also be some questions about implementation, but I think that can be ironed out if/when we agree that this is desirable. cmadler (talk) 13:54, 10 November 2010 (UTC)

• Support - This is basically hard to argue with, although I have slight concerns that some people who do a variety of tasks around DYK aside from reviewing might miss out, but fine. Only one question, who does the checking? Per Cbl62 below, I was assuming this was for self-noms. Mikenorton (talk) 14:57, 10 November 2010 (UTC)
  • Presumably checking this would become part of a review. cmadler (talk) 15:42, 10 November 2010 (UTC)
    • That makes sense, review one hook and then nominate your own. Mikenorton (talk) 15:51, 10 November 2010 (UTC)
• Support as to self noms only; Strong Oppose as to noms of others. Where an experienced user nominates his own work, I think the suggestion is a good one. I do not support imposing such a requirement on nominations of others' work. The process of going out and searching for DYK-worthy articles by new users is a selfless act that should be supported by the project. It would be unwise, IMO, to mandate additional burdens on those who engage in that activity. Cbl62 (talk) 15:47, 10 November 2010 (UTC)

On the other hand, users civic-minded enough to nominate the articles of others probably wouldn't mind being asked to do a few reviews. Gatoclass (talk) 15:53, 10 November 2010 (UTC)

Asked is different from mandated. I'm fine with self noms being mandated, because it's a "work-reward" quid pro quo system. If you want to have your work featured, you have to do the work. But in the case of nominating others, its a "work-work" system. In order to be eligible to do "work" that is beneficial to the project in the form of nominating and encouraging the works of new users, we are going to require that you also do other "work." Not sure if it's a principle of economics or human psychology or just common sense, but I fear a "work-work" system simply sends the message that it's not worth taking the time to nominate the work of others. If people want to do "volunteer work" for DYK, I don't think we should be imposing mandates. On the other hand, where a user wants a "reward" from DYK, I have no problem mandating that they also do the work.Cbl62 (talk) 16:21, 10 November 2010 (UTC)

Agreed; you've more eloquently made the point I was struggling to make above. cmadler (talk) 16:37, 10 November 2010 (UTC)

Well, actually, nominators do get a DYK award for their nominations - and I've always thought this a tad inequitable, given that it takes a lot more work to write an article than just nom somebody else's. So I think adding a requirement to do a review would add a little value to that award. Gatoclass (talk) 17:15, 10 November 2010 (UTC)

Yes, but it's not like the DYK awards are redeemable for cash. I agree with Cbl62, nominating others' work to be highlighted on the front page is good for the project in the same way reviewing others' hooks is: it helps others' hard work be seen, with little public recognition for themselves. I think it's overkill to require them to review in addition to their nominating. Encouraging them to review is fine. 28bytes (talk) 17:27, 10 November 2010 (UTC)

(edit conflict) "just" nom someone else's? You have to find it first! And the stats that have been compiled above don't exactly suggest that DYK is requiring a lot of work from its regular self-nominators at the moment, if they can turn out up to two DYKs a day while still eating, sleeping and performing any other necessary bodily functions. So let's have a little bit more respect for those editors who are working to introduce DYK to users who would not otherwise nominate their new articles, as opposed to those editors who have debased the DYK "award" to satisfy their Smartie addiction. Physchim62 (talk) 17:31, 10 November 2010 (UTC)

• Clarification point. Are we talking about an article-for-article or hook-for-hook quid pro quo? I suggest that it be article-for-article. A hook with 5 new articles to review takes a lot of time to review. Someone self-nominating a hook with 5 articles should have to review 5 articles nominated by others. Is that what others had in mind? Cbl62 (talk) 16:26, 10 November 2010 (UTC)

• I'd be OK with either option, but I agree that article for article makes more sense. cmadler (talk) 16:37, 10 November 2010 (UTC)

I was also thinking in terms of article-for-article. Gatoclass (talk) 17:12, 10 November 2010 (UTC)

• Support I assumed that others already did so, as a common courtesy. I think it should be article-for-article, as Cbl62 says. I would even add that "those with more than 5 DYKs are also "encouraged" to review more than the minimum single quid-pro-quo for each of their own". First Light (talk) 16:35, 10 November 2010 (UTC)

• Support. This seems reasonable. As First Light says, many of us already do this, but I can't see any harm in making it official, with the first 5 self-noms excluded from the requirement as Gatoclass suggests. I oppose the requirement for those nominating others' articles, at least for now: let's try it for self-noms first, and then expand it if needed. 28bytes (talk) 17:19, 10 November 2010 (UTC)

• I like the idea of requiring that people who want to collect DYK baubles should give back in exchange, but I have some practical concerns about this idea. My first concern is that I have encountered some contributors of good DYK content (particularly of articles translated from other languages) who are not sufficiently adept in English to do a good job of writing and evaluating hooks. I think DYK and en.wikipedia benefit from highlighting their work, but it could be counterproductive to ask them to do quality-control evaluations of other people's work. A second, and more serious, concern relates to "quid pro quo," specifically that a review quota could encourage a culture of trading favors between contributors. As a reviewer, I have been castigated on several different occasions by DYK contributors who informed me that I was taking the review process too seriously, telling me that I should "go along to get along"; that is, I should approve their work in spite of problems with it, and maybe they'd return the favor some time. A quota for hook reviews could increase the tendency to cut corners in review -- and to "trade your approval of my hook for my approval of yours." --Orlady (talk) 20:18, 10 November 2010 (UTC)

• The first thought had crossed my mind too, but they can at least check the lengths, dates etc & leave a note that a prose ?and plagiarism check is still needed. Maybe they can do 2 on this basis. Logrolling will always be hard to prevent; it's no more of a risk with this in place. Johnbod (talk) 20:34, 10 November 2010 (UTC)

• In general, I support the idea (for self-noms only, per Cbl62). However, I think that two things need to happen before this system can be implemented: (1) {{NewDYKnom}} and all related instructions must be updated to reduce the possibility that nominators will find out about this new requirement after they have nominated an article; and (2) a short and clear guide for reviewers must be written as Wikipedia:Did you know/Additional rules is too long and detailed. -- Black Falcon ^(talk) 20:48, 10 November 2010 (UTC)

A strong "yes" to both of your suggestions. First Light (talk) 20:54, 10 November 2010 (UTC)

• Support This seems very sensible all round, it's nice to see some agreement on something here for a change! I'm a little unsure as to the practicalities, and the points that Orlady has raised need to be taken into account, but presumably, we can excuse any editors who are not proficient in English. Regarding favours - I don't see how this is any different to how the system works now, so it shouldn't be any more problem of a problem (AFAIK, it isn't a problem at the moment). SmartSE (talk) 23:30, 10 November 2010 (UTC)
• Support. As Gatoclass mentioned, there is the concern of subpar reviews, but ultimately, any additional commentary and discussion on DYKs is better than none, even if it's just a quick check on the size of the article, the date, or the number of citations. Practicality is also a problem, this is a good idea but how do we implement it? Manually?--hkr Laozi speak 02:45, 11 November 2010 (UTC)
• Support. Rd232 ^talk 09:07, 11 November 2010 (UTC)
• Implemenation suggestion: To address the concerns that this would be a pain to verify, I wanted to mention that a "check contributions" link could be added to the nomination template pretty easily: check this user's TTDYK contributions. Click that, and if they have decent edit summaries, it's instant verification. 28bytes (talk) 16:02, 11 November 2010 (UTC)
• Support I usually do at least 2 or 3 reviews per nomination and I expect that regular DYK nominators should be doing this too. I think that experienced DYK contributors should spend the time to help other hooks which takes significantly less time than creating a DYK. Royalbroil 05:23, 12 November 2010 (UTC)

Queue 2: Suggested modification

Hi, just realized that there is an article "Red hair" that can be linked to in the "Wednesbury unreasonableness in Singapore" hook. — SMUconlaw (talk) 18:34, 11 November 2010 (UTC)

Linked. Materialscientist (talk) 23:19, 11 November 2010 (UTC)

Great. Thanks! — SMUconlaw (talk) 09:45, 12 November 2010 (UTC)

Prep 3

We're getting a lot of hooks about the Free Territory of Trieste, which is a good thing, but 2 of the hooks are in the same prep 3. Is it possible to put them in different queues? Yoninah (talk) 08:57, 12 November 2010 (UTC)

Good catch, I'll swap one out. 28bytes (talk) 10:47, 12 November 2010 (UTC)

Done. 28bytes (talk) 11:07, 12 November 2010 (UTC)

Actions speak more loudly than talk page promises

• Template talk:Did you know#Geoffrey Crawley

So … after all this talk above about discouraging DYK entries that are padding and so forth, is it going to actually be put into practice? What are you going to do about, for example, a DYK hook that was in a featured article six months ago? Bear in mind that how much is in fact done, by the people who say that the problems that have been discussed recently can be fixed without major changes, will be a measure to everyone else of how much more pressure should be applied. Uncle G (talk) 22:37, 11 November 2010 (UTC)

• Erm, I hardly think this one is a fair criticism of the DYK process. For a start, it's not been reviewed yet. Secondly, you are assuming that all DYK reviewers should be aware of all featured article content, which is obviously ridiculous. Thirdly, we already know that featured article content is far from perfect, whatever the FAC regulars will shout at any opportunity, so maybe an additional check on this factoid is in order ;) Physchim62 (talk) 22:50, 11 November 2010 (UTC)
  • Kiddo, you're constructing straw men. The featured article that this was already in six months ago is linked in the hook. Uncle G (talk) 01:15, 12 November 2010 (UTC)
    • Please stay civil. The fact that the FA is linked is not relevant as we do not require the hook to be truly original (for better or worse). Whether or not we should AGF that the nominator did not know it was featured on MP is a separate issue. He might reply to you "good catch. Thanks. Here is an ALT". I don't see anything criminal here. Materialscientist (talk) 01:22, 12 November 2010 (UTC)
      • That was perfectly civil. Try not to confuse pointing out a strawman argument for what it is with incivility. I also refer you to the discussion above where people have stated that bad hooks should be refused. Physchim62 is, ironically, one such person. Here's a chance for those who have opined that to put their principles into practice.

        One incidental question: You and Physchim62 are both acting as if not even reviewing the fact stated in the hook against the articles it is linked to is within the remit of DYK, because it's somehow ridiculously difficult to do. I have to ask: What on Earth do you think is within the remit of reviewers in that case? Because it seems like nothing at all, if not even basic checking of the hook against the articles is out of scope for reviewers, for being too hard. It's unfortunate that Physchim62 has pre-supplied the phrase "utterly pathetic", in boldface no less, to associate with such a review process. Perhaps xe should think about removing the boldface before the association to the DYK review process sticks. ☺

        Has 400KiB of discussion been for naught if even this very basic thing is deemed to high a hurdle for DYK to jump, and the people who want change don't actually do anything about it? Uncle G (talk) 01:49, 12 November 2010 (UTC)

        • A simple look at that nomination shows that it has not been reviewed at all. For you to immediately bring your problems to the talk page is disruptive, and shows the arrogance of someone who has spent too much time in the toxic environment of WP:FAC. I suggest you get some fresh air and do something to actually improve the online encyclopedia that we all love. Physchim62 (talk) 02:06, 12 November 2010 (UTC)
          • Are these repeated personal attacks based upon some downright foolish presumptions a sign that you don't have an actual argument? Resorting to personal attacks usually is, I observe. You certainly don't have your facts straight. Here's one that's not. Kiddo, it's you that isn't doing anything to improve the encyclopaedia here. I'm the one that just checked a hook, remember. Uncle G (talk) 02:15, 12 November 2010 (UTC)
            • Personal attacks? Sorry if you feel that way, old man, but I was merely looking out for your health. Thanks for stooping to leave a comment at T:TDYK, it is certainly appreciated. Physchim62 (talk) 02:29, 12 November 2010 (UTC)
              • That's a clear "Yes." answer to that question, I see. Uncle G (talk) 15:24, 12 November 2010 (UTC)
• I would add that the response at WT:FAC to their own quality control problems has been utterly pathetic compared to the discussions that have gone on on this page. What's the FAC solution? To get people to "sign" a statement saying their contributions aren't copyvio. Unworkable, useless, completely ridiculous. Maybe the FAC regulars should come down form their supposedly ivory towers and see real-life quality control at work: they might even learn something, although I doubt they'd admit it. Physchim62 (talk) 23:01, 11 November 2010 (UTC)
  • So your actions, which speak more loudly than talk page promises, are, when something at DYK actually is drawn to attention, and the time comes to actually put some of these much-discussed principles into practice, to do nothing and divert discussion onto a hobbyhorse about FAC. You might have said that the do nothing option is "not viable". Yet here you are doing nothing. You've "support"s all over the notions of rejecting hooks and more stringent review. But the first time that it comes to enacting that … you take no actual action. It's put-your-edits-where-your-talk-page-promises-are time. Uncle G (talk) 01:15, 12 November 2010 (UTC)
    • For all my criticisms of DYK – and I'm fairly well known around here for my criticisms of "Did you care?" – the current discussions offer the possibility of consensual root-and-branch change that will benefit the entire encyclopedia. No sign of any movement at "Featured crap", I'm afraid, dispite my efforts to be constructive at WT:FAC. Your contribution here can be classed as constructive: the above complaints, simply because your comments weren't acted on within the four minutes that separated the edits, are simply pathetic, for not to say completely hypocritical given your long-standing experience of FAC. It takes a lot for me to stand up in defense of DYK, but your criticisms here are way out of line. Physchim62 (talk) 01:35, 12 November 2010 (UTC)
      • Kiddo, you're now confused — confusing me with someone who is hypocritical for starters. You should really get your facts straight before leaping into the personal attacks. I'll let others clue you in. My complaints are not that the nominators didn't act within four minutes but are aimed directly at you, the person who complains about DYK, but when push comes to shove and it is time to actually put your principles, discussed at length on a talk page, into practice, you do nothing. You've still done nothing. You've not found a new hook, suggested anything, or even added a supporting or opposing opinion to the entry. And you're one of the people who wants change. I said it before and I repeat: What you do will speak more loudly than this talk page discussion; and you will be known by what actions you actually take to change DYK in the way that you want it changed, especially when cases in point come along right as you are having these KiBs of discussion. You do nothing. Uncle G (talk) 01:58, 12 November 2010 (UTC)
        • Old man, do you think you're doing anything with these comments? I've spent much of the last ten days trying to make DYK a better process. I could simply have filed a RFC to get it shot off the Main Page, but no, I have spent my time in trying to convince people to work in what I see as a better way to fulfill their objectives, and where those objectives are unfulfillable, to explain why there are choices that have to be made. I take it as a personal insult when you insinuate I have done nothing in this process simply because I haven't made a pointy rejection of one single nomination that you didn't like. Physchim62 (talk) 02:17, 12 November 2010 (UTC)
          • If the point that you've done nothing stings, then good. It should. It should spur you into actually putting your espoused-at-length principles into actual practice. You've spent much of the last fortnight talking, but the first time it comes to putting your principles into action: nothing. And that's the point of course. Lots of fine principles are bandied about, but one isn't known by the lengthy espousal of fine principles. One is known by what actions one takes. You're all known by the inaction when it comes to recycled DYK content that isn't "new and novel" as the guidelines put it. It's all OK, you opine. Just wave it down the DYK conveyor belt like everything else. Hook checking, when it comes to your actually discussing a concrete hook, is a mere formality, a matter of using little ticks and crosses. You don't even try to look for alternative, new, hook material.

            You're known by the fact that when it actually comes down to brass tacks, hundreds of KiBs of discussion are worth exactly the not-paper they are printed upon, because the people who want change won't actually do anything about it, and the people who think that they can correct problematic DYK standards without changing the system at all don't do anything either (except try to stifle discussion of that very point). As I said, right at the start of this section, your inaction, on both of your parts, will have fairly obvious consequences.

            Uncle G (talk) 15:24, 12 November 2010 (UTC)

    • If your comment about my "'support's all over the notions of rejecting hooks and more stringent review" are meant to be a personal criticism, then the response is quite simple: I cannot reject hooks for being too boring until I'm sure of the consensus behind it, otherwise I would (rightly) be accused of disrupting Wikipedia to prove a point. Physchim62 (talk) 01:40, 12 November 2010 (UTC)
• I thought this would be about plagiarism or something. In fact it is (as far as I can see) that the hook fact is already mentioned in the FA article, and in a linked article. That the hook fact needs to be something not referred to anywhere else in WP is a wholly novel idea. If you support this you should suggest it here. I won't be supporting, as I don't think we should assume that our readership has read (& remembered) the whole of en:WP. In the same way the fact that a particular reader might already know a hook fact is no disqualification so long as it is not common knowledge. The hook is the most obvious one for the article & seems ok to me. Johnbod (talk) 02:31, 12 November 2010 (UTC)
  • Your "novel idea" description is wrong, as previously explained, and as mentioned by the very guidelines you point to. And you're constructing the same straw man that Physchim62 tried to construct above, when in fact this was a simple case of comparing the hook against the articles linked to. It is saddening that when it comes down to it, the desperate reaches with which doing nothing is actually defended, even by the people who actually want something done, extends to deforming the guidelines completely out of shape and throwing away even the fundamental idea that hooks should be compared against the very articles that they link to. Practice is woefully at odds with principle. Uncle G (talk) 15:24, 12 November 2010 (UTC)
• Mud-slinging exercises are the last thing we need right now. Kindly knock it off folks. Gatoclass (talk) 06:54, 12 November 2010 (UTC)
  • Pointing out that the upshot of this is that no-one does a thing in practice (rather different to mud-slinging) is very much what you need, however much it stings and however much you don't like it. Try to stifle this, and the people who complained so loudly about DYK being a problem will be back and putting more pressure upon you to do something. You've got at least three problematic entries on your conveyor belt right now. People know that poor review is a problem. People discuss at length, with two polls, multiple sections, and hundreds of KiBs of wikitext that poor review is a problem. But point you to problems of poor review, and do you work on them? No you don't. You even try to hide where this is pointed out. In the meantime, your discussion of the problem at hand has spiralled so far in upon itself that a major topic now is whether a particular editor can summarize a point that someone else has made. You're off-track and you're not doing a thing in practice about the actual problems. The conveyor belt is still running at full tilt and you are still waving things through. And you're even defending that when it comes to actual cases. Uncle G (talk) 15:24, 12 November 2010 (UTC)

I repeat; you are misreading the rules. Plain & simple. Johnbod (talk) 15:47, 12 November 2010 (UTC)

If you want to slow the "conveyor belt" down, you can either join in the attempts to reach consensus above, or you can start an RFC somewhere else. Playing around with an active nominations page is not the way to make your point. Physchim62 (talk) 15:56, 12 November 2010 (UTC)

I disagree. The nominations page is exactly where problematic hooks should be addressed. Kudos to Uncle G for getting in there and pointing out problems with individual hooks. This thread, however, is way over the top and should be re-hatted. 28bytes (talk) 16:33, 12 November 2010 (UTC)

The problems have to be real ones though, not problems that are invented simply for the purpose of making a point against DYK. Several editors, myself included, have looked at that nomination and failed to find any problem of the type that Uncle G pretends to exist. Physchim62 (talk) 17:07, 12 November 2010 (UTC)

As I have already added in the discussion of this article on the DYK page itself, the Geoffrey Crawley article was created independently of the Cottingley Fairies article based on multiple reliable and verifiable sources about Crawley's life that happened to discuss the story of his role in debunking the Fairies as a hoax as a major part of his life story. That the Crawley article would have overlap with the Fairies article is inevitable based on the fact that the sources focus on the story of his role in the Fairies incident. No material was taken from the Fairies article nor was anything "padded". The fact that Cottingley Fairies has been reviewed as a featured article and appeared on the main page several months ago has no relevance to the creation of this article nor its nomination to appear on DYK. I still fail to see what the issue is here. Alansohn (talk) 16:53, 12 November 2010 (UTC)

Uncle G's novel interpretation of the DYK rules is that the hook fact must not have been included in any earlier en:WP article. Johnbod (talk) 17:50, 12 November 2010 (UTC)

Fine, we can always add that to the list of proposals above and see if it gets any support ;) because it's certainly a novel interpretation of the rules! Put me down as "oppose", as practically unworkable and irrelevant to the readers of the DYK section. Physchim62 (talk) 18:39, 12 November 2010 (UTC)

Queue 1, Hook 7, List of first overall NBA Development League draft picks

Can anyone add the link to the NBA Development League All-Star Game into the word "D-League All-Star" in the 7th hook of the queue 1? Thanks. — Martin tamb (talk) 16:01, 12 November 2010 (UTC)

 Done Gatoclass (talk) 17:41, 12 November 2010 (UTC)

Truth in Numbers?

