Scribd Logo
Upload

Welcome to Scribd!

  • 8 views

    Lt5 Uncertainty

    Uploaded by

    Rubles
    This document discusses principles for making decisions under uncertainty. It defines dominance, where choosing one option is always at least as good as another across all possible outcomes. It also defines the maximin principle, which involves choosing the option with the least bad worst-case outcome. Finally, it defines the minimax regret rule, which is to choose the option with the smallest possible maximum regret compared to the best outcome. Examples are provided to illustrate each definition.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Lt5 Uncertainty

    Uploaded by

    Rubles
    8 views2 pages
    This document discusses principles for making decisions under uncertainty. It defines dominance, where choosing one option is always at least as good as another across all possible outcomes. It also defines the maximin principle, which involves choosing the option with the least bad worst-case outcome. Finally, it defines the minimax regret rule, which is to choose the option with the smallest possible maximum regret compared to the best outcome. Examples are provided to illustrate each definition.

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document discusses principles for making decisions under uncertainty. It defines dominance, where choosing one option is always at least as good as another across all possible outcomes. It also defines the maximin principle, which involves choosing the option with the least bad worst-case outcome. Finally, it defines the minimax regret rule, which is to choose the option with the smallest possible maximum regret compared to the best outcome. Examples are provided to illustrate each definition.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    8 views2 pages

    Lt5 Uncertainty

    Uploaded by

    Rubles
    This document discusses principles for making decisions under uncertainty. It defines dominance, where choosing one option is always at least as good as another across all possible outcomes. It also defines the maximin principle, which involves choosing the option with the least bad worst-case outcome. Finally, it defines the minimax regret rule, which is to choose the option with the smallest possible maximum regret compared to the best outcome. Examples are provided to illustrate each definition.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 2

    DECISIONS UNDER UNCERTAINTY

    BEN BLUMSON

    Definition 1 (Dominance). A choice weakly dominates another if and only if


    its consequences are at least as preferred in every outcome. A choice strictly
    dominates another if and only if its consequences are strictly preferred in
    every outcome.1

    Example 1. You are looking for a job as an accountant. You know for sure
    you can get a job at P W C, but you doubt whether you can get the job at
    EY . There are no costs to applying for either job, or to turning down an
    offer if you get both. Q: What should you do? A: You should apply to both,
    since that weakly dominates applying to one.

    Example 2. You want toast for breakfast, and you prefer avocado toast
    to kaya toast, and you prefer kaya toast to vegemite toast. You are certain
    that you can get kaya toast at Ya Kun, but you’re not sure whether you
    can get avocado toast at Jimmy Monkey, or if they will only have vegemite.
    Jimmy Monkey is far from Ya Kun, so you can’t go to both.

    Definition 2 (The maximin principle). For each act, look at its worst pos-
    sible consequences. Then choose the act which has the least bad worst
    consequence.2 There is also a lexicographic version of this principle, which
    breaks ties between least bad consequences by looking at the next least bad.3

    Date: September 8, 2021.


    1
    See for example Resnik 1987, p. 25 or Peterson 2009, pp. 41–2.
    2See for example Resnik 1987, p. 26 or Peterson 2009, p. 43.

    3See for example Resnik 1987, p. 26 and Peterson 2009, p. 45.

    1
    2 BEN BLUMSON

    Example 3 (The Veil of Ignorance4). You are deciding what kind of society
    you will be born into. Some societies are very wealthy, but the wealth is
    distributed very unevenly. Other societies are less wealthy, but more egali-
    tarian. How does the minimax principle advise you to make this tradeoff?
    What does the maximin principle advise you to do if you can take $1 from
    the poorest person in order to give $1, 000, 000, 000 to each other person?
    What does classical decision theory, also known as maximizing expected
    utility, recommend in the same situations?

    Definition 3 (The minimax regret rule). The maximum regret in an out-


    come is the difference between the utility of the best and worst consequences
    in that outcome. The minimax regret rule says to choose the act with the
    least maximum regret.5

    References

    Harsanyi, John C. (1953). “Cardinal Utility in Welfare Economics and in the


    Theory of Risk-Taking”. Journal of Political Economy 61.5, pp. 434–435.
    doi: 10.1086/257416.
    – (1975). “Can the Maximin Principle Serve as a Basis for Morality? A
    Critique of John Rawls’s Theory”. The American Political Science Review
    69.2, pp. 594–606. doi: 10.2307/1959090. JSTOR: 1959090.
    Peterson, Martin (2009). “Decisions under Ignorance”. An Introduction to
    Decision Theory. Cambridge: Cambridge University Press, pp. 40–63. doi:
    10.1017/CBO9780511800917.004.
    Rawls, John (1971). “The Original Position”. A Theory of Justice. Harvard
    University Press, pp. 102–168. doi: 10.2307/j.ctvkjb25m.7. JSTOR:
    j.ctvkjb25m.7.
    Resnik, Michael D. (1987). Choices: An Introduction to Decision Theory.
    University of Minnesota Press. JSTOR: 10.5749/j.ctttshgd.
    4See Harsanyi 1953, Rawls 1971 and Harsanyi 1975.

    5See Resnik 1987, pp. 30–2 and Peterson 2009, pp. 49–3.

    You might also like