DATA ANALYSIS

IN SPORTS.
 S.K. Somaiya
college
Group members :
❖ Varsha Jadhav
❖ Vedang Sawant
❖ Tai Mohammed Moheed
 Objective
❖ To conduct EDA and determine the best-offs of various
variables.
❖ Comparison of average money spent by each team in each
year.
❖ Toss and match wins, home and away wins comparison
❖ Prediction of position of a player in bowler and batsman
standings, total runs scored by each team.

Methodology & Techniques.

❖ Exploratory Data Analysis.
❖ Analysis Of Variance.
❖ Multiple linear regression model and Random Forest
Regression.
❖ R-software, Excel, Python and SPSS software were used to
perform Data Analysis.
❖ The dataset was collected from Kaggle.
 Top 3Lorem
hitterIpsum
batsman Top 3 Batsman with
Lorem Ipsum
with most 6’s & 4’s. most 100’s.

Shikhar
CH Gayle
Lorem Ipsum Lorem Ipsum Virat
LoremKohli
Ipsum CH Gayle
Lorem Ipsum Virat
Lorem Kohli
Ipsum David Warner
Lorem Ipsum
Dhawan

TopLorem
3 Man of the Top 3 Paid
Ipsum Lorem Ipsum
Matches. Players

Kevin
CH Gayle
Lorem Ipsum ABLorem
de Villiers
Ipsum MS Dhoni
Lorem Ipsum Yuvraj Singh
Lorem Ipsum Ben Stokes
Lorem Ipsum Lorem Ipsum
Peterson
 Top 3 Bowlers
Lorem Ipsum Top 3 Teams.
Lorem Ipsum

Kolkata
Lasith Piyush Mumbai Chennai Super
Lorem Ipsum Amit
LoremMishra
Ipsum Lorem Ipsum Lorem Ipsum Knight
Lorem Ipsum Lorem Ipsum
Kings.
Malinga Chawla Indians.
Riders.

Top 3 Venues
Lorem Ipsum TopLorem
3 Umpires.
Ipsum

M
Eden Wankhede HDPK
Lorem Ipsum Lorem Ipsum Chinnaswamy
Lorem Ipsum S Ravi
Lorem Ipsum Lorem Ipsum C Shamsuddin
Lorem Ipsum
Gardens stadium stadium Dharmasena
 Maximum Toss Winners

Is Toss Winner Also the Match Winner?

Choice after winning the toss.

Opted to Bowl 1st 61% time.
Opted to Bat 1st 39% time.
 Number of times bowling
Number of times batting team team has scored above 200
has scored above 200 runs. runs.
Individual Scores of Top Batsman in each Innings.
Score Distribution For Each Teams by Innings.
 Match status of all seasons Home & Away wins
Sunrisers Hyderabad Kolkata Knight Riders

Royal Challengers Bangalore Royal Challengers Bangalore

Rajasthan Royals Kings XI Punjab

Mumbai Indians Rajasthan Royals

Kolkata Knight Riders Sunrisers Hyderabad

Kings XI Punjab Delhi Capitals

Delhi Capitals Chennai Super Kings

Chennai Super Kings Mumbai Indians

0 50 100 150 200 250 0 10 20 30 40 50 60 70

Won Lost tied no result away_win_percentage home_win_percentage

Toss & Match PRICE (IN CRORES)

60 16
14.5
50
11 11.5 11.4
40 9.72 9.5
8.4
30
5 5.3
20

10

0
CSK MI KKR RR RCB KXIP SRH DC

toss win match win toss win match lost

 Comparison of average
money spent in each year.
 Variables under study.
❖ For prediction of total score at the end of the
inning variables such as runs, wickets, overs, runs
in last 5 overs, wickets in last 5 overs, striker &
non-striker runs are considered.

❖ For prediction of batsman position in batsman

standings variables such as matches, innings, not
outs, highest score, 100s, 50s, 4s & 6s are
considered.