I would appreciate it if this DYK could please be reviewed and confirmed.

1. Regarding DYK hook located under Template_talk:Did_you_know#Articles_created.2Fexpanded_on_November_10.
2. I have expanded the article, not just five times but over 10 times. From this [4] to this [5].
3. The hook is interesting and noteworthy, that the former director of the Central Intelligence Agency provided commentary.
4. The article is notable - it had an AFD that closed recently, as Keep, see Wikipedia:Articles for deletion/Truth in Numbers? Everything According to Wikipedia.
5. The hook is verifiable and cited to an online secondary sources, see [6].
6. The DYK hook meets all of the criteria. I put in much effort to the page which was from a post-AFD stub-class-state, to expand the article and improve its quality.
7. Respectfully request that this DYK entry be reviewed and confirmed.

Thank you for your time, -- Cirt (talk) 04:47, 13 November 2010 (UTC)

It's in the queue so it will get reviewed in due time. (The nomination was made barely two days ago.) - Dravecky (talk) 06:51, 13 November 2010 (UTC)

Upside down queue

Whoa, where did this come from? I realize I don't spend a lot of time here, but as someone who spends a lot of time on the nom page, how did I manage to miss this change?

Anyhow, I like the idea, but it does make entering a nom much harder, imho. I have to hunt around for the proper area.

Is there any reason this couldn't be automated? When you click "enter a nom" a small page shows up just for that nominator, where they can select from one of the templates and enter the data in fields. Then it figures out where it should go on the page.

Maury Markowitz (talk) 17:23, 13 November 2010 (UTC)

Wikipedia talk:Did you know/Archive 59#Thoughts on DYK noms, including the declining rate. It's impossible for templates to conditionally place an edit in a particular spot. Shubinator (talk) 17:32, 13 November 2010 (UTC)

Don't use a template then. Maury Markowitz (talk) 17:40, 13 November 2010 (UTC)

The above discussions (just a small point)

Since most of the above discussions are about overhauling DYK, should they be moved to a separate page when archived or whilst they are currently going on? Or split up between the many current archives? Simply south (talk) 18:17, 13 November 2010 (UTC)

DYK Hall of Lame: back again

The first one's a corker:

... that, according to legend, Monte Titano (pictured) in San Marino was given as a gift to Saint Marinus, a Croatian stonemason fleeing anti-Christian persecution, who established a hermitage there?

1. The point of it is that a whole mountain was given as a gift to someone. But from the hook itself this is impossible to convey unless the poor reader either (a) knows that "Monte" is Italian for "mountain"; (b) clicks on the link (you shouldn't have to to get the point of the hook); (c) by some chance works out that the tiny little thumbnail is a pic of a mountain—its unusual ridge feature makes it very hard to see this, and there are more suitable pics in the article, or (d) clicks on the thumbnail to see an enlargement (you shouldn't have to divert somewhere else to get the point of the hook). Why isn't it spelled out plainly? Readers will just not bother when a hook falls flat like this.
2. Why are nine words linked in this one sentence? Isn't the objective to get the reader to visit the themed article, i.e., Monte Titano, rather than splashing lots of other "chain" links into the hook? All of the other links are at the themed article, in full view, easy to access. Has anyone thought about why a reader would want to return to access "stonemason" from the main page hook rather than from Monte Titano?
3. Why isn't "hermitage" simplified to "religious retreat"? The term "hermitage" is still used in the article (and linked). Readers should not have to go on expeditions to fathom what the hook means.
4. The nationality of the stonemason is included (Croation), but the location of the mountain itself, in Italy, is still a mystery. Not only that, the location is missing from the lead of the article until you read for 30 seconds or so. Tony (talk) 10:40, 12 November 2010 (UTC)

Despite the off-putting section title you provided, I think these are all some of these are excellent suggestions for improving that hook. If this were in a prep queue I would go ahead and make those changes. Unfortunately since it's on the main page and I'm not an admin, someone else with the bit will need to do that if they agree.

For future reference, this page shows the hooks that will be going on the main page in the future; there's usually a lead time of at least 36 hours, so checking those for problems will give us a chance to fix them before they appear on the main page. Thanks, 28bytes (talk) 11:19, 12 November 2010 (UTC)

Note that someone has removed "Croatian" from the hook, I think as part of the usual nationalistic wrangles. Johnbod (talk) 11:24, 12 November 2010 (UTC)

The mountain is not in Italy. As identified in the hook and in the first sentence of the article, it is in San Marino. MANdARAX • XAЯAbИAM 11:30, 12 November 2010 (UTC)

Indeed. Has this been changed now? We have another Religion in San Marino hook up. Points 3&4 are pants frankly, & I'm not that impressed by 1 & 2 either. Johnbod (talk) 12:28, 12 November 2010 (UTC)

Yeah, it's all moot now, this was the previous set of hooks that was replaced during the normal update 45 minutes ago. 28bytes (talk) 12:45, 12 November 2010 (UTC)

Which is why I would reiterate to anyone interested in helping DYK improve hooks that the best time to do so is before they go up on the main page. 28bytes (talk) 12:50, 12 November 2010 (UTC)

(e.c.) Hence the feeling that threads like this aren't meant to be constructive. A constructive contribution would comment in the 36 - 60 hours the hooks spend at T:DYK/Q, or at T:TDYK, and not be launched as a 'gotcha' when the hook is most of the way through its time on the main page. EdChem (talk) 12:54, 12 November 2010 (UTC)

Well, since the OP has never edited TT:DYK, I simply assumed he was unfamiliar with the fact that the future prep set queue was available for anyone to view and comment on before going to the main page. I'd prefer if he'd pitch in and help review or improve hooks on the nominations page rather than just telling others to fix it, but he's certainly within his rights to point out problems. My only point was that pointing out problems earlier, before they go on the main page, would be in everyone's interest. 28bytes (talk) 13:28, 12 November 2010 (UTC)

• I'm sorry to comment after the fact, but I've never explored the DYK process. I can only see what's on the main page. There is a feeling in many threads above that everything is tickety-boo with DKY hooks. The purpose of my comments was to point out why this is not the case. It is just the first one on the list. I don't dare to look further. Systemic fixes are required, not just fixes to this particular example. Tony (talk) 15:24, 12 November 2010 (UTC)
  • Tony, no one is saying there aren't problems. I'm sure you see both above and below this section kilobytes and kilobytes of discussion of what the problems are and how we might fix them. 28bytes (talk) 16:17, 12 November 2010 (UTC)
  • Tony, I have to emphasise, you really should go through the DYK process first before saying that "systemic fixes are required". It's not that systematic aren't required, but a statement like "I've never explored the DYK process" is very concerning.--hkr (talk) 22:58, 12 November 2010 (UTC)
    • It's far easier to make broad, sweeping judgments about a topic you don't understand than about a topic you know in detail. --Orlady (talk) 23:06, 12 November 2010 (UTC)
      • And it's far easier to see the inconsistencies from outside the process than from inside. Physchim62 (talk) 23:17, 12 November 2010 (UTC)

←I desperately need to go through the process ... yes. But looking at the product, in a way, is all that matters from the perspective of pointing out areas in which, down the chain somewhere, things need to change. I want to add that I find the newness restriction (five-day-old articles) means we are showcasing relatively raw articles, or using only those that have been prepared elsewhere and thrown into article space in a mature form. Is the motivation to encourage article creation (rather than refinement) still as important as it was a few years ago ... we have 3.2 million of them. And effective hooks need a large pool of articles if a constant stream of interesting facts is to be used.

My other immediate issue, easy to fix, is that the hooks seem to be stuffed full of wikilinks. The problem with this is not that they trivial, but that the only link we want people to follow, initially, is the themed link. That link-target will presumably have all of the other links in it, usually at the opening or in the first part of the article. I pointed this out in relation to "stonemason" above. And "Croation". In fact, in all cases other than Monte Titano. Readers are much more likely to follow links in the full context of the article than from a short hook that is speckled blue/black on the main page. The appearance issue also counts, apart from the logical use of linking. The analogy is at The Signpost's Wikipedia:Wikipedia_Signpost/2010-11-01/Features_and_admins︱F and A page, where in the blurbs for FAs we tend not to link anything but the article name. Why? Because we want readers to go straight to the article to find out more, not to host of different articles. Tony (talk) 05:02, 13 November 2010 (UTC)

Frankly Tony, if you had followed the San Marino link you would have known not to make your erroneous point #4 above! I agree there is some overlinking, but a linked "hermitage" is better than a very vague and probably not accurate "religious retreat". Johnbod (talk) 08:55, 13 November 2010 (UTC)

I agree, I'm not debating the merits of the other arguments, but some of the points are invalid or slightly pedantic.--res Laozi speak 09:12, 13 November 2010 (UTC)

Yes, but very few people comparatively speaking read the Signpost. There are probably hundreds of thousands if not millions who see Wiki's main page, and you can't assume they are all interested in the same thing. We are an educational project after all, if someone is curious to know more about Croatia or stonemasonry, we should accommodate that. Gatoclass (talk) 05:49, 13 November 2010 (UTC)

The exact absolute and relative readership of Signpost:Main Page is irrelevant. In the example above, it was just ridiculous to stuff that one sentence full of links. I count six... surely a maximum of three would do fine. The objective of the hook is to lead readers to the article just created, but if the reader goes elsewhere than the intended article, that represents a lost opportunity. Yes, I am assuming that once they have drifted away, the majority don't come back – it's a reasonable assumption to make. --Ohconfucius ^¡digame! 09:42, 13 November 2010 (UTC)

Naughty readers, not going where they're supposed to! Perhaps we should just remove all links except featured content from the Main Page... Oh and we can petition the developers so the Search box is removed from the Main Page, that way readers will have to click on a link to featured content to be able to do their search! And then if we remove all the links from all the other Main Page sections except DYK, we can guarantee that each hook will get, on average, 140k hits (that's about the same as a fairly popular ITN story at the moment)! Problem solved!

Sorry, minor stylistic elements like the number of links in a hook are utterly insignificant compared to the catastrophe which are DYK's click-through rates. Why does the average DYK hook get less than a thousand hits? Because DYK is treating its readership with contempt in the creation and selection of hooks. Simple as that, sorry. Physchim62 (talk) 10:57, 13 November 2010 (UTC)

Are they really such a catastrophe? I've taken a look through some FA page hits recently - I was shocked at how low their numbers are. But an FA hogs a quarter of the page and is up for 24 hours. If you really think the job is to score as many page hits as possible, wouldn't it be better to just scrap FA and use its cyberestate for something more interesting? Gatoclass (talk) 17:45, 13 November 2010 (UTC)

That everything else gets less than ITN is no reflection on the rest if people choose to use WP as their newspaper. I didn't think Phschim made his point on this last time it was discussed - despite a worse position on the page total DYK figures seem to average higher than TFA - and personaaly I wish he'd avoid hyped-up language on this matter. Johnbod (talk) 18:43, 13 November 2010 (UTC)

DYK as a whole – that's 32–36 different hooks on different subjects, gets slightly more hits that TFA, a single article with a long prose summary of the contents (probably sufficient information for most readers), obviously on a single subject (and so inevitably not to the interests of the majority of a random group of people), plus three plain links below it (which also get considerable numbers of hits, far more than most DYK hooks). The damning comparison is with OTD: DYK has a hugely resource-intensive process to produce all these hooks, and yet OTD hooks (from a process run on a shoestring) are still twice as popular with our readers. People aren't using OTD as their newspaper, after all! And it is arguably in a slightly worse position than DYK on the Main Page. I'm sorry if Johnbod gets upset when I call a spade a spade, especially as he has worked hard to try to get a practical improvement in the situation. Physchim62 (talk) 18:56, 13 November 2010 (UTC)

ITN vs DYK is an invalid comparison because people are searching for the event through other means as well. Phys, you may have a point, I don't really know... but the use of hyperbole turns people's ears off. Try stating your points with more seriousness and less drama. You may be surprised. Ed ^{[talk] [majestic titan]} 19:36, 13 November 2010 (UTC)

OTD gets twice as many hits as DYK? You've actually done some research on that? I just picked a day at random and compared the total number of hits - OTD got 65k over 24 hours and DYK got 39.2 over 12, which would average out at 78.4 over 24. So on that particular day, DYK came out ahead, and I have to say with a pretty ordinary collection of hooks. Are you sure you're not just making this up as you go along? Gatoclass (talk) 19:54, 13 November 2010 (UTC)

Well yes. Here are some statistics for click-thorugh rates covering ten days in January for all Main Page sections. The median click-through rate for DYK hooks was 1.0k; there were 39 sets of hooks (one day only had three sets), so we multiply that figure by 3.9 to get a notional click-through rate if the hooks had been allowed to stay up for 24 hours: that give 3.9k "hits"/24h. Over the same time period, the median click through rate for OTD hooks was 7.8k/24h, exactly twice as high. Note as well that the median click-through rate for the entire sections (that is, summing the click-throughs for all the hooks) over that period was 64.9k for OTD but only 44.3k for DYK. Physchim62 (talk) 14:17, 14 November 2010 (UTC)

Why would you want to be fiddling with "medians" when you have absolute numbers to compare? Sorry, but I'm rather sceptical of this result. Gatoclass (talk) 15:04, 14 November 2010 (UTC)

Because using medians is a standard statistical technique for dealing with data sets which have a "high-end tail": that is, no hooks will get fewer than zero hits, but some hits will get an unusually large number of hits. The arithmetic mean gives more weight to the exceptionally high results than the median. To illustrate this, let's compare the arithmetic means for that ten-day period: DYK 1.535k hits per hook, OTD 10.477k hits/hook. You still see the same effect, although now OTD hooks are only 1.75 times as "popular" as DYK hooks (after accounting for the difference in visibility time). Physchim62 (talk) 16:46, 14 November 2010 (UTC)

Note also that many of the hits on ITN items have nothing to do with the main page, but are generated by people who are looking up items that are in the news. I'd wager that the same thing is true of OTD items, albeit to a lesser extent -- news media and other non-Wikipedia outlets highlight significant events of the day, and that undoubtedly leads some people to the relevant Wikipedia articles. On the other hand, when a brand-new article on an obscure topic gets 2,500 page views on the day it appears in DYK, you can be very sure that almost all of those hits are attributable to DYK. --Orlady (talk) 02:05, 14 November 2010 (UTC)

I quite accept that a straight comparison between DYK and ITN is impossible, for the reasons stated. To give an example, one ITN story has generated hits equivalent to more than the number of hits on the Main Page that day! Obviously that has nothing to do with anything ITN was doing! But ITN does run more obscure or quirky news stories from time to time: see Gävle goat or hijra for a couple of examples. It's difficult to set a hard criterion for what is a "quirky" ITN story, so I shaln't attempt to make a more formal comparison. And ITN falls flat on its collective face from time to time, but I tend to 'define' that as an ITN story that only generates a peak of 10–15k, far from the sort of numbers that DYK sees as normal. Physchim62 (talk) 14:17, 14 November 2010 (UTC)

I don't think we should be spending too much time analysing what is essentially a side issue, but I can't help observing that OTD contains articles which are normally on well-known and "high interest" subjects, while DYK articles these days are almost all by definition not on such subjects. Johnbod (talk) 02:28, 15 November 2010 (UTC)

• "the catastrophe which are DYK's click-through rates"—well, it's plain as day to me that the poor click-through rate is caused by diluting the object of the hook among many other links. Those other links are at the target article. If you don't direct the readers to it, of course they'll react just as they've been proved to react when confronted with too many choices in real life—they're less likely to choose anything; it's a turn-off. Honestly, when I look at a messy blue/black hook, I'm wondering which link is the subject. If you link just once in each hook, you will not only get higher click-through rates, but you will make it look a heck a lot more attractive to readers. Tony (talk) 02:14, 14 November 2010 (UTC)

Overlinking in hooks

Somewhere in the masses of discussion, the issue of whether we should include links for other articles in hooks was mentioned. Despite the way it was raised, I thought the point was worth a considered discussion. Only linking the 'hook' article(s) will draw attention to the article being highlighted, and any other links in the hook are almost certain to be included in the article as well. However, it removes the option of piped links for acronyms (like MMA, for example) or readers going directly to articles they choose. It seems to me that this topic is worth covering in new draft guidelines. Thoughts? EdChem (talk) 10:54, 13 November 2010 (UTC)

That's the least of our worries at the moment. I certainly would not support only linking the hook article - Tony, who introduced the subject above (Hall of Lame Revisited) neatly demonstrates, without intending to, why this would be a really bad idea. But we sometimes have too many mundane links in hooks, & at other times too few. Johnbod (talk) 11:03, 13 November 2010 (UTC)

I don't understand my neat illustration. Maybe I'm being thick. Tony (talk) 12:05, 13 November 2010 (UTC)

One of the links you complained about would have told you that San Marino is an independent country, not part of Italy. You then went on to complain that the location in Italy was not mentioned. Looks neat to me. Johnbod (talk) 13:56, 13 November 2010 (UTC)

Mmm ... I'm thinking about how readers of the main page actually read these hooks. I do not think they are as "studied" in their approach as this implies—I'd approach hook writing with the goal of trying to get across the gist in the five second it takes to read the anchor sentence; and I would try not to require the reader to hit a link to find out the basics. I'd expect on the eastern ?something of the Italian peninsula (a phrase I think the article uses, got no time to look right now) to be recognisable to most readers. I'd add that to the sentence and unlink "San Marino". I just do not think readers read hooks in a zigzag to and from wikilinks to get that info. What will happen is that they'll pass over the whole thing because it falls flat on first appearance. We could, actually, organise a test of volunteers to determine just how they do read the hooks. I'm pretty sure I'm right (how arrogant). Tony (talk) 14:29, 13 November 2010 (UTC)

Of course people don't normally hit all the links, but if they are uncertain, or as ignorant as you were here, then they can. The linking should be just the same as in normal article text but, like articles, tastes differ & some have too few and others too many. Johnbod (talk) 14:32, 13 November 2010 (UTC)

Note that only one of the hooks in the current queue (including prep areas) lacks internal links apart from the title article, so it would be a substantial change. Mikenorton (talk) 15:41, 13 November 2010 (UTC)

I don't see why we should prevent Main Page readers from clicking on links that interest them. It's up to DYK to come up with hooks that make people want to click on the bolded article. Nor am I convinced that it makes any practical difference to the click-through rate. I looked for an example on DYK of a phenomenon we see at ITN: that, when there is a person's name linked in the hook, it gets a big peak in hits, sometimes more hits than the "featured" link. I found "referring to Che Guevara T-shirts and other clothing, Aleida Guevara said that Che 'probably would have been delighted to see his face on the breasts of so many beautiful women'?" The featured hook got 10.2k (significantly better than average, even for the top slot) while Aleida Guevara got 8.9k. You can't simply say that the 8.9k hits would have got to the featured link: what about people who simply didn't want to read a whole article about Che Guevara T-shirts? Or the people who read both articles? Physchim62 (talk) 19:33, 13 November 2010 (UTC)

Yes, I think this is a non-issue. IMO we should be linking as many articles as we reasonably can on the mainpage, as we are an educational project and we should be making our information as accessible as possible. Gatoclass (talk) 20:32, 13 November 2010 (UTC)

I agree with Gatoclass -- it's not necessarily a bad thing when people click on "other" links in the hook. It's quite reasonable that people would click on Aleida Guevara -- wanting to know who she is. My experience tells me that it's not just people's names that get more attention than the intended target -- it's anything in the hook that people are curious about. One of "my" DYK hooks was "... that the element promethium was discovered in 1945 by Manhattan Project chemists Jacob A. Marinsky, Lawrence E. Glendenin and Charles D. Coryell?" -- although the featured new articles were about people, it was hardly surprising that the article on promethium got far more hits than the articles about the people. Regardless, the hook caused some people to poke into corners of Wikipedia they might never have visited otherwise. --Orlady (talk) 01:49, 14 November 2010 (UTC)

Indeed. I just noticed that the figures for my last DYK, 2 days ago, were (day before:dyk day): "... that the "secretive character" of the Beggar Looking Through His Hat (pictured), attributed to 17th-century artist Jacques Bellange (21:1,700), may have appealed to its former owner, American KGB spy Michael Straight (36:3,100)?" Is that a bad thing? Should I have omitted the 2nd link? Of course not! Johnbod (talk) 02:34, 15 November 2010 (UTC)

We must be getting something right!

