You are on page 1of 12

IIM BANGALORE

Regression Analysis
of Pricing of IPL
Players
Project Report

Himanshu Rana, 1211344

Prashant Yetukuri, 1211364

Amandeep Singh, 1211323

Soumya Basu, 1211385


Pricing of Players in the Indian Premier League

Executive Summary

In the project, price for the players in IPL were analysed against various
factors. All the variables were taken for multivariable regression analysis
using SPSS, using the backward method.

It was observed that not all factors that were directly related to their
performance on the field drove the price of a player. There were specific
factors which had a direct impact on players remuneration. These factors
ranged from performance measure of players such as Strike rate (in case
of a batsman) to physical attributes of players such as age. We applied
techniques of multiple linear regression to determine such factors which
were deterministic in pricing the players.

For most of the specific questions we referred to the general relationship


that was found using the regression analysis and this result was further
confirmed by simple linear regression.
Best Regression model(s)

The following are the independent variables which are derived after doing
regression analysis.

Where,

Country = 1 for India, 0 for non-India


Age_3=1 for age < 25 years, 0 for others
T_Runs is test run scored
Runs_ODI is run scored in ODIs
ODI_Wickets is wickets taken in ODIs
RUN_S is runs scored
BASE_PRICE is base price
YEAR = 0 for year <2011, 0 for others

Linear regression model has been developed using Backward variable


selection method. The criterion used for Backward method is Probability of
F-to-remove >= 0.100
As seen from the above table in our model the R Square value of is 0.618
and Adjusted R Square value is 0.592.

Team variable is removed


Cricket in the T20 format is considered a young mans
sport, is there evidence that the players price is
influenced by age?

From our analysis we have seen that the price of a player is greater
if the player is less than 25 years of age.

Variabl B Std.error Beta t Sig.


e
AGE 173039.2 76474.8 .140 2.263 .025
27 62

As we can see from the data the Significance value is less than 0.05
which indicates that the players price is influenced by age.
What is the impact of ability to score SIXERS on the
players price?

According to our analysis, there is no significant linear relation was


observed as per the below analysis.(Ability is taken as runs scored in
sixers / runs scored)

SOLD PRICE vs ABILITY TO SCORE SIXER


2000000
1800000
1600000
1400000
1200000 SOLD PRICE
1000000 Linear (SOLD PRICE)
800000
600000 R = 0.02
400000
200000
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
What is the impact of the predictors batting strike rate
and bowling strike rate on pricing? Identify the predictor
that has the highest impact on the price of players.

Batting strike rate and bowling strike rate do not have significant impact
on the pricing. This can be explained by very low R 2 = 0.014 for bowling
strike rate and R2 = 0.034 for batting strike rate.

Bowling Strike Rate vs Sold Price

2000000
1800000
1600000
1400000
1200000
1000000
800000 Linear ()
R = 0.01
600000
400000
200000
0
0.00 20.00 40.00 60.00 80.00 100.00120.00
Batting Strike Rate vs Sold Price

2000000
1800000
1600000
1400000
1200000
1000000
800000 Linear ()
R = 0.03
600000
400000
200000
0
0.00 50.00 100.00 150.00 200.00 250.00

The base price has the highest impact on Selling Price of the players
which can be explained by highest R2 value compared to others.

2000000
R = 0.27
1800000
1600000
1400000
1200000
1000000
800000 Linear ()

600000
400000
200000
0
0 500000 1000000 1500000

When compared to others, the base price has the highest R square value
Identify the player who is highly overpaid and the player
who is highly underpaid.

As per analysis, error is highly negative for SC Ganguly and highly


positive for SS Tiwary. Comparing their credentials, SC Ganguly is
highly underpaid and SS Tiwary is highly overpaid.

ODI
A OD BAS SOL
T- - RU Auct PREDIC
cou g I- E D ERR
. RU RU NS- ion TED
ntry e WK PRI PRIC OR
NS NS- S Year PRICE
3 TS CE E
S
Gan 11 -
guly, 72 36 10 13 200 124095 4000 8409
SC 1 1 12 3 0 49 1 000 5.522 00 56
Tiwa
ry, 83 100 579363 1600 1020
SS 1 0 0 49 0 6 1 000 .062 000 637

Are players of Indian origin paid more than players from


other countries?

To identify the relation of being Indian with price, we have introduced


country which takes value of 1 for India and 0 for others. It is observed
that a positive coefficient for country variable is obtained and also this
figure is significant considering the value of probability of F for Country
variable.

Variabl B Std.erro Bet t Sig.


e r a
COUNT 197426. 53353.. .239 3.70 .000
RY 227 170 0
This in synch with well-known notion that for same attributes Indians are
paid more than their foreign counterparts.
If a regression model was built after removing Virat
Kohli from the sample, what would be the impact on the
co-efficient for the predictors, INDIA and L25? How
would you interpret this impact?

After removing Virat Kohli from the sample, value of country


increases from 197426.277 to 219100.800 and Age_3
(Age<25years) gets insignificant. Insignificance of Age 3 (Age<25
years) can be explained by removal of overwhelming factor of Virat
Kohli.

How much should Mumbai Indians offer Sachin


Tendulkar if they would like to retain him? Is the model
sufficient to predict the price of Icon players?

According to model, the sold price of Sachin is 1800000 but fair


value of Sachin Tendulkar is 1598610.008 which minimum amount
should be paid. Icon player should get 15% premium over second
highest pay. But this factor is not considered in our analysis. Hence,
our model cant predict the value of icon players.
APPENDIX 1:

You might also like