Scribd
Upload
0 views1 page

Problem 3 Approach

The document describes a linear regression approach for predicting cricket match scores, focusing on matches with 10 overs or less. Key factors influencing the score include the quality of the batting and bowling teams, as well as the impact of wide and no ball runs, total runs, and wickets. The model is structured using a data frame with binary representations for teams, ensuring that only relevant teams are included in the training set based on the test set.

Uploaded by

shuvradeepkgp.24
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
0 views1 page

Problem 3 Approach

The document describes a linear regression approach for predicting cricket match scores, focusing on matches with 10 overs or less. Key factors influencing the score include the quality of the batting and bowling teams, as well as the impact of wide and no ball runs, total runs, and wickets. The model is structured using a data frame with binary representations for teams, ensuring that only relevant teams are included in the training set based on the test set.

Uploaded by

shuvradeepkgp.24
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

The approach I used for predicting the score is linear regression.

Since most of the test cases


have around 10 overs or less, so for the test and training data sets to have similar conditions, I
limited the training set to 10 overs per match. I assumed the following factors affect the total
score of the game
● Batting team: A good batting team will generally put up a higher likewise a bad batting
team will have a low total
● Bowling team: A good batting team will generally restrict the opponents to a lower
likewise a bad bowling team will have a high target
● Total number of wide runs at a certain point: This gives the batting team an extra ball to
play with and is slightly more advantageous than other extras(excluding no ball)
● Total number of no ball runs at a certain point: This gives the batting team an extra ball
to play with and gives the batsman a free hit so it is slightly more advantageous than
other extras(including wide)
● Total number runs at a certain point: Will tell us how the match is progressing in terms of
run rate
● Total number of wickets at a certain point: More wickets mean lesser runs

Thus I used these values in the data frame while structuring the model.
Batting team and bowling team are represented by keeping a 0/1 against all the team names.
Also the teams used in the training set are limited to the ones present in the test set so in case
there are extra or lesser teams in new test sets, the non essential columns must be removed
from the pickle file.

You might also like