The current set on main page, headed by "Americanization School" has the best overall set of hooks I can recall seeing for a long time. I don't understand the NBA one at all, but that's normal and no doubt plenty do. I also lived in Polkerris] for 6 months which redeemed a rather less interesting one, for me anyway. Johnbod (talk) 00:51, 14 November 2010 (UTC)

A lot depends on the effort that goes into creating a balanced update. I know there are lots of people who think that DYK should be more "interesting", but what many of them don't realize is just how many editorial decisions have to be made to build a quality end product. Unless you have some really dedicated people around who are willing to make the effort, update quality inevitably suffers. Gatoclass (talk) 07:20, 14 November 2010 (UTC)

Average of less than 1000 views per hook?

I have begun an investigation to consider whether the claimed statistic of less than 1000 views per hook is supported by a detailed analysis of the views of the articles in a hook. I started with a preliminary look at three hooks in which I was involved. No claim of these being representative is made, but I found the preliminary results sufficiently surprising to think them worth sharing. The details are at User:EdChem/DYK clicks. I think the results support putting the effort into examining all hooks of a single day. Before I go to that effort (as it is time consuming), I invite comment on whether my methods strike others as reasonable / supportable, and whether the potential data are worth taking the time to collect. EdChem (talk) 11:25, 14 November 2010 (UTC)

How are you doing this? By looking at the spike in page-view traffic? That might be reasonable, but you'd need to look at quite a few hook-days to get reliable trends. Whatever the number, I can show you exactly how to triple the numbers instantly. Tony (talk) 12:46, 14 November 2010 (UTC)

These are the stats I collected in January. Physchim62 (talk) 13:34, 14 November 2010 (UTC)

I have now collected and input the data for one complete update set of nine hooks. Obviously more sets need to be analysed to see if the results are typical or not, but the short version is: This single set of 9 hooks attracted 21,100 views for the bolded articles, and a further 21,954 for the non-bolded articles. This is from a 6 hour period, yet the totals approximate those for a full day from Physchim62. There are, of course, many possible explanations, the most obvious being the set I used was atypically effective. Note that I consider the non-bold views under-estimates because in cases where a clear estimate of additional views can't be made, I have taken them to be 0. Also, in most cases, I have deducted the average daily views for the rest of the month from the 'spike' in views, so that only the extra views reasonably attributable to DYK views are shown. In other cases, the views show clear patterns over each seven-day period, allowing an estimate of the expected views absent DYK on a given day to be made. I have generally gone for a conservative estimate, so that it is possible that there are more DYK-induced views than I have estimated. Undoubtedly, much more data are needed to support a conclusion, but I am sceptical that the estimate of less than 1000 views per hook is at all accurate. EdChem (talk) 14:19, 14 November 2010 (UTC)

You are including hits on non-bolded hooks, which I did not: all the figures in my table would be higher if non-bolded hook hits were included. However, in terms of "project effort" in creating the hook, only the bolded link counts (unbolded links are rarely checked for anything at all). Also, our treatment of multi-article hooks are different, and EdChem seems to be looking at and arthmetic mean (I used medians). Physchim62 (talk) 14:28, 14 November 2010 (UTC)

Physchim62, though I don't concur that non-bolded views should be disregarded (as I think it is clear they come from DYK), I have carefully presented views from bolded and non-bolded articles separately before presenting the aggregate. Even on just the bolded views, the two update sets I have now looked at average about 2000 views per hook (across 2 consecutive sets and 18 hooks, bolded views = 35,624 and non-bolded views = 42,803), and no single hook has less than 1000 bolded views. The complete data set (2 updates, 18 hooks) remains too small to support any definitive conclusions, but do you accept that my data are (so far) inconsistenet with your 'less than 1000 views per hook' result? Remember that I am discounting views by (typically) the average views for the article for the rest of the month, so am only counting views over and above what might have otherwise occurred. As for means v. medians, I recognise the statistical reason for preferring medians, but I am so far presenting mostly raw data. When there are sufficient sets analysed, the resulting data can be analysed to see whether it is sufficiently close to normal to justify description by mean and standard deviation, or instead by median and inter-quartile range. EdChem (talk) 05:34, 15 November 2010 (UTC)

If you want to measure total DYK impact, you should also include hits on the photo featured with the lead hook. I have sometimes found spikes of 10,000 hits or more on the photo. Cbl62 (talk) 06:31, 15 November 2010 (UTC)

For example, the Ophiophagus hannah image accompanying the one you started to work on had 3,200 hits while on DYK. Cbl62 (talk) 06:33, 15 November 2010 (UTC)

Thanks, Cbl62 - I was including image views in the non-bolded views, though I can see how a case can be made that they be included as bolded views for the lead article of an update. EdChem (talk) 11:55, 15 November 2010 (UTC)

There are three reasons for my choice of methodology, one statistical, one Wikiprocess-related and one practical.

• It is impossible to eliminate double-counting if you include more than one link. How many of your summed hits correspond to the same reader viewing several of the linked articles? The Main Page sections are insignificant compared to Wikipedia traffic as a whole, so we are not trying to use the Main Page to increase total traffic; all four of the sections (to different extents) are (among other goals) trying to get readers to read articles they wouldn't otherwise have read, and that objective is best measured by an estimate of individual readers.
• The various process to create the Main Page sections concentrate the vast majority of their resources on the bolded link, not on the others. Hence, to see whether a given process is an efficient use of project resources, it is fair to concentrate on the part of the output (the "featured" article) on which the resources have been spent.
• It is already very tedious to get a large enough data set while only looking at the bolded link; including the non-bolded links greatly increases the workload, and so reduces the availability of statistics.

I'd also say that I don't think the inclusion or exclusion of non-bolded links would significantly change the relative hit rates in my January analysis. I only used the bolded link for each of the sections (and only the most popular bolded link for multiarticle hooks). Physchim62 (talk) 11:47, 15 November 2010 (UTC)

• So the first set of hooks in which only the hook-article is linked will get many many more hits. Is this not the whole point? Tony (talk) 06:36, 15 November 2010 (UTC)
  • PS I can't work out from Psyschim's table what the underlying hit-rate of each hook-article might be, compared with the spike on the day caused by the hook. Are the data retrievable only on a daily basis? If so, beware the effect of time-zones. The previous or following day might be part of the spike. Tony (talk) 06:40, 15 November 2010 (UTC)

As far as DYK is concerned, you can see a smaller "day-after" effect on the spike, but only for the final set of hooks in the UTC day. This is presumably due to server caching of the Main Page by ISPs. I didn't take it into account in the January stats because it is so much smaller than the variability in hit rates between hooks; I'll try to include it if I do a new set of stats as long as I can come up with a way to make a similar correction on OTD. For TFA and ITN, there is a much larger "day-after-peak" effect, because the article is still linked from the Main Page, so I treated this separately. Physchim62 (talk) 12:10, 15 November 2010 (UTC)

I presume you're using Heinrick's thing, which now seems to be working consistently for the first time in god knows how long. Tony (talk) 12:14, 15 November 2010 (UTC) PS, there'll be a day-before effect, too. Tony (talk) 12:15, 15 November 2010 (UTC)

Yes, stats.grok.se There's sometimes a "day-before" effect on ITN, both because of the subject matter and because of the way ITN works, but not for the other sections. Why do you think there should be? Physchim62 (talk) 13:19, 15 November 2010 (UTC)

If the hooks are posted first thing in the day, UTC, there will be some spiking in North America on the previous "date". Tony (talk) 14:10, 15 November 2010 (UTC)

No, because the statistics are collected per UTC date, not local date. The only way you can get a "day-before" effect is if significant numbers of readers are looking at an article before it is posted on the Main Page. That happens (sometimes) with ITN, because news media are talking about the topic before the ITN story goes up, and (depending on the topic) readers are looking for information on Wikipedia on the basis of news reports. I don't see why it should happen with the other sections, and I haven't seen any evidence for it. Physchim62 (talk) 14:36, 15 November 2010 (UTC)

More reviewers needed

The great debate above seems to have distracted attention from the actual running of the project, we are currently down to only 18 verified hooks with a number of queues empty. Can we get some more people reviewing please? Gatoclass (talk) 06:54, 16 November 2010 (UTC)

I don't think as many people come here, so maybe we should ask on the suggestions page. For instance, when you review and approve someone's hook, take that time to ask them to review another if they haven't recently. I know there are a few users who nominate hooks every day but never review or comment on anyone else's. - PM800 (talk) 09:29, 16 November 2010 (UTC)

Thanks for the suggestion, but I'm not about to start leaving notes all over the Suggestions page for people to get involved. We are (supposedly) in the process of establishing a quid pro quo system of reviewing in any case, so nominators will be getting more involved eventually, but in the meantime we will have to continue relying mainly on our regular reviewers, who do actually read this page. Gatoclass (talk) 11:12, 16 November 2010 (UTC)

We lost our most productive contributor in maintaining reviews and preparing sets, etc, in Rlevse. At the same time, wholesale change for the project was suggested / demanded by various editors, some with reasonable comments and suggestions, others disparaging the project and its contributors. The suggestion for including recently (and comprehensively) sourced BLPs, possibly with a 2x or even 3x expansion requirement, has been left to linger and die despite broad support. My own nomination as a test case has been neglected to death which doesn't exactly improve my motivations (and I was already feeling disillusioned following the harrassment across this page). The stats analysis arguing for <1000 views per hook appears to me to be methodologically flawed, but no one else appears to have noticed. Speaking just for myself, I recognise the need for more reviewers / reviewing, but I am struggling to summon the will to contribute to the effort needed. Maybe I am not the only one feeling discouraged. EdChem (talk) 11:37, 16 November 2010 (UTC)

I think we are all feeling a bit exhausted and irritable after the last few weeks of discussion. But the show must go on. If you need a break, by all means have one, hopefully someone else will step up to the plate in the meantime. As for your "ignored" proposals, they may have some merit that can be revisited sometime down the track, but in line with your own feelings, I think a lot of people are tired of discussing the various proposals at the moment and could use a break from it. Gatoclass (talk) 14:27, 16 November 2010 (UTC)

I'll tackle some reviews when I get home tonight. @Ed, keep your spirits up. I for one appreciate your efforts here. 28bytes (talk) 22:18, 16 November 2010 (UTC)

Queue 3 correction

In the sixth hook of Queue 3, "feet" should be "foot". (I was fixing it in Prep, but got an edit conflict as it was moved to Queue.) MANdARAX • XAЯAbИAM 22:47, 16 November 2010 (UTC)

Fixed by changing the hook to use {{Convert}}. --Allen3 ^talk 22:53, 16 November 2010 (UTC)

Hub Hart in queue 3

The Hub Hart article in queue 3 has an apparent inaccuracy. The hook reads ...that James "Hub" Hart started his Major League Baseball career a few months after he was named an All-American for football? The problem is that he does not appear to have been an All-American in football. The various All-American selections have been collected at 1904 College Football All-America Team with citations to the sources. Hart was not on any of the lists. The sole source cited in the article for the supposed "All-American" selection is based on a 1982 article written by a Georgetown undergraduate student in the school's student newspaper, "The Hoya." Erroneous identifications of players as All-Americans is rampant. I suggest that this hook be returned to the Suggestions page until this is sorted out. Cbl62 (talk) 04:36, 17 November 2010 (UTC)

Further evidence that the hook is wrong. Multiple lists of All-American football players from Georgetown make no mention of Hart. See Hoya-Saxa list of Georgetown All-Americans, Hoya Football list of Georgetown All-Americans, and NCAA list of Georgetown All-Americans (pages 13 and 32). Cbl62 (talk) 04:47, 17 November 2010 (UTC)

Thanks. Returned (that queue was lacking a quirky hook anyway ..). Materialscientist (talk) 04:54, 17 November 2010 (UTC)

Changing DYK

It seems, from discussions all over the place, that there are two proposals to be (re-)made. So here they are. Discuss. Uncle G (talk) 12:12, 1 November 2010 (UTC)

Proposal 1: Slowing down the output rate

DYK is largely output driven. One of the ironies of the recent discussions was that partway through a robot interrupted to nag people that the output end was in danger of emptying. That seems to be causing a rush for reviewers. Can we make it more input-driven? If not, can we slow down the output rate?

Discussion (Prop. 1)

• I think slowing the output rate is a Good Thing in itself, and also a necessary part of any move to "improve" checking, accountability, etc. The Devil, as always, is in the details! Slowing down the output at the current rate of input involves refusing a lot more submissions than at present, preferably as early as possible in the procedure to avoid wasting reviewer resources (always in short supply all across WP). I've noticed two proposals for how to do this (apologies if I've missed any):

• Refuse hooks that are too mundane (only the hook, not the whole article, needs checking in order to refuse, but probably needs multiple reviewers on the hook to be fair)
• Increase the length requirement (can be done automagically, but length is a poor indicator of quality and no indicator at all for the sort of problems that have arisen recently)

Both ideas have their supporters and opponents: my strong preference is for the first. Physchim62 (talk) 13:54, 1 November 2010 (UTC)

If we slowed down the output rate, what would you have the new figure as?

Mundane hooks are somewhat subjective really. I've seen instances of hooks being labelled as boring by others, yet they'd drawn me in. Multiple reviewers for one hook would be nice, but it's not practical. There are simply too few reviewers at present. Paralympiakos (talk) 14:06, 1 November 2010 (UTC)

I agree that refusing boring hooks is a good idea in principle, but for reasons I have expressed many times here (see my comment dated 00:24, 29 October 2010 (UTC) above) I don't think it's a workable proposal under the current system. If we're going to do it, someone needs to come up with a much more specific proposal about how exactly DYK will select or reject hooks. Otherwise it's just going to be a drama fest. rʨanaɢ (talk) 14:19, 1 November 2010 (UTC)

My ideal output rate would be a single DYK section (7–10 hooks) every 24 hours. I guess that that's probably too big a change to make all at once, because we also need to address the expectations that editors have when they submit to DYK. But I think once it becomes clear that hooks are being rejected for being simply too mundane, submitter attitudes would also quickly change: editors would have to consider "is this really 'Main Page interesting'?" before submitting hooks. Physchim62 (talk) 14:29, 1 November 2010 (UTC)

Maximum of 10 in a 24 hour period. People's expectations will adjust very quickly. It will free up multiple reviewers to work on the best submissions, and lead to better, more interesting articles being featured. The only argument against this is "but we've always done it this way." Which is a poor argument indeed. Bali ultimate (talk) 14:32, 1 November 2010 (UTC)

Statements like "the only argument against this is 'but we've always done it this way'" shows how carefully you've read the preceding discussions. Actually participating in constructive discussion and weighing the pros and cons of all options would be more useful than repeatedly making strong demands about how you want DYK to be changed, like a broken record. rʨanaɢ (talk) 16:10, 1 November 2010 (UTC)

There's a big problem here. If one cares about quality then reducing throughput is the single simplest, lowest cost change that could be made to improve this place. You have an insufficient number of competent reviewers and the competent ones you have are strapped for time. That's a fact. The question then becomes, how do you fix it? All i see are hidebound status quo defenses. If the argument is a philosophical one -- that you want to encourage new article creation, irrespective of quality, then we'll just have to disagree (and lots of new articles are created completely outside of people interested in this process). I happen to agree with Edward Abbey that growth for its own sake is the ideology of the cancer cell. There's a further problem that so many of the hooks don't work with a "Did you know?" format (which promises something enlightening or surprising and counter-intuitive.) I saw one today that was of the nature of "did you know that a very obscure so and so was a nephew of an equally obscure so and so?" If you really want to feature such new articles, then just create a "new on wikipedia" and just include the most interesting possible one-sentence summary of the content. But i digress.Bali ultimate (talk) 16:32, 1 November 2010 (UTC)

As for how it would work, let's keep it simple! A reviewer scannig the nominations page comes across a hook that he or she feels is just too mundane for a section entitled "Did you know?", so they add something like:

• The hook seems just too mundane to me. ~~~~

at the top of the discussion. After, say, three reviewers have found the hook too mundane, the nomination automatically fails. Physchim62 (talk) 14:37, 1 November 2010 (UTC)

That seems like a sensible idea to trial. I was looking over at WP:ITN yesterday and noticed that this is basically what they do when selecting articles. If this leads to too much discussion on whether a hook is interesting, rather than people reviewing the articles, it may have the opposite effect to what is intended though. I agree with Physchim62 and that hope nominators will start to adapt accordingly, if we up the standard for hooks. This would be a major change from what I thought the function of DYK was (to encourage new content) rather than it being especially interesting, but I guess maybe the project has reached a stage where it is becoming more mature and that we should be trying to improve the quality of DYK. As for how many articles we should feature, I think that cutting it down to two sets a day is probably a good idea to start with, numbers above seem to be plucked out of the air, rather than based on any reasoning. I'm not sure about increasing the length requirement, because sometimes you can have a really interesting hook about something that little is known about. Generally speaking, of the articles I review, most are way over >1500 at the moment anyway, so I'm not sure this would change anything. If someone could run some stats on that though it might be helpful. SmartSE (talk) 15:20, 1 November 2010 (UTC)

(e/c, in fact my edit was deleted for some reason, probably by the wiki-software) The selection of articles going to the front page should be based on the quality of the article rather than the cleverness or mundaneness of the hook. Hooks can easily be rewritten, even by a reviewer. Badly written or poorly referenced articles on the other hand shouldn't be on the main page, even if they have an awesome hook. If there is anything this discussion should be reaching for, it should be to encourage reviewers to review the article, first and foremost.First Light (talk) 15:32, 1 November 2010 (UTC)

_{::::Sorry don't know how that happened. SmartSE (talk) 18:16, 1 November 2010 (UTC)}

If you see DYK as a showcase for articles (which I don't, I see it as a service to our readers) then it's failing, because the readership of articles on DYK is far below that which would be expected given its position on the Main Page. You can't have higher readership or better reviewing at the current throughput rate. If we want to give Smarties to good little editors, fair enough, but there's no need to squat a Main Page section to do that. Physchim62 (talk) 16:13, 1 November 2010 (UTC)

I also see DYK as a service to our readers, first and foremost. Any content on the main page should be putting the readers first. For that reason, it's incumbent that the articles featured there at least meet minimum standards of quality, referencing, and freedom from copyright violations. A clever little hook leading to a poorly written or referenced article is a disservice to our readers, in my opinion. It's also a disservice to new editors to reward them for a clever hook and a poor article. First Light (talk) 17:26, 1 November 2010 (UTC)

No one is suggesting that poor articles are put up. All that's being suggested is that some nominations are refused because reviewers don't think the hooks will interest the readers. The idea is to have fewer articles to check for quality. Physchim62 (talk) 18:40, 1 November 2010 (UTC)

There are two problems with removing "boring" hooks. One is that someone who is bored by science, for example, will probably find many of the science related hooks boring. Someone who thinks pop culture articles are mundane will find those hooks to be mundane and boring. I'm actually arguing for the sake of others, since I find interesting hooks somewhat easy to come up with. I think you'll find that the vast majority of boring hooks could be rewritten to be less boring. Now, if the article were mundane, boring, badly written, poorly referenced, plagiarized, or a copyvio, then the reader will be let down by the clever hook. To make a long story short, I think we're putting the cart before the horse by basing accepting or rejecting an article on the hook alone. A reviewer should be reviewing the article first, since this DYK is about highlighting new articles and not new hooks. First Light (talk) 20:39, 1 November 2010 (UTC)