❖ For prediction of bowler position in bowler

standings variables such as matches, innings,
overs & wickets are considered.
 Prediction of Strike Rate.
❖ Variable under consideration are matches,
innings, not outs, highest score, 100s, 50s, 4s & 6s.
❖ R-Squared value is 42.8% when multiple linear
regression was fitted to the model.
❖ R-Squared value is 87.89% when Random
Forest Regression was fitted to predict the
strike rate.
❖ Random Forest regression proved to be a
better fit than multiple linear regression model.
Prediction of position of
batsman & bowler in
standings
Correlation Plot Between Variables
Prediction of Total Runs for
Each Inning.
❖ Variables under consideration to predict total
runs are runs, wickets, overs, runs in last 5 overs,
wickets in last 5 overs, striker & non-striker runs.
❖ R-Squared value is 50.88% when multiple
linear regression was fitted to the model.
❖ R-Squared value is 67.46% when Random
Forest Regression was fitted to predict the
strike rate.
❖ Random Forest regression proved to be a
better fit than multiple linear regression model
to predict the total runs for each inning.
 Limitations
❖ Data which was considered in analysis is from
2008-19.
❖ Lack of information on pitch & weather
conditions in matches.
❖ Some teams were not participated in all
seasons from 2008-19.
❖ Data structure was not appropriate as it can
large number of empty cells that was replaced
by zero initially.
❖ Complete record of money spent by each
team in all seasons was unavailable.
 Observations
❖ CH Gayle is the best player with maximum runs (5 centuries), maximum
boundaries and maximum man of the match title but than too he is not the
most expensive player.
❖ MS Dhoni and Gautam Gambhir have never scored a Century.
❖ V Kohli has played only for Royal Challengers Banglore in all seasons.
❖ Mumbai Indian is the best team and Eden Gardens is the best venue and S
Ravi is the favorite umpire.
❖ Yuvraj Singh is the highly paid cricketer and Lasith Malinga is the bowler
with maximum wicket.
❖ Mumbai Indians seem to be very lucky having the highest win in tosses
followed by Kolkata Knight Riders. Pune Supergiant's have the lowest wins
as they have played the lowest matches also. This does not show the higher
chances of winning the toss as the number of matches played by each team
is uneven.
❖ RCB is the team which cross 200 runs while setting up a target maximum
number of times followed by CSK, KXIP and MI. While chasing the target
KXIP cross the target of 200 maximum times followed by RCB.
❖ According to Box & Whisker plot the batting by CSK in innings 1 looks to be
the best. Innings 2 also conveys the same story.
❖ KKR spends the maximum money on teams followed by MI and KXIP spend
least money.
❖ Correlation between total runs and the striker batsman runs is maximum
with 0.38, followed by run last 5 overs with 0.37 and non-striker batsman
runs with 0.29 and these validates our data is true.
 Interpretation
• The toss winner is not necessarily the match winner. The match winning
probability for toss winning team is about 50%-50% almost same.
• MI & CSK won maximum matches in all seasons of IPL. RCB, DC & KXIP
lost maximum matches as compared to wins. Maximum number of tie
had occurred in RCB & DC matches. Maximum no results were found in
RCB matches.
• Match wins of MI & CSK are more as compared to toss wins. Match
wins of DC & KXIP are less as compared to toss wins. SRH & RCB had
approximately equal amounts of match & toss wins.
• KKR, CSK & SRH won more away matches as compared home matches.
MI won approximately equal home & away matches.
• Average Money Spent during each season of IPL is same but money
spent on each team differs and this is validated by Two-Way ANOVA
Model.
• Random Forest Regression is the best fit to predict the total runs with
67.46% accuracy, also Random Forest Regression seems to be the best
fit to predict strike rate with 87.89% accuracy.
• Multiple linear regression is the best fit to predict both bowler &
batsman position in standings with 87.47% and 70.01% accuracy
respectively.
Thank you !