(ec) I think that going to 3 updates of 9-10 might be viable, taking out the hooks / articles that really aren't worthy. "Boring" is unbelievably subjective, and if you want to use it to cut by 60+ % is asking for major fights. It is also changing the function of DYK from encouraging new content creation to - well, I'm not sure what it is changing the function to. I understand that highlighting new GAs has some appeal (for example), but DYK is supposed to be encouraging readers to start contributing, and for editors early in their careers as Wikipedians. New articles that are headed for GA and FA are great too, but early career Wikipedians deserve some encouragement, IMO. We have a nomination page of 200 to 250 articles at present - which are the 120 to 150 of those nominations that are unworthy? EdChem (talk) 15:42, 1 November 2010 (UTC)

I'm sorry, but I just don't see how this proposal is going to reduce the workload of reviewers, which is the real problem. Someone will still have to go through the noms to decide which are the most viable - that alone could turn out to be a nightmare. Then every ref. in every remaining article will have to be thoroughly vetted to see there is no plagiarism. And what is going to be the net benefit of all this? I would suggest, minimal. This will completely destroy the object of DYK as it currently exists, if we are going to do this, we may as well dump DYK altogether and just promote GAs. Gatoclass (talk) 16:51, 1 November 2010 (UTC)

I'm not sure I understand your criticism here. Surely it is quicker and simpler to answer the question "would this hook interest a significant number of our readers, and so lead them to read the full article" than to answer the question "does the entire article satisfy all the DYK criteria for posting". The smaller number of nominations that get through to the second stage of checking would more than make up for the effort on the first stage of "checking" (really just expressing an opinion). Either you cut down the number of article, going onto the full article checking or you find more reviewers, there's no other option for improving reviews. Physchim62 (talk) 18:55, 1 November 2010 (UTC)

There is a way that "boring" could be used as a criterion to reduce workload: if nominations die after a set time without review. It may be assumed that reviewers are at least partly choosing what to review based on what looks interesting to them. A bot can move, say, week-old nominations without any review comments to a subpage, where they get another week for someone to rescue them by moving back to the main page (not the original nominator). Then they just die for lack of interest (bot deletes nom from subpage). Besides workload, other advantages of this approach are flexibility and avoiding arbitrary standards: if we run short on noms, then noms which would otherwise be judged too boring will be get sufficient attention to get through anyway. Rd232 ^talk 17:41, 1 November 2010 (UTC)

I don't see how simply ignoring a nomination is any improvement on stating an opinion that it's not DYK material. At least if someone states an opinion, there can be discussion about it; simply ignoring things is no solution at all. Physchim62 (talk) 18:55, 1 November 2010 (UTC)

Sure its a solution. If no-one can be found to express an interest, it's uninteresting by definition. Rd232 ^talk 20:34, 1 November 2010 (UTC)

Not really. There's a big difference between "I don't think this hook will interest a significant number of our readers" and "I don't want to review this article". In any case, ignoring nominations has always happened, and DYK is still churning out 32–36 hooks a day: ergo something new has to be done if DYK output is to be limited. Physchim62 (talk) 20:56, 1 November 2010 (UTC)

The 'ignoring' strategy has been discussed before, I myself have suggested it in the past. One problem is that people could easily game it by making i-scratch-your-back-you-scratch-mine arrangements with other editors. rʨanaɢ (talk) 19:03, 1 November 2010 (UTC)

That's true of just about anything, including both the status quo and more complex attempts at voting. I'm mostly inclined to say it doesn't matter that much, at least in terms of the dimensions introduced by the ignoring approach. In terms of reviewing, there's an argument to require at least two reviewers to sign off, which would limit these issues a bit. Rd232 ^talk 20:34, 1 November 2010 (UTC)

Agree. It needs to be more than one editor to be fair to nominators, but it would still be multiple editors making much quicker judgement calls than reviewing for an entire article. Physchim62 (talk) 20:56, 1 November 2010 (UTC)

Neutral - but see my comments on proposal 3 below. - The Bushranger _{Return fire}^{Flank speed} 21:55, 1 November 2010 (UTC)

I support reducing the output, but strongly oppose the criteria be based on "Is it boring?". As a lot of other users have mentioned, boredom is ridiculously subjective and can be easily gamed. I like the simpler solution: Increase requirements for DYK size and the amount of citations needed. Upping the requirements would reduce the flood of DYKs quickly cobbled together and force users to work on the quality of their articles.--hkr Laozi speak 22:12, 1 November 2010 (UTC)

I don't think the idea of "Is it boring?" is ridiculously subjective: the idea that quality can be measured by character count is equally subjective, if not more so. And where are these DYKs that are "quickly cobbled together" coming from? A DYK nomination has to come within five days of article creation. Ah yes, of course, all the people who are preparing articles offline, posting them fully made and immediately nominating for DYK! Hardly the sort of "new users" that we're supposed to be encouraging, IMHO. In any case, increasing article length increases the workload for reviewers, which is not exactly what we're trying to do at the moment. Physchim62 (talk) 22:40, 1 November 2010 (UTC)

You're implying that editors spend all five days working on their articles, when there's nothing stopping a user for doing the bare minimum in an hour, before placing their article up for DYK. In fact, I would argue that having any kind of deadline, less than a month, actually encourages users to rush their articles. Very few people have five straight days of free time to work writing and improving an article, for most people Wikipedia's a hobby, not a job. Also, I think the smaller workload created by reducing the output would more than make up for the slight amount of extra work required to check up on a few more citations.--hkr Laozi speak 23:14, 1 November 2010 (UTC)

And when I said "up the requirements", I wasn't only referring to size, but the amount of citations and reliable sources. Forcing users to check on and use multiple reliable sources can help improve the quality. I'll admit it's debatable by how much, but what other alternatives do we have for improving quality? Judging how interesting the hook is has no reflection on the quality of the article, only on the quality of the hook.--hkr Laozi speak 23:23, 1 November 2010 (UTC)

What you seem to be implying is that it's already impossible (for the editor with the average amount of spare time for WP) to create a DYK within the five-day deadline unless they prepare it offline... If that's the case, then DYK as it stands is just acting as Smarties for already established editors – who know their way around wikicode – and is already useless for encouraging new contributors. Any increase in the character limit would only make that situation worse: we would be creating a Main Page section just for the regulars.

As for the reviewer workload, the time needed to check an article for points like grammar and copyvio is proportional to the length. So if you increase the character criterion to 2,500 characters (as an example), you are increasing reviewer workload by up to 40%. Doo you really think there would be 40% fewer nominations with a 2,500-character minimum than at present? When so many of the current nominations are from "regulars"? Physchim62 (talk) 23:28, 1 November 2010 (UTC)

I'm not saying that it's hard to create a DYK within a 5 day deadline, but it is hard to create a DYK of good quality. Quality takes time, and five days is not enough. I haven't advocated a character criterion, and I never said that size should be the only deciding factor for DYKs. The number reliable sources, I think, is a much better gauge, as it shows that the author has done the research. Having higher requirements for sources also reduces cases of close paraphrasing, which has been a problem on DYK.--hkr Laozi speak 23:39, 1 November 2010 (UTC)

• Very strong oppose to this whole idea. You cannot fairly apply a "too boring" rule: it would be far harder to apply fairly than our least clear speedy deletion criterion. What's more, no good reasons for slowing DYK production have been given: give me a good reason to reduce the number of articles that appear at DYK. Every argument that I see is essentially trying to reduce DYK itself, in the spirit of "DYK is harmful"; no such statement has any worth. Nyttend (talk) 01:43, 2 November 2010 (UTC)
• Oppose as unworkable (for reasons discussed above) and contrary to the purpose of DYK, at least as I see it. As I see it, the purpose/benefit of DYK is to bring attention to new articles -- to improve them and to invite addition of relevant links, backlinks, and categories that the article creator might not have been aware of. That purpose is best served by giving a large number of articles a brief period of exposure on the main page.
  Notwithstanding that objection, if there is going to be an effort to deleted nominated hooks, the best way to do that is to be brutal towards nominations that are perceived as having serious problems. I've had some bad experiences here in which I found serious problems with a nomination (such as massive plagiarism in the article or severe misinterpretation of a cited source), but ended up being severely criticized for trying to reject the hook instead of graciously spending 6 hours of my life fixing the problem. I believe that it would be beneficial for DYK to take a harder line on junk. --Orlady (talk) 03:48, 2 November 2010 (UTC)
• Oppose, agree with these comments by Nyttend (talk · contribs) and Orlady (talk · contribs). Cheers, -- Cirt (talk) 03:51, 2 November 2010 (UTC)
• Oppose creating a subjective "interesting" standard. As noted below, Physchim62 finds an article on an early 20th century college football team to be boring and unworthy, but it got 8,700 views while on DYK and has had about 20,000 views since March. While we may be a bit twisted, some of us find sports and sport history fascinating. Different strokes. So I oppose a subjective interesting standard. If we are going to cut back on output, as Orlady noted above, take a harder line on the "junk", as determined by objective quality standards (not by subjective views of "interest" level). Cbl62 (talk) 06:28, 2 November 2010 (UTC)

Did you know... that readers aren't particularly interested in churches (967 readers) or moths (1139), are bored stiff by statesmen (766) unless they're diplomats having a drink (2621), but perk up for computer games (2616) and go wild for nipple tumours (4554) and vandalised genitals (8282)? - Pointillist (talk) 14:31, 2 November 2010 (UTC)

• Stop voting, people. This is a discussion. The proposal asks two questions. "Oppose" and "Support" are meaningless answers to either one. Uncle G (talk) 10:50, 2 November 2010 (UTC)
• Support slowing down the time frame of hook promotion, perhaps require only newly-promoted-GAs, and expand length of time of hooks on Main Page. -- Cirt (talk) 15:17, 2 November 2010 (UTC)

Proposal 2: More transparent logs

It's difficult to track when and by whom a nomination was checked, discussed, and approved. "What links here" from the article is no help. There are no archive pages. There's no permalink in the notice on the article's talk page. And scanning the edit history of such an oft-edited page is inordinately tedious if the discussion was months or years ago, and relies upon people using edit summaries that mention the individual articles. Can we switch to a system where it's easy, after the fact, to locate the DYK discussion and approval?

Discussion (Prop. 2)

More transparent logs are definitely desirable. Repeating some of what I suggested elsewhere:

• On COI: the administrator who promotes a set of updates from the prep area to the queue should be checking the history to see if the editors who prepared the queue match the editors with the DYKnom and DYKmake credits. Adding a DYKrev note would help with detecting COI problems.
• Carrying process information into the template at the article talk page would be good, so that there is a record that is easily found that says something like

"Article nominated for DYK by XXX on XXXDATE, article creation / development credited to user(s) XXX (diff). Nomination reviewed and approved by XXX on XXDATE (diff). Selected hook processed for main page appearance by XXX, and moved into the queue by administrator XXX on XXXDATE (diff)."

If the hook was subsequently moved back to the nominations page, and then re-queued, the DYKrev and admin information could be updated to reflect the final preparation / queueing before the main page appearance.

• This information would not only be very helpful for accountability purposes, but it would also allow us to more easily see if any editor(s) are regularly acting ouutside accepted procedure - they could then be counselled / advised / admonished (as appropriate).
• It would also (on the positive side) allow us to give greater recognition to those who are working hard on the reviewing and administrative tasks that are essential and yet get really no credit or appreciation. We could see who is doing good work, and recognise it.
• Recording this information in a central archives would also be useful, and such an archive could be incorporated into an altered page structure. I couldn't agree more with the criticisms that finding information in the history of T:TDYK at present is (at best) an irritatingly difficult process.
  

EdChem (talk) 12:37, 1 November 2010 (UTC)

Further thought... maybe we need a bot to assist with admin tasks, one that builds the archive at each update. We could mark each nom when it is moved to prep, and the bot would then remove it from T:TDYK and start an archive entry, noting who moved it to prep, etc, adding the DYKrev based on who gave the DYKtick or the AGFtick, etc, etc. EdChem (talk) 12:47, 1 November 2010 (UTC)

I haven't read through this whole thing yet. But regarding the COI issue you raise, that is relatively rare already; so far in the whole ANI discussion only one instance has been raised, and that has already been corrected. There are already rules against it, I don't think we need to make a big deal of adding COI checks to the process; . Promoting one's own article is bad, people shouldn't do it, if someone is noticed doing it they get a warning, simple.

As for your list of records that should be kept on the {{dyktalk}} template, personally I think that is too much information and it would be a real pain for editors to have to copy and paste all that into each template for each article every time they promote hooks. As I said at ANI (and I'm not sure, to be honest, why that discussion is now being duplicated here), it would make more sense to set up a transparent archiving system such that, once you know the date of an article, you can find the rest of that information with a single click; that would circumvent all the copy-pasting.

Anyway, for the moment all discussion of this archiving system is moot, because I have been looking at {{dyktalk}} and I'm not even certain it will be technically possible to implement what either of us was (without starting the whole thing again from scratch). Ultimately that question will have to be answered by Shubinator, who best understands how DYKUpdateBot works...whatever DYKUpdateBot does to get the hook it puts on the talk page (like this), it will also need to do that to get whatever other information we add to {{dyktalk}}. rʨanaɢ (talk) 14:13, 1 November 2010 (UTC)

I'm not really sure what the question is. If a human can do it (and here I'm just talking about credits, not copyvio screening), it's likely the bot can too. The date would need to be embedded into {{DYKmake}} for the bot to read it though. The dyktalk issue is fairly separate from the bot itself; if dyktalk can't be modified to suit our needs, no human or bot will be able to do it. Shubinator (talk) 04:34, 2 November 2010 (UTC)

Skimming through what has been read it seems you want a WP:TFD like system which would locate all the hook history info for one day at one page, could transclude active nom information through to main suggestion page. Then you'd either close each hook TfD style or move (similar to how you remove them now) to a "Completed hook" section which by includeonly/noinclude tags wouldn't transclude to the active nominations page. Rambo's Revenge (talk) 14:48, 1 November 2010 (UTC)

Would it be possible for nominators to add a template to the talk page of the article, and for these to be transcluded to T:TDYK for discussion? That way there would be a permanent, easily found discussion of the hook, unlike now where they are hidden in the history of T:TDYK. Obviously this would involve some (probably lots) reprogramming of bots, so may not be possible to implement quickly. SmartSE (talk) 15:27, 1 November 2010 (UTC)

I'm not sure how transclusions of that many templates would affect the loading time for those on slower connections or users of older versions of Internet Explorer. Not speaking against the proposal, just pointing out an issue that should be dealt with if we're moving forward on this. GeeJo ^(t)⁄_(c) • 15:37, 1 November 2010 (UTC)

Smartse, what you are suggesting is essentially individual subpages for each nom, and that proposal has been shot down every time it was suggested over the past 2 years (at least), partly because of reasons like what GeeJo brings up. The proposal that I made was for subpages by date, like what Rambo's Revenge describes above (my full proposal is at Wikipedia:Administrators'_noticeboard/Incidents/Plagiarism_and_copyright_concerns_on_the_main_page#Setting_up_date_subpages_and_archiving). rʨanaɢ (talk) 16:15, 1 November 2010 (UTC)

Sorry I didn't realise you had made a similar propsal to me. I tend to avoid ANI: it is the historical (and current IMO) definition of WP:DRAMA. Rambo's Revenge (talk) 16:27, 1 November 2010 (UTC)

Ah ok, nevermind then, just thinking aloud. Your suggestion at ANI sounds like a good one to me, but I'm not sure whether you mean for us to have an archive where every thread can be viewed at once, or whether you'll still have to trawl the history to find the thread. It would be good if all the day's noms could be viewed on one page, but currently removing threads from T:TDYK is an important part of moving hooks to the main page and making sure they are only moved there once. Can anyone think of how we can make a decent archive and solve this problem? SmartSE (talk) 18:13, 1 November 2010 (UTC)

Shouldn't be a problem. Excluding special holding area there are 10 regular days at present, so that would translate as 10 templates transclude. I've seen FXC pages transcluding over 40 featured candidacies. Rambo's Revenge (talk) 16:13, 1 November 2010 (UTC)

• I'm wary of this idea, but not completely opposed. My main problem is that I can't see how it would address the concerns that prompted all this discussion. Better "accountability" may well be a good thing in its own right but, on its own, it won't lead to better reviewing: if we assume good faith then all the current reviewers are already doing the best they can, so just making it easier to blame people shouldn't change anything! Physchim62 (talk) 21:05, 1 November 2010 (UTC)
• Weak Support? - I would support a {{DYKrev}} tag being added to DYKmake and DYKnom, but not anything more. - The Bushranger _{Return fire}^{Flank speed} 21:57, 1 November 2010 (UTC)

Proposal 3: introduce some GA DYKs

It's been suggested before to introduce some Good Article DYKs, with a generally positive response from those not involved in DYK, and generally negative from those who are (see eg here). The objections there seem to be (i) DYK is about showcasing new articles and (ii) DYK as is couldn't handle the extra workload. The first point I disagree with fundamentally (it's certainly not intrinsic to the concept), and so do plenty of others - it shouldn't be just about promoting new articles; newly improved or newly verified as Good quality should be showcased as well. The second point is non-trivial but I have a suggestion which I think could work well: make the additional workload part of the GA nomination process, so that GA DYKs come here ready to be slotted into the DYK queue. This would require GA structures to figure out their own DYK process as part of GA review, using appropriate criteria that suit everyone. The hooks would be listed under an appropriate separate heading ("from our newest Good Articles"), with as few or as many GA hooks as available (up to a max of 50% of the available DYK slots). GA hooks could be displayed longer than new article hooks, if we end up with a situation of there being many queues with no GA hooks at all.

Why should we do this? a) for the readers: better quality highlighted on Main Page. Quite apart from GA review probably being stronger than DYK review (so copyvio and reliability problems etc less likely), GA articles are generally stronger than DYK articles. It may be argued that Today's Featured Article already showcases Featured Content, but it does so in a very different way from DYK, and has only 1 slot while DYK has 10, of which I'm proposing up to 5 be used for GAs (depending on availability of new GAs). In any case GAs are not FAs, and many never will be, or won't be for a very long time (and presumably previous featuring as DYK would be taken into account in choosing Today's Featured Article, so might never be featured unless we run out of new FAs that haven't been). b) for the editors: the element of intrinsic reward is greater for new articles than GAs, and so the latter should get at least as much recognition. DYK new articles tend to be written by just one person, so they already have some satisfaction from getting their work published. GAs are far more likely to have a collaborative element, improving other people's work; and bringing in DYK recognition would encourage that, so that more articles currently at C or B class would be brought up to GA standard by people not previously involved with them. In any case, Wikipedia is an encyclopedia, and the older and more established it gets, the more over-emphasising new content at the expense of good content seems to be sending the wrong sort of message (devaluing content maintenance and improvement, which becomes ever more important the older Wikipedia gets). Rd232 ^talk 18:00, 1 November 2010 (UTC)

Discussion (Prop. 3)

• We don't have enough people to thoroughly review existing DYKs, so the solution is to add GAs to the mix? Sorry, but this makes no sense at all. Gatoclass (talk) 18:11, 1 November 2010 (UTC)

GA articles are (by definition) already reviewed - and to a higher standard than new articles. This isn't a problem.--Scott Mac 18:12, 1 November 2010 (UTC)

It may not be "a problem", but it does nothing to resolve the current problem at DYK, which is lack of manpower to review all the submissions. And adding more articles for promotion is inevitably going to create more work, not less. Gatoclass (talk) 18:19, 1 November 2010 (UTC)

Did you read the proposal properly? Hooks would come ready to slot into the queue, and the structure of the proposal is that GA hooks would expand or contract as available up to 50% of the queue. The additional daily decision-making at DYK will generally be pretty minor. Rd232 ^talk 18:32, 1 November 2010 (UTC)

• I strongly support this. Incentive for people to improve content as well as create new content or expand stubs. There is not extra work for DYK, just change the criteria to accept hooks from any article awarded GA status in the last month. The hooks then just need checked like any other hooks. Allow 50% of the hooks to be from GA (of course if GA hooks are not available their quota can be made up from extra new article hooks). DYK will be the same for the readers, but will now encourage article improvement (which will increase the number of people participating and, probably, therefore increase scrutiny).--Scott Mac 18:11, 1 November 2010 (UTC)
  • Changing the criteria would be an easy way to introduce GAs, if we didn't have the manpower issue. Sorting out hooks as part of the GA process is something which could work well I think, and not create extra work at DYK. Linking the two processes will also probably be good for both, in terms of enhanced visibility and cross-fertilisation of editors (if you see what I mean...) Rd232 ^talk 18:37, 1 November 2010 (UTC)
  • @Scott: So getting the GA project is not incentive already, but editors can only ever be motivated to improve content by offering them baubles? I still don't see where people started getting the idea that a couple hours on the main page is the only way to make people edit the encyclopedia... rʨanaɢ (talk) 22:03, 1 November 2010 (UTC)
    • A non-argument so absurd as to verge on disruptive. Clearly DYK "baubles" (such a wonderfully dismissive way of putting it) motivate some people; so the question is how best to use the incentives they provide. It is not a requirement for them to be effective that these incentives motivate everyone never mind everything done on Wikipedia. Rd232 ^talk 22:22, 1 November 2010 (UTC)
      • Pointing out that there are other ways to motivate people other than main page time is "disruptive"? Have it your way. But getting this worked up isn't going to be constructive, either. rʨanaɢ (talk) 22:28, 1 November 2010 (UTC)
        • Re-interpreting an obviously unconstructive and churlish comment as a merely stupefyingly self-evident point is not helping matters. Rd232 ^talk 22:40, 1 November 2010 (UTC)

• I'd support this, but it will only work (unless the main page is totally redesigned) once we've resolved the first proposal. I'd also be a bit concerned that it might pressurise GA reviewers to pass articles without enough scrutiny, in the same way that some DYK reviewers might do now. SmartSE (talk) 18:24, 1 November 2010 (UTC)
  • Proposal 1 would help, since clearly this proposal involves taking some of the slots; but it's not quite essential to the point of saying "it will only work if...". I don't see how it would pressurise GA reviewers, because GA has its own pace and logic, and because the GA hooks are explicitly designed not to have a daily quota which GA might fail to meet. If there are no GA hooks on any given day, that's OK, the process takes as long as it takes. If there are often no GA hooks, they can be displayed longer than new article hooks, so there's at least 1 or 2 a day on average. Rd232 ^talk 18:32, 1 November 2010 (UTC)

• I'd support this, but I'd go further and scrap the idea of new/expanded articles, and only have GAs, which are generally of higher quality than stubs. We should not be displaying stubs on the main page. Aiken (talk) 18:39, 1 November 2010 (UTC)
  • DYK doesn't display stubs. And you just said that DYKs are "generally of higher quality than stubs". So I'm not sure what you're trying to say. rʨanaɢ (talk) 22:06, 1 November 2010 (UTC)
    • Some are basically "glorified" stubs, and meet the bare minimum of 1500 characters (which isn't much). Aiken (talk) 11:03, 2 November 2010 (UTC)
• Support in principle, but needs to be aligned with proposal 1. Another thing to consider, if a new article quickly becomes GA, does it "compete" for a "new article slot" or a "GA slot" or both (thereby doubling its chances of mainpage appearance)? Sasata (talk) 18:53, 1 November 2010 (UTC)
  • I doubt it would happen that often. I think logically it could be either but would make more sense to showcase as GA. Rd232 ^talk 20:20, 1 November 2010 (UTC)

• Yes, please. The main page content would dramatically improve in quality, and those who spend a great deal of time improving an article to GA level would get equal recognition with those who create the much easier and typically lower quality DYK articles (I know, having done both). I think this idea would need to be discussed in an RfC, rather than solely at Talk:DYK, to get much broader community input. First Light (talk) 20:50, 1 November 2010 (UTC)
• Strong support - DYK shouldn't just be "New and greatly improved", it should be truly "Did You Know that...?". This is a way to do it. And there's an easy way to use Proposal 1 and still keep DYKs at their current output rate - slash "New and Improved" DYKs by 1/2, and replace the 1/2 removed with "New GA" DYKs. - The Bushranger _{Return fire}^{Flank speed} 21:58, 1 November 2010 (UTC)
  • I don't follow you. By "it should truly be 'Did You Know that...'", are you trying to say that DYK hooks should be interesting? Then how does this proposal help—are you aware that GAs can be just as dull as DYKs? rʨanaɢ (talk) 22:09, 1 November 2010 (UTC)
    • Interesting is subjective. What I mean is, "Did you know this fact". I'm sure as we all know from school 'learning something' doesn't always mean the same thing as 'Wow!'. ;) - The Bushranger _{Return fire}^{Flank speed} 22:19, 1 November 2010 (UTC)
      • So you are saying DYK currently isn't "did you know this fact", but if we add GAs then it will be? What makes you say that? rʨanaɢ (talk) 22:26, 1 November 2010 (UTC)
        • I'll answer that because it's a point I haven't made in the proposal, and it's important. The philosophy inherent in "Did you know?" is simply presenting interesting facts to readers; the restriction to new articles or massively expanded articles is artificial and based on encouraging editors (both current and to a lesser extent future, from the "look you can make one too" effect). Expanding the domain to all Wikipedia content would be philosophically correct from the reader's point of view but practically unhelpful in incentive terms for editors. Expanding the domain to include newly promoted GAs is therefore a small step towards the practice implied by the underlying philosophy, whilst retaining and indeed enhancing its useful in incentive terms. Rd232 ^talk 22:54, 1 November 2010 (UTC)
  • As it stands, there are not enough GAs being promoted to fill 50% of the DYK slots at current DYK output level (32–36 hooks/day). 50% GAs can only work if the total DYK output is two section (16–20 hooks) per day or less. Physchim62 (talk) 22:11, 1 November 2010 (UTC)
    • Hmm, fair point - The Bushranger _{Return fire}^{Flank speed} 22:19, 1 November 2010 (UTC)
    • As the proposal stands, the point has been made repeatedly that this does not matter, as 50% is a maximum not a quota. Please pay attention. Rd232 ^talk 22:17, 1 November 2010 (UTC)
      • Sorry, I only meant to reply to Bushranger's enthusiasm, not the proposal as a whole. It get's a bit confusing to know who's replying to whom here! Physchim62 (talk) 22:55, 1 November 2010 (UTC)
        • OK :) Rd232 ^talk 01:39, 2 November 2010 (UTC)
          • There's no reason that the 50% Good Articles couldn't remain the same for 24 hours, while the regular DYKs could do their usual rotation of every 6-8 hours. First Light (talk) 04:11, 2 November 2010 (UTC)

• Rd232: I find it a little strange that in making this proposal you have simply copied and pasted all your statements from Wikipedia:Administrators'_noticeboard/Incidents/Plagiarism_and_copyright_concerns_on_the_main_page#break without including any of the rest of that discussion there. In particular, you're continuing to oversimplify the "oppose" arguments and ignore the ones I that specifically raised in that discussion. Is there any reason you're repeating all your points word-for-word here, other than going to the other parent? rʨanaɢ (talk) 21:59, 1 November 2010 (UTC)
  • Rjanag, I find your failure to elaborate your concerns here particularly strange in that at the ANI subpage you had said you weren't elaborating there because it wasn't the correct place (which is true of course). Now, do you have an actual point? If so, are you going to tell us what it is? Rd232 ^talk 22:17, 1 November 2010 (UTC)
    • I elaborated my concerns rather explicitly at the ANI thread before you restarted the discussion here. My main point is the one I have already said multiple times: GA is a meaningless designation for readers and DYK is not; as far as readers are concerned, putting GAs on the front page just duplicates what TFA is already doing, whereas DYK serves a different purpose that is clear (and its purpose never was to showcase "quality", so the argument that DYK isn't as high quality as GA seems moot). All the proposals I have seen so far to put GAs on the main page are for project-internal reasons and don't necessarily stop to consider whether it would make sense for readers.
    • That's all I'm saying about this for now, as I'm not really interested in wasting space by repeating myself at length when you have already heard my views. rʨanaɢ (talk) 22:34, 1 November 2010 (UTC)
      • I took your points on board by explicitly comparing the proposed GA DYK hooks with TFA. It is clearly not duplicative, as argued in the proposal. It is clearly not just for project internal reasons, it is showcasing good content in an interesting way for readers. And GA status is no more meaningless to readers than Featured Article status (and by the by, it's a more intuitive name I think; "Featured" is less obvious, I'd rename it Excellent if it were up to me). You claim I've ignored your points, but in fact you've repeatedly ignored mine. Rd232 ^talk 22:47, 1 November 2010 (UTC)
• Support - this makes all kinds of sense. → ROUX ₪ 22:30, 1 November 2010 (UTC)
• Oppose. One of the best things about DYK is that it's open to everyone, not just Wikipedia insiders/regulars. There are about 250 editors who have each individually contributed 25 or more DYK hooks. It's a way for new editors to get their feet wet with new articles and receive recognition for that. I suspect that the GA process, on the other hand, is for the most part the work of the insiders/regulars. How many editors have contributed 25 or more Good Articles? Probably a handful in comparison to the 250 who have contributed 25 or more DYKs. If I had a suggestion, it would be to try to encourage more DYKs from new and different editors. Even though I've been a heavy DYK contributor, I'd be fine with a rule that imposes a limit of no more than 20 (or pick the appropriate number) DYK hooks by a single editor per month. That would make room for new editors, reduce the submission of repetitive or cookie-cutter articles by insiders/regulars, cut the total output, and promote more variety. Cbl62 (talk) 06:40, 2 November 2010 (UTC)
  • Hm. You argue that DYK is open for everyone, not just WP regulars and then go on to argue that the fact that 250 editors have contributed 25 or more DYKs (by which time they're hardly new editors) proves something that supports your argument. In any case GAs are obviously not comparable to DYKs in terms of effort. Probably GAs do have fewer newcomers involved than DYK, but I'm not sure the difference is all that great (what proportion of DYK contributors are relatively new?). In any case, the proposal is to introduce some DYKs drawn from GAs, up to a maximum of 50%. We could have a maximum of 10% or 20%, which hardly affects the current DYK quantities drawn from new articles. Rd232 ^talk 08:46, 2 November 2010 (UTC)
    • Hm. I maintain my point. In addition to the 250 who have contributed 25 or more DYKs, there are thousands who have contributed smaller numbers. In this way, DYK is a forum for the new editor and editors of all type to have their content featured on the main page. This broad participation is incredibly valuable to encouraging development of editors. By shifting to GA, we move in the opposite direction of giving over DYK to the insiders/regulars (an even smaller number) who are involved in the GA process. Cbl62 (talk) 15:22, 2 November 2010 (UTC)
• I think the main problem with DYK is in the conception and setup. Its open nature, and the loose criteria right now give nominators almost universal right to secure a listing on teh front page if the basic criteria are met. I see this 'every one's a winner' approach as one of the main root causes of our ills today. I think in the interests of quality control, endowing DYK nominations with one or more 'quick fail' criteria is surely the way to go. However, most of thse criteria will need to be qualitative/objective. Thus I think something along the lines of GAN would be the way to go. The shortage of reviewers is another way of "natural selection" – by definition, if the nomination is not interesting enough to secure reviewers within say 14 days, it will die. --Ohconfucius ^¡digame! 12:25, 2 November 2010 (UTC)
• Oppose, the purpose of DYK is to showcase new material. -- Cirt (talk) 12:29, 2 November 2010 (UTC)
  • • No, it isn't. The purpose of DYK is i) to provide Interesting Facts to readers ii) to incentivise editors to improve Wikipedia. The declaration that this incentive can only be used to generate entirely new articles is arbitrary, as neatly illustrated by the current exception which allows old articles expanded 5x to qualify. Adding content newly verified as Good Quality makes complete sense. We can certainly discuss how they're included (whether separate heading, or what proportion, whether as fixed proportion or maximum as I suggested), but the philosophical claim that the purpose of DYK is to showcase new material is simply wrong. (If that were true, we'd have a random selection drawn from all new articles, with the opening sentence quoted: you can't argue against that if you don't accept my points i and ii.) Rd232 ^talk 13:47, 2 November 2010 (UTC)
  • What's wrong with expanding the scope? Particularly when the current scope encourages (if only through inaction) plagiarism and copyright violation? → ROUX ₪ 12:32, 2 November 2010 (UTC)
    • Because this sort of proposal would change DYK to the point that it is unrecognizable and does not serve its own independent purpose any longer, per comments by Gatoclass (talk · contribs), who had said it could almost be completely scrapped if we were going to do that, unfortunately. -- Cirt (talk) 13:02, 2 November 2010 (UTC)
      • Ah. All change is bad, gotcha. No point in engaging you further on this until you approach it in good faith. Cheers. → ROUX ₪ 13:12, 2 November 2010 (UTC)
        • The same could be said in reply. -- Cirt (talk) 13:13, 2 November 2010 (UTC)
• In addition, the move towards needing GA level articles for contribution to DYK is inevitable. The rate of new article creation has slowed and will continue to slow as has been discussed at great length elsewhere on Wikipedia. The simple fact is, the longer Wikipedia is in existence, the fewer new subjects there are to write about; the focus of the project has been moving slowly from creation in its early phase to maintenance and improvement. The survival of DYK (which as a trivia nerd I personally find a really fascinating way to learn new things) depends on a continual stream of new content. As that new content lessens, other avenues will need to be explored in order to keep DYK relevant and useful. → ROUX ₪ 13:02, 2 November 2010 (UTC)
  • The amount of DYK nominations shows this statement to be simply inaccurate. -- Cirt (talk) 13:05, 2 November 2010 (UTC)
  • It's not inevitable whatsoever. We still have to wait over a week for noms to hit the queue. There's a steady backlog and that will only increase when the wikicup returns in a few months. There's no shortage whatsoever, to my mind. Paralympiakos (talk) 13:09, 2 November 2010 (UTC)
    • I will need to dig them up, but there are hard numbers showing that new article creation has been on a steady decline for some time. → ROUX ₪ 13:12, 2 November 2010 (UTC)
      • Here at DYK, we have increased the number of hooks on the page, from six, to a range of up to 8, to an expected 8, to now 9. There is no shortage of new creation. -- Cirt (talk) 13:21, 2 November 2010 (UTC)
        • See here. Distinct downward trend in article number for ten years. Image showing Wikipedia growth by month, obvious decline. → ROUX ₪ 13:32, 2 November 2010 (UTC)
          • Not good data for this discussion. How does that compare to number of DYK hook nominations over time? -- Cirt (talk) 13:34, 2 November 2010 (UTC)
            • Don't know and don't care. The simple fact is, DYK requires new articles or 5x increase in old ones. I would be willing to bet the majority are net new articles in any case. Fewer new articles == fewer possible DYK hooks. This is a pretty simple proposition. → ROUX ₪ 13:36, 2 November 2010 (UTC)
              • Wrong. As I have said, the amount of DYK hook nominations have increased over time. -- Cirt (talk) 13:38, 2 November 2010 (UTC)
                • Really? Cite please. → ROUX ₪ 13:42, 2 November 2010 (UTC)
                  • Apparently you accepted this at face value, as your prior response called this "irrelevant", lol. -- Cirt (talk) 13:43, 2 November 2010 (UTC)
                    • Also, see Wikipedia_talk:Did_you_know#8_hooks_or_9.3F. -- Cirt (talk) 13:45, 2 November 2010 (UTC)
                    • No... irrelevant means that your statement doesn't matter, not that I accepted it. Got a cite or not? That link doesn't say noms are increasing, it says lowering the number of slots increased the backlog. → ROUX ₪ 13:48, 2 November 2010 (UTC)
                      • Your admission that the backlog has increased, contradicts your prior statements about a prediction of less hook nominations in the future. -- Cirt (talk) 13:58, 2 November 2010 (UTC)
                        • Um... no. If you have 100mL of liquid and pour it into a 110mL container, you will have no overflow. If you pour it into a 90mL container, you will have 10mL overflow. So, again, have you got any actual proof that nominations are on an increasing trend? → ROUX ₪ 14:01, 2 November 2010 (UTC)
                          • Um... no. There is a backlog. The backlog has increased over time. Compare with [7]. -- Cirt (talk) 14:10, 2 November 2010 (UTC)
                            • I don't think you understand. And since you don't appear to have any sort of actual trend showing increase in noms, whereas I have clearly shown decrease in net new articles as well as concern about declining noms on this very page, it seems more clear than ever that you're simply not arguing in good faith here. When you feel like providing proof to support your assertions, I'll be happy to look at it and change my opinion if it actually shows what you assert. Until then, you're just wasting my time. (Cue snarky comment from you, "The same could be said of you") → ROUX ₪ 14:22, 2 November 2010 (UTC)

(od)the number of DYK slots was increased because of the backlog, was it not? Roux's point about new articles doesn't quite fit here because the selection of new articles for DYK is such a small proportion of all new articles (the arbitrariness of the tiny selection one of the things I dislike about it). Where the decline in new articles over time fits in is to point to the increasing importance over time of maintenance and improvement of existing content. The latter we have an obvious way to incentivise, by using some DYK hooks from recently promoted GAs. Rd232 ^talk 14:19, 2 November 2010 (UTC)

It does fit. Fewer new articles = smaller pool of articles to create DYK hooks from. Simple math. → ROUX ₪ 14:22, 2 November 2010 (UTC)

too simple math: DYKs are not a fixed proportion of new articles. And DYKs are such a small selection of all new articles that I can't see a reduction in new articles having much effect. It's far more up to the relatively few people who choose to be involved in the DYK process, adding nominations and especially reviewing them. Rd232 ^talk 14:28, 2 November 2010 (UTC)

To roux, I think there are still oodles and oodles of articles out there, but I do worry that the ones we make are more and more esoteric over time, hence why I support recently promoted GAs. Casliber (talk · contribs) 23:17, 14 November 2010 (UTC)

I'm not opposed to featuring hooks from recently promoted GAs. Set a certain time to feature new GAs, give them there own nomination section. Grsz¹¹ 14:31, 2 November 2010 (UTC)

• Support changing DYK to showcasing newly-promoted-GAs, and simultaneously increasing time on Main Page of sets of selected DYK hooks. -- Cirt (talk) 15:20, 2 November 2010 (UTC)
• Strongly Oppose changing DYK to showcase GAs unless they also (somehow) fit the existing 5-days/5x-expansion criteria. DYK is for Wikipedia's newest articles, period. If there's a need to create a special section for "Today's Not-Quite-Good-Enough-To-Be-Featured Article" on the mainpage, argue for that, but don't try to hijack DYK because you wish your "good" article got the tiny bit of attention granted to a newly created or expanded article. - Dravecky (talk) 20:14, 2 November 2010 (UTC)
  • This territorialism is growing tiresome. Perhaps you should familiarize yourself with what people have been saying before you start a) slagging them off, and b) going on about 'hijacking'. → ROUX ₪ 20:19, 2 November 2010 (UTC)

• Support including GAs as a way to refocus on quality rather than quantity or newness. The fivefold-expansion rule has always seemed rather arbitrary to me. An article could be dramatically improved with no expansion whatsoever; Some rambling pieces of textual rubbish would probably end up shorter after a good rewrite. --Hegvald (talk) 14:23, 3 November 2010 (UTC)

• Oppose. Not that I don't think that GA is an important part of Wikipedia, but I generally disagree that GA's deserve special mainpage recognition. The existance of DYK and Featured Articles on the main page provides a nice bookend effect: at DYK you can find the newest articles, and at FA you can find the best articles, in their closests-to-finished state you will get (yes, I know that even FAs aren't finished, but you get my drift). Adding GAs to the main page muddies this nice symmetry. I don't really think we need to place articles on the main page at intermediate stages. The current system works fine, and doesn't need any tweaking in this regard. --Jayron32 04:38, 4 November 2010 (UTC)

  I don't think it is a matter of GA's deserving recognition; instead they are a vast resource which can be used on the mainpage to encourage editors to contribute to Wikipedia. Geometry guy 01:06, 6 November 2010 (UTC)

• Support Wikipedia is at a point where we should encourage quality over quantity. The amount of GAs will be insignificant, but it's a project that deserves more exposure. Lampman (talk) 12:49, 5 November 2010 (UTC)

• The main argument against this proposal seems to be that DYK is for new articles only. Apart from being circular reasoning, this is patently false: anyone who has followed DYK recently will know that articles are increasingly newly expanded articles, not recently created ones. So by banning recently promoted Gas from entry, we’re basically saying that recently expanded articles are ok on the main page, while recently improved articles are not. This preference for expansion over improvement seems to me a complete perversion of the most basic principles of Wikipedia. Lampman (talk) 13:56, 5 November 2010 (UTC)
• Support. DYKs have simply too many issues for this to persist, and I think GAs are more useful to readers as a generally higher-quality resource. I'm not swayed by the various arguments above: DYK is hardly any more "newb-friendly" than GAN, seeing as a few people churn out the lion's share of them. Der Wohltemperierte Fuchs^(talk) 14:58, 5 November 2010 (UTC)
• Encourage. I have for some time suggested that the main page should make use of the resource that is the rapidly expanding pool of Good Articles. This is an idea whose time may have come. Supplementing DYK's by articles which have already been reviewed, and pre-identified at GA for quality and interest could lighten the workload here while also providing greater choice of articles and hooks to present each day, so that we don't reach (or have to scrape) the bottom of the barrel. I agree that such a change may have to be introduced in conjunction with other reforms to the selection and scheduling of DYKs, but the synergy created with GA reviewers might boost contributions to both DYK and GA processes. Indeed the interaction could be quite exciting! Geometry guy 01:06, 6 November 2010 (UTC)
• Support, I too have advocated for this in the past. The current criteria for DYK is sketchy (e.g. 5x expansion for existing articles). GAs and FAs would actually complement DYK quite well considering that an extensive review process already happens for articles to achieve GA/FA. -- Lil_℧niquℇ №1 | talk2me 20:33, 12 November 2010 (UTC)
• Support. Featuring some GAs for 24 hours is better than showing a bunch of DYKs that stay on the main page only for 6 hours because a lot of us will miss 1 update while we sleep. OhanaUnited^{Talk page} 06:59, 14 November 2010 (UTC)
• Strong support. Adding a GA that has been promoted in (say) the past 5 days fits under the definition of "newly improved" for mine (which is what expanded articles are, i.e. not "our newest articles" anyway. The other reason I like this idea is often a GA improver is looking for more sets of eyes before propelling an article towards FAC, hence this is a nice nudge in the right direction. Casliber (talk · contribs) 23:09, 14 November 2010 (UTC)

Another suggestion: Use newly promoted Good Articles as DYK source

Just before I log off for the day, I might as well throw this out. As I said above, if we are going to start promoting GAs, we might as well just scrap DYK altogether and merge it into GA. We could scrap the "new article" concept, which means we could get rid of perennial niggles like the 5-day nomination window and the x5 expansion. DYK would then become "from Wikipedia's most recently promoted articles". We could have a minimum article length of 5,000 characters which would get rid of all the little stubby articles. GA has a more robust (though far from perfect) review process, articles which pass GA could be reviewed a second time for plagiarism issues and so on, then go into a promotion pool. Updates would be made whenever necessary, probably only one or two a day. We would then have the manpower of both the GA and DYK people, we wouldn't have the periodic lobbying from GAers for entry to DYK, and it would put an end to much of the ongoing controversy regarding the robustness of DYK's review process.

I am not saying I would support this myself, but I certainly think it would be more viable than the halfhearted merge that is being suggested here and on many occasions previously. Gatoclass (talk) 19:03, 1 November 2010 (UTC)

A radical paradigm shift, but maybe the time has come. Sasata (talk) 19:10, 1 November 2010 (UTC)

If DYK were to be scrapped, then I could think of much better uses for the free space than "featuring" GAs, so no. I'm not opposed to a mixed DYK section of new articles and new GAs, so long as it lives up to it's title of "Did you know?" Physchim62 (talk) 19:20, 1 November 2010 (UTC)

"much better uses" - like what? Rd232 ^talk 20:28, 1 November 2010 (UTC)

Well my pet idea is to shift "selected anniversaries" over and use the space for a "Recent deaths" section, as we often get 50k+ readers on the article of someone who has recently died without it being on the Main Page, and there is currently no quality control for such articles unless they make it on to ITN (rare). Other suggestions that have been made include "Recent sports" and "selected portals". But none of this is really relevant to the current discussion. Either we try to improve DYK or we scrap it altogether; but this is not the place to decide what to replace it with if we scrap it altogether. Physchim62 (talk) 20:39, 1 November 2010 (UTC)

Another option would be to have a section where people to nominate popular articles that are in good condition (or recently improved to good condition): there are many possibilities, and I find it slightly presumptuous that it is always "featuring GAs" that comes up. Physchim62 (talk) 20:43, 1 November 2010 (UTC)

I take your points, but I find "nominate popular articles that are in good condition (or recently improved to good condition)" confusing in your contrasting that with GA status. GA status is not synonymous with "good articles" but it's a good way of screening for these purposes. Anyway if you're not incentivising improvements, you're losing some of the power of the Main Page. Rd232 ^talk 23:02, 1 November 2010 (UTC)

An excellent suggestion. New articles are not what we should publicise on the main page, and GAs deserve more recognition. Aiken (talk) 19:28, 1 November 2010 (UTC)

I'm not convinced. I think if it works very well with GAs and that process beds down then later on we could choose to make it all GAs, without the risk of major transition hiccups. But at the same time, there are good reasons that DYK is long established with new articles, and I'm not sure we should get rid of that completely, it does have value as an illustration of Wikipedia's constant expansion, planting an idea which can attract new editors. Selection and review processes can be sharpened of course. Rd232 ^talk 20:28, 1 November 2010 (UTC)

Would it be possible to have two separate sections, one for GAs and one for traditional DYKs? However, this won't solve the current DYK problems...--hkr Laozi speak 21:54, 1 November 2010 (UTC)

That's Proposal 3 above. Rd232 ^talk 22:25, 1 November 2010 (UTC)

No, I mean, as separate sections. Not just mixing in GA DYKs with New Page DYKs, but having a completely different section linking to Good Articles on the front page.--hkr Laozi speak 22:58, 1 November 2010 (UTC)

• Strong oppose - but see my comments about Proposal 3 above. - The Bushranger _{Return fire}^{Flank speed} 21:59, 1 November 2010 (UTC)
  • Please stop voting. Everyone was having a real discussion until you started voting. Uncle G (talk) 22:30, 1 November 2010 (UTC)

There is no universally correct answer for this debate. It's reasonable to define WP's usefulness as breadth*depth; both are important. That said, it makes sense to revisit the question, for example yearly: which is needed more, breadth or depth. If the consensus is depth, we showcase x (around 9) hooks relating to GAs at a time for z hours apiece (where z is chosen to control queue size). If the consensus is breadth, we do what we did for the last year. For the moment, focusing on depth would likely sidestep most plagiarism/copyright/quality issues.

I invite folks to find the actual diffs, but I expect that the genesis of DYK was not surrounded by deep philosophical searchings. Therefore, wondering about what the original purpose of DYK was is less helpful than wondering about the best use of the space and/or click-throughs. I'd say that it is obvious that at some point encouraging depth will be more practical that encouraging breadth. It seems equally obvious that at some point (perhaps when we're getting close to that 5-millionth article) we may want to encourage breadth. In my mind, the usefulness of exploring the breadth depth option combined with flexibility down the road ftw. Cheers. Haus^Talk 03:07, 2 November 2010 (UTC)

• Oppose, DYK should remain a separate and independent process, and not be "merged", in this fashion. -- Cirt (talk) 12:33, 2 November 2010 (UTC)
  • Explaining why would go a long way to helping this discussion achieve consensus. → ROUX ₪ 12:58, 2 November 2010 (UTC)
    • For one thing the whole entire purpose of DYK is to showcase new material. -- Cirt (talk) 13:04, 2 November 2010 (UTC)

A lot of GAs are already new articles. Merging the two processes would not in any way preclude the nomination of new articles, it would just mean the emphasis would shift from new articles to good articles. Gatoclass (talk) 14:25, 2 November 2010 (UTC)

• Support changing process to limit to showcasing newly-promoted-GAs, and then also increasing timeframe of selected hook sets on Main Page. -- Cirt (talk) 15:21, 2 November 2010 (UTC)
• Strong Support - replacing DYK with a selection of recently-promoted Good Articles would be a far better use of the space on the main page. What should be important is quality, not mere newness (and rapid expansion) and DYK encourages the latter. Robofish (talk) 17:28, 2 November 2010 (UTC)
• Oppose for same reasons I cite above. I like that the main page features the new articles; it highlights the fact that we are constantly growing. If the DYK process needs tweaking, that's fine, but scrapping it and replacing it with GA is a bad idea. --Jayron32 04:40, 4 November 2010 (UTC)
• Comment I don't think we promote enough articles to GA to fill an entire queue. If we threw B-class articles into that category I would strongly support. Marcus Qwertyus 21:14, 8 November 2010 (UTC)

It took me a while to locate the monthly Wikipedia:Good article statistics, which go back years. Cheers. Haus^Talk 18:20, 9 November 2010 (UTC)

Cool. So it could work but we'd have to have reduce it to about 2 queues a day. Marcus Qwertyus 20:22, 9 November 2010 (UTC)

• support. It seems like a good compromise of all the comments made previously. -- Lil_℧niquℇ №1 | talk2me 03:07, 14 November 2010 (UTC)

• Wow - definitely has merit, but I have a niggling worry that the gulf between stub creation and GA is too big. I do like it alot though. Casliber (talk · contribs) 23:20, 14 November 2010 (UTC)
• Support per my comments above and elsewhere on this page. Haus^Talk 02:09, 15 November 2010 (UTC)

Fifth Proposal: No more time deadlines

• Based on the above discussion, I have an idea for another proposal: I think we should scrap the five day deadline entirely, it encourages people to rush out articles, and ignores recent, good articles that have already passed the five day deadline. Quality takes time, after all. Doing so would make the process more like FA and GA nominations, but for shorter articles and with lower quality criteria. Regardless of when they were created, articles with enough support votes end up in DYK, and ones that fail, don't. It could be argued that this would increase the workload, but if we reduce the daily output to 10 DYKs a day, the workload should be about the same as the AfD process.--hkr Laozi speak 23:50, 1 November 2010 (UTC)
  • The time limit is not much of an issue really; in reality, very few articles are actually passed and shown within 5 days of being created or nominated. Plus, even with longer time, I doubt reviewers would spend more time checking articles more carefully; what happens (in my experience) is you do what is needed to check that an article meets the established criteria, and then you move on. Having twice as much time to review an article doesn't necessarily mean you would spent twice as long doing it. rʨanaɢ (talk) 00:01, 2 November 2010 (UTC)
    • I see your point, but I think having no time limits would allow much older articles, but of good quality (not neccessarily a Good Article), to be nominated. However, this couldn't be implemented without also upping the criteria for DYK nominations, as you've proposed, and I support. I guess what I'm suggesting is a DYK for a third tier of articles, short articles with a quality in the B range to A range (which would include GAs, but not make it exclusive to them).--hkr Laozi speak 00:31, 2 November 2010 (UTC)
• Oppose, the purpose of DYK is to showcase new material. -- Cirt (talk) 12:29, 2 November 2010 (UTC)
• Support, insofar as expanding the time horizon of DYK hook sets, while requiring all DYK hooks to come from newly-promoted-GAs. -- Cirt (talk) 15:24, 2 November 2010 (UTC)
• Oppose, although an extension from 5 days to 7 might make sense. - The Bushranger _{Return fire}^{Flank speed} 16:02, 19 November 2010 (UTC)

Identity crisis

If I've interpreted this situation correctly, it seems the problem is that copyvios and plagiarism are slipping past reviewers. The most-endorsed solution seems to be to reduce DYK output. If you cut from 36 hooks/day to 10 hooks/day, that means at the current rate of submission 72% of hooks must be rejected. It is debatable whether doing this fairly and objectively is even possible, and even if it were implemented, arguments over which hooks deserve to be featured would quite likely mean more work for the reviewers, not less. So, given limited personnel and a serious problem, I think we have to decide what DYK is supposed to be.

• 1) Is it to feature good (not necessarily GA specifically, but some sort of standard) articles? In that case, I think the best solution would be to ditch DYK altogether and work with the GA project completely, or you would still face the problem of rejecting massive numbers of nominations.
• 2) Is it to feature new articles? In that case, the number of DYK nominations have to come down to give reviewers more time to check for copyvio/plagiarism. One possible way to do this would be to make the nomination process much harder. Another possible way would be to disallow self-noms, and only allow reviewers who have gone over the article and can vouch for it to nominate.
• 3) Is it to encourage new editors? In that case, one option would be to make it so that only articles from editors with 25 or fewer DYK credits can enter. Thus, writing articles for DYK wouldn't be any more difficult, but far fewer nominations would be possible.

Just some thoughts. Make of them as you will. -- Yzx (talk) 00:51, 2 November 2010 (UTC)

I agree with your assessment. It's never going to be possible to reach a consensus on what should be done with DYK until a consensus is reached about what DYK (or at least its spot on the main page) should be about. But for several years now there has been a wide variety of views about that. rʨanaɢ (talk) 01:43, 2 November 2010 (UTC)

I think that's true for all of the Main Page sections! Physchim62 (talk) 02:20, 2 November 2010 (UTC)

If the concern is purely plagiarism/copyvios, the simplest solution, simpler than reducing output, is to require more citations and distinct reliable sources as a criteria. Close paraphrasing seems to come from articles that rely primarily on one or two distinct sources as references. Forcing editors to use multiple, different sources per section could prevent that from happening.--hkr Laozi speak 04:30, 2 November 2010 (UTC)

I doubt that is feasible; most DYK topics are pretty obscure & often all the online sources available are used. Typically online sources repeat the same basic sort of material on many sites, & forcing editors to include several such sources will only make things harder to control all round. I don't really see that "Close paraphrasing seems to come from articles that rely primarily on one or two distinct sources as references" although of course they often may. The simple fact is that DYK reviewers have not been nagged to check for plagiarism & people following the suggestions page have mostly only seen completely blatent examples (mainly just from ignorance of the policies) rejected for it. If reviewers just add it to their checks, & people see more rejections, the culture will respond pretty quickly without the need for the more drastic solutions outlined above. This is essentially the approach that FAC is taking btw. Has anyone noticed how GA seems to have stayed apart from this hurricane, unlike FA & DYK. This might be because GAs don't share the same problem, but personally I very much doubt that. Bringing GAs on might just be going from the frying pan to the fire. They are after all much longer to check. Johnbod (talk) 04:41, 2 November 2010 (UTC)

What is the current count of GAs approved per day? I like GAs and think it would be neat to have them linked to the main page but I like the DYK aspect as well and more. DYK hooks are often interesting, obscure (as put above) and I like reading through them; I learn new things. I am an honest believer in slowing the whole process down. Slowing the process down along with increasing the standards in more than one way will change the reputation (as put above as well).--NortyNort (Holla) 12:45, 2 November 2010 (UTC)

Looking at the recent history of GAN, it looks like 8-10 approvals per day. Gatoclass (talk) 14:21, 2 November 2010 (UTC)

If the efforts at DYK were moved to GAN, this would be increased. And they could stay on for a full day. Aiken (talk) 14:23, 2 November 2010 (UTC)

Which would be why I made my proposal below. No need to reject anything for 'not enough slots', just throw them into a queue from which a bot will choose randomly each DYK cycle. In terms of not enough GA-class promoted per day, would be quite simple to say 'anything promoted in the month before implementation' to ensure a cushion. → ROUX ₪ 14:25, 2 November 2010 (UTC)

Well I haven't read your proposal closely but I would be opposed to any kind of "bot selection". A bot cannot possibly make judgements about what is and is not appropriate to promote. I just can't see any role for a bot at all. Gatoclass (talk) 14:29, 2 November 2010 (UTC)

...you should probably read it more closely, then. The bot would simply be responsible for making random selections from already-approved hooks. → ROUX ₪ 14:39, 2 November 2010 (UTC)

DYK lottery

One more idea: I said above I couldn't think of a way to reject massive numbers of hooks fairly and objectively, but I can think of one. Make subject area categories sorted by day and have hook nominations under them. Have a bot (or random number generator, or reviewer with their eyes closed) pick one from each at random. That hook will be reviewed, and if rejected, another will be randomly picked. Once a day's set of hooks are complete, the other eligible hooks will be discarded. It won't be pretty, but it'll get the job done. -- Yzx (talk) 01:28, 2 November 2010 (UTC)

Any review is subjective, so even your process would not be completely objective. It would be an interesting system for picking TFAs... Physchim62 (talk) 01:35, 2 November 2010 (UTC)

It would not be reviewed for interesting-ness, only for adherence to set criteria (refs, no copyvio/plagiarism, etc). This will mean some boring hooks, yes. But rejecting three-quarters of all hooks based on "interesting-ness" would be much worse. Even ignoring the subjectivity of that judgment, that would turn DYK into a literal competition with winners and losers. If people think sensationalism and "reward culture" is bad now... -- Yzx (talk) 01:41, 2 November 2010 (UTC)

If we could objectively review articles against a set of criteria, we would do it by bot or script, just as DYK already does for the character minimum or 5×-expansion criterion. But you can't, you need subjective human judgement on questions of grammar, sourcing, copyvio, BLP, linking etc. Under my proposal, there would only be "winners and losers" among the editors who feel that it is somehow a right to get an article on the Main Page just because it's new: my personal opinion is that no editor has such a "right" to get an article on the Main Page. Why "reward" the editor who creates a new article on a single early-20th century season of a college football club, destined to be read by half-a-dozen people a month at most, and not the editor who expends an equivalent amount of work cleaning up an article that's read by 2,000 users every day? Or the editors who create a featured topic, adding coherence to Wikipedia's coverage? All this to create a section which our readers ignore, presumably because they don't find it as interesting as… Selected anniversaries! We have Freedom of Religion, one can worship at the altar of supposed objectivity, but it sure creates some funny results. Physchim62 (talk) 02:14, 2 November 2010 (UTC)

Why punish those that write about an early 20th century college team? DYK is fine as it is, we don't need to stop these articles from being DYK just because you think they'll only get half a dozen views a week. Paralympiakos (talk) 02:23, 2 November 2010 (UTC)

? Physchim62 identified a way in which the current system fails to reward important work. Focussing (at least to some extent) limited rewards on important work does not constitute "punishment" of the less important but still valuable work. Rd232 ^talk 02:28, 2 November 2010 (UTC)

Indeed, and I don't have any great ideas as to how to "reward" article cleanup or featured topics. Why should we choose to reward article cleanup over new articles, indeed? The point I was trying to make is that there is an inherent subjectivity in the way we allocate Main Page space as it is, so it is already pointless aiming for a completely objective system later on down the line. It's long been understood that Main Page space is a limited and valuable resource, which is why it is (I think) generally accepted that the readers, and not the editors, should be the priority in its allocation. Physchim62 (talk) 03:33, 2 November 2010 (UTC)

This would be a horrible idea. Absolutely against it. I'm sorry. Paralympiakos (talk) 01:55, 2 November 2010 (UTC)

Physchim62's comment about the early 20th century college football team demonstrates the "different strokes" issue in trying to select "interesting" hooks. To Physchim, the hook is utterly boring, but it got 8,700 views while on DYK and has had about 20,000 views since March (far higher than the half dozen a day posited by Physchim). (It happens to be one of the most important teams in the history of the sport.) That's quite good in comparison to the typical DYK hook on churches of England or rare species of mushrooms. While Physchim finds sports history deadly dull, others (obviously, myself included) find it quite interesting. Cbl62 (talk) 06:02, 2 November 2010 (UTC)

• Lottery: good idea. Your suggested implementation: poor. Why not simply have a queue of DYK hooks that get randomly selected for display? All we would need to do is choose whether X number of hooks will stay on the Main Page for each day, or simply load a random set of three whenever the page is loaded. Every time a new hook is approved, it gets added to the queue and stays there until either: it's appeared on the Main Page once (in the case of the first choice); or it gets bumped off when it's the oldest DYK left and a new one is added to the pool. If this were coupled with bringing in GA-level articles, more editors would reap the reward of knowing that their work was shown on one of the most highly visible sites on the entire internet. (Naturally this presupposes more stringent reviewing of hooks.) → ROUX ₪ 02:36, 2 November 2010 (UTC)

I had in mind fairness and hook balance. If all the DYKs are in a single queue, randomly drawn sets will proportionately favor more popular subject areas (particularly if you get a bad draw, which is always possible), whereas my system guarantees one of each subject per day. Furthermore, an editor can game the system by entering large numbers of hooks for repetitive articles. Under my system, an editor has a hard cap of one DYK credit per subject area per day, and that only with extraordinary luck. -- Yzx (talk) 02:48, 2 November 2010 (UTC)

Fairness is exactly the point. Every DYK hook that gets checked for accuracy and lack of copyvio gets tossed in a pool. Particularly if they're set to simply display a random 5 on every pageload, every checked hook makes it to the Main Page. Keep the pool at a reasonable maximum number (say, 200?); every new DYK that gets approved gets tossed in the pool, and another one removed on the FIFO principle. In a single move we have eradicated the time pressure (no rush to check or approve anything; a couple hundred DYK hooks are rotating on the main page), and increased fairness (everyone with a decent hook will have theirs appear on the Main Page). → ROUX ₪ 03:02, 2 November 2010 (UTC)

That approach is appealing (and rather more workable than Yzx's), but I wonder how much effort it would be to implement. (Incidentally different hook selection on every page load would be right out for caching reasons; selection for display would be about as often as now.) The approach does reduce time pressure on reviewing, which particularly if DYK processes are changed so review takes longer would be helpful. Rd232 ^talk 08:38, 2 November 2010 (UTC)

It shouldn't be too difficult. A bot could handle the adding/removal of hooks fairly simply, I think; it would just need to scan the hook page for approved hooks and drop them into the queue. Then each time the DYK cycle repeats (8 hours, 12 hours, whatever), the bot just grabs X number of hooks from the queue and drops them into whatever subpage transcludes to the Main Page. If we assume for the moment that we are going to start including GA-class articles, we suddenly have a much larger wealth of possible hooks to draw on, especially if to begin with we include as candidates any GA-level articles promoted, say, a month before the implementation date so as to build up a backlog. It would also be trivial to ensure that, much like TFA, GA DYKs with specific date associations get displayed on that date using something like:

{{exists|page=DYK/DYK-GA/{{CURRENTMONTHNAME}} {{CURRENTDAY}}, {{CURRENTYEAR}}
|then={{DYK/DYK-GA/{{CURRENTMONTHNAME}} {{CURRENTDAY}}, {{CURRENTYEAR}}}}
|else={{DYK/DYK-GA/DYKlist}}…}}

(Note: I used GA for this because new articles are less likely, I think, to be ready in time for a specific date. There's no reason they couldn't be, however, using similar syntax). This process would completely eliminate time pressure by ensuring there is a backlog (with hooks recycled as necessary; if we assume a pool of minimum 100 available hooks at any time, the bot can simply recycle back into the pool from the main page transclusion any time the pool drops below 100. Or 75, or OVER 9000, whatever), and in conjunction with other proposals above, removing the rush from first creation of a new article to DYK hook. This allows for more thoughtful checking of references and plagiarism, and will also allow more editors to see their work end up on the main page, albeit unpredictably in terms of actual timing. The benefit to readers should be obvious: readers will get to see a greater breadth of articles represented on the main page. → ROUX ₪ 12:56, 2 November 2010 (UTC)

• Oppose, this seems like throwing the baby out with the bathwater. -- Cirt (talk) 12:35, 2 November 2010 (UTC)
  • Perhaps you would care to actually explain your positions, instead of just blanket opposing every suggestion? → ROUX ₪ 12:56, 2 November 2010 (UTC)
    • Perhaps you could stop focusing on me and replying to every single one of my posts with pithy responses. -- Cirt (talk) 12:58, 2 November 2010 (UTC)
  • Hear hear: please explain, per WP:NOTVOTE. In particular, are you opposing a lottery per se, or Yzx's approach to it, which involves some DYK noms not getting on the Main Page at all? What about Roux's/my approach noted above, which doesn't? Rd232 ^talk 13:20, 2 November 2010 (UTC)
    • Arbitrarily throwing out hook nominations is a bad idea. How about applying that to GAN? How about marking as "failed" all GAN reviews on candidates that have not yet been reviewed if the nom is over one month old? How would that go over at GAN? -- Cirt (talk) 13:23, 2 November 2010 (UTC)
      • What is with people making obviously unhelpful comments in this discussion?? GA review is not comparable to the review of DYK hooks in the way that you imply. Worse, the suggestion I asked you to respond to was Roux's/my lottery idea, which does not involve throwing out any DYK hooks, it instead involves a lottery to display DYKs from a pool, so there's less time pressure on reviewing due to the constant need to fill a daily quota. Rd232 ^talk 13:29, 2 November 2010 (UTC)
      • I have subheaded my proposal, below, for clarity. → ROUX ₪ 15:11, 2 November 2010 (UTC)
• Support, this sort of idea of automation as proposed seems reasonable. -- Cirt (talk) 15:50, 2 November 2010 (UTC)
• Oppose I fail to see how this will solve any problems. DYK reviewers are currently having a hard time as it is getting through all the entries. This suggestion would mean that they'd have a to review a bunch of entries that would not even get featured. That would be a great disincentive for reviewers, who are already in short supply. If we want to reduce the number of DYKs, perhaps to accommodate a certain number of GAs, the best way to do that would be to tighten the requirements. That would have the added benefit of securing better quality DYKs. Lampman (talk) 14:16, 5 November 2010 (UTC)
• Strong Oppose,. per Lampman's comments. - The Bushranger _{Return fire}^{Flank speed} 16:02, 19 November 2010 (UTC)

DYK lottery system, new process proposal

Here is what I have proposed, subheaded and edited for clarity.

1. Bring DYK hooks in from recently (say, past two weeks, and past month for the first two weeks that this process is in place) promoted GA-class articles, hooks to be reviewed for plagiarism and copyvio as part of the GAN process; the only approval needed at DYK will be "Is this a good hook?"
2. Scrap the daily quota of DYK hooks
3. Tweak the approval process to, assuming no problems with copyvio and/or plagiarism, solely evaluate whether a hook is accurate and interesting on its own merits, not 'Is it interesting enough for one of the only available slots?'
4. Move all approved hooks to a queue from which a bot will randomly select as many hooks as there are slots on the mainpage for each DYK cycle (24-27 per day, FIFO principle)
5. The same bot will transfer those hooks to wherever DYKs are transcluded to the main page
6. Repeat 4 and 5 ad infinitum; hooks removed from the front page can then be moved to an archive, perhaps organized by day or month. Should the queue ever drop below an arbitrary minimum (say, 200), the bot will recycle the oldest (or newest, or shortest, or least popular based on pageviews, whatever) DYK(s) from the archive to make up the pool
7. Specific DYK hooks can still be scheduled to run on specific days, either at the request of the nominator or by suggestion from reviewers. Possible code to handle this below

code
{{exists |page=DYK/DYK-queue/{{CURRENTMONTHNAME}} {{CURRENTDAY}}, {{CURRENTYEAR}} |then={{DYK/DYK-queue/{{CURRENTMONTHNAME}} {{CURRENTDAY}}, {{CURRENTYEAR}}}} |else={{DYK/DYK-queue/DYKlist}} }}

This has numerous benefits:

1. Remove the time pressure from DYK, allowing for more careful and thoughtful evaluation of sourcing and plagiarism/copyvio issues in DYK hooks (with the assumption that such sourcing issues in GA-derived hooks would be dealt with before the hook comes to DYK for approval)
2. Give more editors more of a chance to have their work exhibited on the main page
3. Provide a broader exposure of articles on the main page
4. Provide incentive to editors to improve extant articles, as well as create new ones (yes, I am aware of the 5x path to DYK; improving to GA doesn't always include that much straight-up addition of content)
5. Every hook that DYK regulars approve will make it to the main page, removing the competition (and resultant bad feelings) for limited spots

Possible objections:

1. DYK is for showcasing new content - not the case; DYK is for presenting interesting content. The 5x avenue to DYK proves that it is not solely about new articles.
2. This will increase the reviewing backlog - not a problem, as backlog would simply not be a concern, due to lack of pressure in getting DYKs ready for main page
3. This will take more editor time - it won't, as a bot (yet to be coded) will handle all the mundane details of moving things around. In fact, this may well free up editor time, due to automating chunks of the process
4. GA-derived hooks will be rejected by DYK regulars - not a concern; the DYK regulars who oppose including GA content will abide by consensus and come to realize that this proposal is aimed at strengthening DYK and not eliminating it
5. More DYK hooks means waiting longer to get them on the main page - so? They will still get there, providing exposure for the article, and the nominator(s) will still be able to point at a specific date and think "Awesome, my work got seen by a gazillion people on that day!"
6. This is an attempt to take over DYK by GA - Simply not the case. DYK is still the funnel; GA would simply be another avenue towards putting hooks in.

Thoughts? → ROUX ₪ 15:11, 2 November 2010 (UTC)

• Oppose, such a large and drastic change to the entire process is not necessary. There are other ways to slow down the time process involved, including increasing number of hours per sets of DYK hook selections, and increasing quality requirements to only showcasing newly-promoted-GAs. -- Cirt (talk) 15:15, 2 November 2010 (UTC)

I haven't altogether got my head around this proposal yet. I can see some merit in the FIFO approach you are proposing, however, I see no need for a bot to do the selection and I think it would be much better for a human to make the selection as occurs with ITN for example. What I don't get is how this is supposed to resolve the problem of too many articles and not enough reviewers. We can't allow the pool of hooks to just go on expanding forever. I'm also a little concerned that a FIFO system will encourage laziness and less reviewing overall. Gatoclass (talk) 15:31, 2 November 2010 (UTC)

The FIFO idea was related to an older version of this proposal which had some flaws, and is no longer relevant; sorry for any muddiness on that point. The bot selection is just to pick randomly from hooks that are already approved, freeing up editor (reviewer!) time from mundane housekeeping tasks. It also ensures truly random selection; if a person is making the selections one can see arguments arising based on which got picked and by whom. The pool of approved hooks won't go on expanding forever; once a given hook has appeared on the main page it gets whisked off to an archive, and only recycled if the pool of available approved hooks drops below an arbitrary benchmark. Maybe I haven't explained properly, perhaps this will help:

Hook nom → review → placed in queue (by bot scraping review page) → randomly appears on main page (FIFO) → archived → recycled into queue if needed.

This would be the process for both standard- and GA-derived hooks, though the latter review would be more streamlined as copyvio/etc will have been handled at the GAN process. The recycling from archive option will ensure that reviews can be more careful, as there won't be any pressure along the lines of "we only have three slots filled for the next cycle." All cycles will always be filled. Have I made things more clear? → ROUX ₪ 15:46, 2 November 2010 (UTC)

You haven't explained why the pool of approved hooks "won't go on expanding forever", but more importantly, you haven't explained how to deal with the expanding pool of unapproved hooks, and how this system ensures that the better quality articles get approved and the others dropped. Also, hook selection is a relatively minor issue; one doesn't need a bot for that. The problem as I've said is too many submissions and not enough reviewers. Gatoclass (talk) 16:19, 2 November 2010 (UTC)

For the same reason they don't go on expanding forever now; they get used up. As for unapproved hooks, presumably partly by freeing up editor time from all the housekeeping, and am open to other suggestions. → ROUX ₪ 16:34, 2 November 2010 (UTC)

• Support, in particular the idea of FIFO sounds most logical. -- Cirt (talk) 15:51, 2 November 2010 (UTC)
  • Well I was aiming for random, but I suppose FIFO would work just as well, given that there seems to be some support for it. Have amended the proposal and my comment above to reflect this. → ROUX ₪ 15:55, 2 November 2010 (UTC)
• How would you prevent a glut of (for the sake of argument) U.S.-centric hooks from ending up on the Main Page at the same time? GeeJo ^(t)⁄_(c) • 20:00, 2 November 2010 (UTC)
  • I don't really see a problem with that (and before you say it, I am very much not from the USA and have general issues with americentrism); DYKs just go on the main page in the order they get approved, FIFO. Or go back to my original version, and do the actual appearance totally randomly. → ROUX ₪ 20:03, 2 November 2010 (UTC)
• Oppose - I can see the benifits, and I can see the rebuttals to the possible objections, but overall, to me, the benifits don't outweigh the negatives by enough to make this worthwhile. YMMV! - The Bushranger _{Return fire}^{Flank speed} 16:04, 19 November 2010 (UTC)

Proposal 5

There is a current request at DYK for editors to review as well as nominate. This could be made into a requirement, in the following way: once you have X DYKs to your name (maybe 5?), you need to have reviewed a hook within the last 5 days in order to nominate. As part of the nom, name the hook you reviewed. This should be combined with something along the lines of Proposal 2, both to make it easier to link to the reviewed hook, and so that hooks have reviewers more explicitly named (to prevent cavalier reviewing). Rd232 ^talk 08:55, 2 November 2010 (UTC)

• Oppose. This is not done at WP:GAN. -- Cirt (talk) 09:13, 2 November 2010 (UTC)
  • And? → ROUX ₪ 12:34, 2 November 2010 (UTC)

unnecessary
The following discussion has been closed. Please do not modify it.
Citing WP:OTHERSTUFF while saying "And?" is a bit amusing, particularly when this involves the suggestions of merging the DYK process with GAN, it is actually directly relevant. -- Cirt (talk) 13:01, 2 November 2010 (UTC) My point is, basically, so what if this isn't currently done at GAN? Your blanket opposes all over this page smack of the typical Wikipedia kneejerk opposition to any kind of change or rationalizing any process, and it's disappointing. → ROUX ₪ 13:07, 2 November 2010 (UTC) Your blanket responses to every - single - one - of my comments, smacks of opposition to my opposition and a need to reply to every single one of my posts, in a back-and-forth and back-and-forth and back-and-forth drawn out threaded process, over and above actual polite, professional, and constructive dialog. -- Cirt (talk) 13:10, 2 November 2010 (UTC) They are nothing of the sort. You are making blanket opposition, apparently because change is bad. You're simply not acting in good faith here, and I see no reason to continue speaking with you until you are. → ROUX ₪ 13:14, 2 November 2010 (UTC) And you seem to have some intense pattern of needing to respond to my comments about this discussion. -- Cirt (talk) 13:16, 2 November 2010 (UTC)

• • GAN does not, to my knowledge, have the same problem of lack of reviewers, and structurally it's much less likely to have issues with some people making lots of nominations and not reviewing much or at all (since a plausible GA nom is so much more work than a plausible DYK nom). If it does, it might help there too. But this is irrelevant at WT:DYK. In other words, Roux's "and?" is succinct and to the point. Rd232 ^talk 13:10, 2 November 2010 (UTC)
    • From prior experience, GAN actually does have a significant issue with lack of reviewers. -- Cirt (talk) 13:11, 2 November 2010 (UTC)
      • Fine, then we can propose a similar logic there, or work harder to bring in WikiProjects (since depth of knowledge matters more at GA than DYK). It's also a similar problem to lack of comments on RFCs; and I've suggested in the past some kind of lottery system (from a population of editors willing to be selected) to try and bring people in, and not just wait for people to wander by. But one problem at a time, eh? Rd232 ^talk 13:16, 2 November 2010 (UTC)
        • When I nominate articles I worked on to WP:GAN, I make an effort to review some others, to contribute to the ever-present backlog at GAN of waiting candidates. But try making that a requirement at GAN, and see what happens. -- Cirt (talk) 13:18, 2 November 2010 (UTC)
          • What, people will oppose because it's not done at DYK or FA? :P Seriously, if it's really necessary, each peer review process ought to be able to sort something out so it doesn't grind to a halt. The proposal I'm making here, at DYK, is more appropriate for DYK than elsewhere, because in general reviewing DYKs is easier than reviewing GAs or FAs. Now can we please stick to talking about DYK here? If you're suddenly keen on improving GA, go propose something there, and drop a link here. Thanks. Rd232 ^talk 13:26, 2 November 2010 (UTC)
            • But you are proposing to merge the two processes of DYK and GAN. -- Cirt (talk) 13:29, 2 November 2010 (UTC)
              • No I'm not! I'm proposing some of the DYK slots be given to DYK hooks constructed by the GA process, to be slotted into the DYK queue as the hooks come available (up to 50% max of queue). That hardly constitutes a merger, does it? Rd232 ^talk 13:32, 2 November 2010 (UTC)
                • Firstly, devoting DYK hooks to GA just makes the nomination backlog problem worse, unless you also want to gut the DYK nominations. Secondly, this idea can easily be seen as step one in GA takeover of DYK by stealth. EdChem (talk) 13:50, 2 November 2010 (UTC)

Take over DYK by stealth..? Good Lord, what kind of bizarre behaviour is that? We are trying to propose ideas to make DYK better, less open to abuse due to copyvio and plagiarism, and more open to more editors (why is creating new content automatically better than vastly improving extant content?). → ROUX ₪ 13:55, 2 November 2010 (UTC)

EdChem, on your substantive point about the backlog: a number of proposals have already been made to improve this which don't involve rejecting any nominations that meet the current standards - most obviously and simply my Proposal 5. More could probably be invented. And the proposal is for GA hooks to take up to 50% of slots, dependent on GA hook availability, a maximum which could be negotiated down. Rd232 ^talk 14:13, 2 November 2010 (UTC)

• As a general approach, I think a reviewing requirement (or some other project contribution) from more experienced nominators sounds reasonable. It would also help out with some of the problems / load created by the WikiCup. EdChem (talk) 13:20, 2 November 2010 (UTC)
• Support, this would be made much easier if we restrict DYK nominations to newly-promoted-GAs, and expand time horizon of selected hook sets on the Main Page. -- Cirt (talk) 15:26, 2 November 2010 (UTC)

• Seems unnecessary. As far as I know, most people do this anyway, it's just simple common sense and politeness to help offset the extra burden you are creating by nominating an article, and there are selfish reasons to do it too—the more other noms are reviewed, the sooner yours will get some attention. Speaking for myself, pretty much the only time I ever review anymore (not just for DYK, but for GAN or FAC as well) is when I've nominated something myself. Besides, I'm pretty sure there's already something on WP:DYK or WP:DYKAR along the lines of "if you nominate an article, consider also reviewing a couple" (although I can't find it at the moment). (strickn, I'm pretty sure I was thinking of PR) rʨanaɢ (talk) 17:05, 2 November 2010 (UTC)
  • So... there's an undisputed backlog, which would not exist if everyone pitched in their fair share of reviewing, but raising a mild request to the level of a requirement is unnecessary? Huh. Rd232 ^talk 17:08, 2 November 2010 (UTC)
    • Your proposal didn't ask for "everyone to pitch in their fair share of reviewing", it asked for certain editors (experienced ones—i.e., those with 5 or more DYKs) to review at least one article within five days of nominating something. If you do the math, you'll see that that still leaves people who might nominate without reviewing, and it leaves people who might review one article and nominate ten. (Much of the backlog comes from precisely these two things: people nominating their first DYK and not aware they can also review, and from the minority of prolific contributors who nominate a ton of articles but rarely review; your proposal doesn't really solve either of those, and if we want to "force" the prolific nominators to review as much as they nominate that could be done just as easily with a friendly message rather than a "rule", given that it's a small number of regular editors.) Not to mention that, as I said, a lot of people are doing what you propose already, so selfish nomination must not be the sole reason for the backlog. And, frankly, most of the people here are mature adults and I think it's a little demeaning to establish a behavioral requirement that essentially says "help out". rʨanaɢ (talk) 17:18, 2 November 2010 (UTC)
    • (note that I have stricken a portion of my message 2 comments above). rʨanaɢ (talk) 17:32, 2 November 2010 (UTC)
      • Mm, in the flurry of discussion I forgot to clarify something that occurred to me earlier: the proposal was supposed to include an additional criterion, that for each new nomination, the experienced (X successful noms) nominator needs to have a review within the last 5 days; I didn't mean "do 1 review, make 10 noms". In this form, it does help those who are "selfish", and making this requirement suitably prominent will help make everyone more aware of the need for reviewing (which I'm pretty sure is mentioned there somewhere). I discount friendly messages to established contributors because I assume such contributors are already well aware of the backlog, and need additional incentives to change their behaviour. Rd232 ^talk 17:44, 2 November 2010 (UTC)

• In addition to the points I raised above, I'm not sure this proposed rule would ensure good reviews (i.e., reviews that catch copyvio, which is the whole issue that started this discussion). A person being forced to do a review so that he can get his own nomination reviewed is, I expect, likely to just do the bare minimum. Call it not assuming good faith, but that's what I imagine would happen. rʨanaɢ (talk) 21:38, 2 November 2010 (UTC)
• Oppose. Not everybody has the time or energy to rummage through DYK noms as well as writing good new articles. Should they be penalised for it? - The Bushranger _{Return fire}^{Flank speed} 16:06, 19 November 2010 (UTC)

A personal observation and request

I've been watching all that is going on around DYK in the last few days, and frankly I am feeling quite upset and very discouraged. The suggestions that three-quarters of what appears at DYK should be rejected strikes me as incredibly unappreciative of what has been being achieved here. I have five DYK credits, and the sub-text of some of what is being said is that four of them should have been rejected as crap. I've done a few reviews, and I am discouraged that so few comments are willing to recognise that good reviewing, with article-improving constructive comments, is being done here. I feel alot like a lot of people have come from outside the project to tell us how bad what we are doing is, instead of approaching us as valued colleagues making a positive contribution to Wikipedia; I hate the feeling of being treated like a child, and I think the work being done here deserves some acknowledgment and respect. It is really hard to feel motivated to contribute when the content generated is evidently considered near-worthless by some and when the project is obviously viewed with disdain. No one is suggesting that DYK can't be improved. I readily admit that improved accountability would be desirable and that we could do better with mentoring reviewers who are missing issues in articles that should be caught. I am quite willing to report copyvios when we see them, and as an academic I am obviously opposed to plagiarism. But, I am also opposed to burning down a project I think has value to deal with relatively isolated problem cases. Perhaps I am alone in how I am feeling, perhaps I am overly sensitive. Even so, I implore everyone to consider more the feelings of the editors whose work you seek to criticise; criticism can be constructive and given with a feeling of mutual respect and collaboration. Please, can we have some more respect for the regular DYK contributors, some recognition for the large amount of good work done, and some appreciation for the collective knowledge and experience of the DYK community? EdChem (talk) 12:56, 2 November 2010 (UTC)

Hear hear. Paralympiakos (talk) 13:05, 2 November 2010 (UTC)

I'm sorry if anyone's ended up feeling that way (and clearly some have). Any Wikipedia process getting a spotlight shined on it will expose flaws, and that's never pleasant. In the heat of that it's easy to feel that the value of the process is being ignored; but aside from a handful of people who would consider getting rid of the current approach altogether, it's really a discussion about improving something acknowledged as worthwhile. Sooner or later, the spotlight will fall somewhere else, but for the moment it's here. Rd232 ^talk 13:13, 2 November 2010 (UTC)

• Comment: Strongly agree with comment in this subsection, by Paralympiakos (talk · contribs). Cheers, -- Cirt (talk) 13:15, 2 November 2010 (UTC)

I understand how you feel EdChem, but I have been contributing to this part of the project for three years now and quite frankly I am fed up with the regular attacks on DYK. As a result I'm inclined to the view that it is time to rethink the process. I'm thinking that perhaps it is time to raise the bar substantially and to merge DYK with GA. We just don't have the manpower to thoroughly scrutinize 50 hooks a day or more and if we are going to have more scrutiny the only way to do that is to raise the bar for eligibility and thus reduce the number of submissions. Gatoclass (talk) 14:17, 2 November 2010 (UTC)

I wouldn't be that defeatist, at least without giving some of the ideas here a go first. Proposal 5, for instance, might work very well to raise reviewing contributions. PS I don't think anyone is "attacking" DYK per se. As I said above, if you put any Wikipedia process under the spotlight you'll see flaws. Rd232 ^talk 14:23, 2 November 2010 (UTC)

• Ed, you seem to be implying that everyone calling for change at DYK is actually calling for the abolition of DYK, but that's just not the case. If anyone wants to abolish DYK, this is hardly the page to do it on! Surely DYK should take the current attention as an opportunity to get views from outside its set of committed regulars, to see if the Process could operate better within the constarints that all Wikipedia processes must work under. Because if DYK is shown to be totally resistant to change on the grounds that the regulars think there's no problem and they've always done it this way, it will only give ammunition to those who will say that, in that case, DYK should be scrapped altogether as obsolete and replaced with something completely different. Physchim62 (talk) 14:29, 2 November 2010 (UTC)

Summarizing some of the problems

Perhaps it would be worthwhile to point out some of the problems with the current process. Firstly, there is the obvious problem of lack of scrutiny, which is where this entire debate originated. We simply can't parse every article for plagiarism and copyvio problems along with all the other checks, this has been proven time and again. The small number of volunteers is basically just overwhelmed with the approximately 50 hooks a day that are submitted.

The second issue is lack of quality; there are regular complaints about this, and I've always found it difficult to justify promoting crummy new articles to the mainpage when GAs get no exposure whatever. We've just had, for example, another spate of complaints about the "cookie cutter" articles users submit for Wikicup. 1500 chars minimum is setting the bar very low, and for every quality 1500 character article we get, there are probably several not-so-good ones.

Some other perennial issues are the 5-day nomination window and the x5 expansion requirement. DYK is supposed to encourage the creation of new content, but I've become aware for myself at least that involvement in DYK has actually substantially reduced my content creation, because I am constantly finding topics of interest I could write a stub on, but don't because I think maybe one day I will have enough info about that topic to make the 1500 character limit. I'm sure I can't be the only one who has found the DYK rules discouraging to content creation in this way. The x5 expansion rule is another bugbear that comes up on a regular basis.

Taking all these factors into account, it seemed to me it might be better, and a lot more straightforward, to simply raise the bar for submissions, to say, 5000 characters, and just drop all the other restrictions. This would mean a substantial reduction in submissions but an improvement in quality. But we also need a more robust process for reviewing articles. We could achieve most of this of this simply by integrating DYK with GA, which would have the added advantage of giving GA articles a path to main page exposure. I don't think that's such a bad idea, and while there may be a variety of ways to go about implementing that, I'm inclined to think that this is the kind of system we should be moving towards. Gatoclass (talk) 15:00, 2 November 2010 (UTC)

Well, Gatoclass, thank you, this is the first coherent summary of the present issue yet. How do you propose, and what do you propose, to change? Perhaps an addition that the article must have been created from scratch prior to GA nomination, and that the article must be a newly-promoted-GA? -- Cirt (talk) 15:06, 2 November 2010 (UTC)

That's interesting. On the first point about scrutiny there are a number of proposals here that might help; the others are more fundamental. I hadn't myself wanted to challenge the basic model of the existing DYK process, not least because of expected strong resistance from current DYK people (entirely normal response to radical change, I mean). But it turns out the reaction to even relatively small changes, which don't challenge the basic model, are meeting enormous resistance - in which case, we may as well think more broadly about perhaps being more radical. Raising DYK standards to 5000 characters would certainly make a dent in the rate of nominations (and lessen concerns about slots going to GA hooks raising backlog), but I'm not sure in itself if it's a good idea. I'm a bit concerned by your comment about negative incentive effects (not making new articles because of not meeting standards now), which would worsen. Rd232 ^talk 15:33, 2 November 2010 (UTC)

Well I think there would probably have to be some sort of time limit on when the GA was created, because some of the older ones would not have received the same scrutiny. What I'm really proposing is dumping the "new article" provision altogether, and replacing it with a "recently promoted" one. We'd still be getting plenty of new content because a lot of GA contributors write their articles from scratch in any case. I haven't yet considered what other caveats if any should be placed on submissions - the idea is simply that if an article has passed the GA review process recently, it's eligible. Gatoclass (talk) 15:26, 2 November 2010 (UTC)

Agreed. -- Cirt (talk) 15:26, 2 November 2010 (UTC)

Personally I always envisaged GA hooks being from newly promoted GAs. Perhaps that wasn't clear in my proposal. Rd232 ^talk 15:43, 2 November 2010 (UTC)

I took that for granted as well. I think that's the only way it would work. Lampman (talk) 14:40, 5 November 2010 (UTC)

I agree with much of what Gato said, but I am skeptical of using GAs as the solution. There are 250 editors who have each individually contributed 25 or more DYKs and thousands who have contributed smaller numbers. DYK is the only place for the new editor and editors of all types to have their content featured on the main page. This broad participation is incredibly valuable to encouraging development of editors and advancing the interest of the project. FA, on the other hand, provides a forum on the MainPage for the more devoted editors, typically a smaller number of regulars, to feature their content. By merging DYK with GA, we move in the direction of giving over DYK to a smaller number of editors who are involved in the GA process -- a group that likely overlaps with the FA process. While we can improve DYK, we should not lose its main values, which include the promotion of new content AND the encouragement/development of new editors. Shifting to GAs is a radical shift, and one that I think is in the wrong direction. I would, however, favor other ways of incrementally improving quality and diversity, e.g., increasing the size requirement to 2500, reducing the slots from 4 to 3 a day, being more disciplined in rejecting poor quality articles, and, importantly, limiting individual editors to 20 (or some such number) of hooks per month. Cbl62 (talk) 15:37, 2 November 2010 (UTC)

My point is that there is a much wider base of editors involved in DYK than GA. The 250 who have contributed 250 are the tip of the iceberg. There are thousand of editors who have had their work featured on the Main Page as a result of DYK. By allowing the smaller group involved in GA to take over some of the DYK slots, we may reduce the accessibility for the broader base of editors. Cbl62 (talk) 19:00, 2 November 2010 (UTC)

See above, a proposal which expands the accessibility for everyone, without any silly territorialism about 'taking our slots' or similar. → ROUX ₪ 19:05, 2 November 2010 (UTC)

You're misunderstanding me. It's not about territorialism (let alone silly territorialism). It's about giving access to the broadest cross-section of editors. If it were about territorialism, I would not be proposing that we limit the number of hooks each editor can have. My intent is to keep DYK open to the broadest number of people, and I simply think that adding GAs is a step in the wrong direction. There are other ways of improving quality without limiting access. Cbl62 (talk) 19:23, 2 November 2010 (UTC)

Apparently I am... because adding more GAs = more editors = more access. How on earth does that limit access? Did you read the proposal above? → ROUX ₪ 19:30, 2 November 2010 (UTC)

Giving some DYK slots to recently promoted GAs is a lot less radical than fidding with size requirements. And I'm mystified by the relevance of the "250 editors with 25 or more DYKs" if your concern is newer editors, since clearly such editors are not new. Raising size requirements substantially is a much bigger deal for new editors I would think than losing a couple of slots. And how is limiting editors to a hard maximum of X hooks per month better than requiring prolific contributors to contribute to reviewing in proportion to their nominations? Rd232 ^talk 15:43, 2 November 2010 (UTC)

Giving some slots to GA might be "less radical", but it does nothing to resolve the problem which is too many submissions and not enough reviewers. The only way to resolve that issue is to raise the bar on DYK requirements. Cbl has suggested another way of doing so above, which is restricting the number of DYKs per person per month, although this would be likely to affect only a small number of contributors. The bottom line is that we must either choose quantity or quality of review, and if we are going to improve the quality of review the only way to do that is to reduce the number of submissions somehow.

I guess we could make a start by raising the length requirement - I personally think 4k characters would probably be the minimum needed to impact the number of submissions. Then we could think about putting in place a more robust review process, closer to GA. If that works smoothly, we could then start looking at perhaps integrating DYK and GA more closely - but I do agree that trying to do it all in a single step would probably be inappropriate. Gatoclass (talk) 16:05, 2 November 2010 (UTC)

There are alternatives to increasing reviews without raising the length requirement - notably Proposal 5. And adding in a couple of slots for GAs can happen alongside any other DYK changes, since the additional workload is all at the GA end. Rd232 ^talk 16:35, 2 November 2010 (UTC)

Proposal 5 does nothing to address the quality of reviews, which is the issue here. In fact, by having inexperienced users do reviews, the quality will get worse, not better.

And I really think we should drop the GA discussion for now. We should focus on one thing at a time, and the main issue here is how (or perhaps even "if") we are going to improve the quality of reviews to ensure that plagiarism and copyvios do not make it to the mainpage. And as I've said, the only practical way to do that is reduce the number of submissions, by whatever means. Anyhow, it's late here and I will be logging off shortly, so I guess this discussion will have to be continued at another time. Gatoclass (talk) 16:51, 2 November 2010 (UTC)

Indeed, proposal 5 is about quantity of reviews (inexperienced reviewers as an issue is from a separate idea to limit established editor involvement, so let's not cross wires). In terms of quality, other proposals, including better transparency and requiring each hook to be signed off by two reviewers, can address that. Overall, part of the problem with this discussion is requiring each proposal to fix everything; some of them fit together jigsaw-style to be complementary in how they address different issues. I've tried to clarify that a little on the new subpage. Rd232 ^talk 22:19, 2 November 2010 (UTC)

I also agree with Gato's analysis and ideas. Recently promoted GAs are more deserving of the home page than DYKs, but then each of them promote different worthy ideals. WP:GA people could work on approving the GAs for the home page. I'm not sure about changing the size requirements for the new articles. But GAs and what we now call DYKs would look good together under a "DYK?" banner:

Did you know....

From Wikipedia's Good Articles:

• ... that the Battle of St. Louis ) was an unsuccessful British-led attack on St. Louis, Missouri on May 26, 1780, during the American Revolutionary War?
  
• ...etc.?
  

From Wikipedia's newest articles

• ... that Ogden H. Hammond, the father of New Jersey congresswoman Millicent Fenwick, survived the sinking of the RMS Lusitania, though his wife did not?
  
• ...etc.?
  

First Light (talk) 16:19, 2 November 2010 (UTC)

That's exactly what I had in mind (except add "newest" in front of Good Articles). Also I imagined newest article hooks being above Good Article hooks, but it's probably better this way (since the TFA will be just above, gives a certain logic). Rd232 ^talk 16:35, 2 November 2010 (UTC)

I don't think two separate sections is the answer. In any case, the GA question is still very much a secondary concern right now. Gatoclass (talk) 16:24, 2 November 2010 (UTC)

This is a bit OT, though appropriate to the section heading. Not sure where else to put it. While I would love to see some kind of "most improved" articles featured on the main page, along with newest articles, I'm not so sure that the DYK process really is broken. Sure, because of the volume of nominations and scarcity of reviewers some things will slip through the cracks. But the average DYK article is much much much better than average Wikipedia (non-stub, not total crap) article simply because the DYK process enforces SOME kind of quality standard (and yes, I would like that standard to be upped - but we're talking gradual reform here, not revolutionary pyres). Having followed this through for the past few days, I get a sense that this particular article slipped through the cracks of this process PRECISELY because of who the author was. People assumed that an arb would never do such a thing and assumed good faith a little too much. There's no reason to radically change DYK - there is just a reason to remind reviewers that they should scrutinize articles they verify no matter who wrote them. More generally, something like an audit of a random selection of DYKs from the past year or so might not be a bad idea, if someone was willing to put in the work. Otherwise we're just speculating here without real information.radek (talk) 18:46, 2 November 2010 (UTC)

• I agree with Cbl62's point about DYK representing the only venue for a new or less experienced editor to get their article on the main page. GA and FA are extremely daunting processes; I've been editing Wikipedia for more than 5 years and closing in on 8,000 edits, and I don't think I've made any major contributions to a Featured Article, and only to Good Articles as part of collaborations/improvement drives. I think it's important to continue to offer some such venue that's reasonably accessible. cmadler (talk) 19:34, 2 November 2010 (UTC)
  • And where are we saying that new articles will no longer be allowed at DYK? Almost everyone here is talking about adding GA. Honestly I am forced to wonder sometimes if anyone ever reads anything before opining. → ROUX ₪ 19:40, 2 November 2010 (UTC)
    • Perhaps you haven't been reading the discussion. In addition to some -- a few -- people suggesting the replacement of DYK with GADYK, there has been a bunch of discussion, particularly in this section, about significantly raising the DYK requirements. In the very first comment in this section, Gatoclass suggested raising the DYK minimum length from 1500 to 5000 characters of prose. cmadler (talk) 20:30, 2 November 2010 (UTC)
  • The Main Page offers limited incentives because of limited space. Nobody has tried to make a coherent argument why those incentives should be applied solely to FAs (just 1 per day) and to new articles (how many per day now? 30 odd? can't spare a single one of those slots for GAs??). The idea that the new article DYKs are a great thing for new editors is hogwash - it's largely incentivising established editors. No process like that can do much for new editors unless it's really specifically targeted at them (eg in welcome messages and such), and perhaps even excluding established editors. (In fact, if the New Editors thing is such a big deal, why not just put a fixed upper limit on DYKs per editor? i.e. "You've got 5 DYKs? Well done, now make way for a new editor." Somehow I don't see established DYKers jumping on that idea.) As to the minimum length - I've disagreed above with that, partly on Gatoclass's own argument about disincentivisation, which would be worsened by a higher character quota. Rd232 ^talk 21:01, 2 November 2010 (UTC)

===Moving the discussion to another page=== So, everyone agrees that DYK has a lot of problems, but there's no consensus on a solution. Regardless, this discussion has been getting too large, and I propose we split it out into: Wikipedia talk:Did you know/DYK reform.--hkr Laozi speak 20:29, 2 November 2010 (UTC)

Why move it? It's a long discussion, yes, but it's in the right place. Physchim62 (talk) 22:18, 2 November 2010 (UTC)

A lot of the DYK reform discussions are scattered all over this talk page, and I think it would be better if we separated them from the non-DYK reform topics and centralise them on a subpage. It would make reading and finding the relevant discussions much easier, for both pages.--hkr Laozi speak 22:22, 2 November 2010 (UTC)

What problems? I've been told that "pushing crap onto main page" is bad. Excellent, I'll leave it to those who don't. The more editors follow this route, the less crap you will have to deal with. Your real or imaginary problem has an easy solution: no nominations = no problem. East of Borschov 22:45, 2 November 2010 (UTC)

Err... my problem? All I've stated is that people have brought up complaints of DYK, and that the discussion should be centralised in a subpage. This section is neither about or addresses the concerns of DYK, it is only a suggestion on having a subpage. I've never said that DYK is "crap", but the plagiarism concerns are real, and need to be addressed, preferably in one location instead of being scattered all around Wikipedia.--hkr Laozi speak 22:52, 2 November 2010 (UTC)

You've said "everyone agrees", that's enough. I, for one, never agreed: DYK has absolutely no problems of its own. It reflects the problems of wikipedia as a whole. East of Borschov 23:08, 2 November 2010 (UTC)

East of Borschov put it well, I was reading to the end of this discussion to say pretty much the same thing. At the DYK ANI discussion, one user discovered and pointed out that the TFA that day (Oct 31) had blatant plagiarism in it. This is a problem throughout Wikipedia and I agree that DYK articles are better than the average stub or unreferenced article. Being this is a good place to catch such problems, DYK should slow down a bit and pay attention to the review more. I don't expect ever article to be free of problems though, nor do I expect an FA or GA too. I am for change in DYK but completely merging it with another project isn't necessarily the answer. DYK has a use and a GA can make its way through DYK as well. A GA could be a substitute for 5x expansion. I am not trying to knock other projects, but Wikipedia isn't perfect and so aren't reviewers. I am for 18 hooks a day, 1 nom per editor per day and an at least 500 character increase in minimum article size. I do agree as well that some hooks are boring and their interest may not be seen or just found in a small majority.--NortyNort (Holla) 10:02, 3 November 2010 (UTC)