ISSN (Online) 2581-9429

IJARSCT
International Journal of Advanced Research in Science, Communication and Technology (IJARSCT)

Volume 3, Issue 1, April 2023

Impact Factor: 7.301

Cricket Match Prediction using Machine Learning

Ms. Shrunkhala Wankhede1, Vaishnavi Desai2, Vaibhavi Desai3, Disha Tangade4
Assistant Professor, Department of Computer Science and Engineering1
Student, Department of Computer Science and Engineering2,3,4
Priyadarshini Bhagwati College of Engineering, Nagpur, Maharashtra, India

Abstract: As cricket is the most popular game in the world. T20 and ODI are most loved by people. The
IPL was launched in 2008. So we decided to develop a machine learning model that can predict both game
scores and outcomes. In this paper, a model with three methods is proposed. The first is the score
prediction, the second is his IPL match win percentage and the last is his ODI match win percentage. The
model used supervised machine learning. We used previous statistics to build this model. The application is
trained on historical data from previous games and uses these algorithms to make accurate predictions for
upcoming games. Finally, several predictors are identified that can be used for data analysis. I used jypyter
notebook. A label encoder was used for pre-processing. I created a predictive model using a machine
learning algorithm. Algorithms used are decision tree classifier, random forest classifier, etc. for team
winning prediction, Lasso regression, and ridge regression for score prediction. The web application is
designed to provide real-time predictions for upcoming matches, allowing cricket enthusiasts to stay up to
date with the latest predictions for upcoming matches. Overall, this project has the potential to
revolutionize the world of cricket by providing accurate match predictions. Our web application enables
cricket enthusiasts to make informed decisions and improve their overall viewing experience.

Keywords: Decision Tree Classifier, Random Forest Classifier, Lasso Regression, Logistic Regression,
Gaussian NB, Gradient Boosting Classifier

I. INTRODUCTION
Cricket is a sport that is widely played and watched across the world, and two of the most popular formats of cricket are
IPL (Indian Premier League) and ODI (One Day International). These formats attract millions of fans globally, and
with the growing popularity of machine learning and data science, it has become possible to predict the outcome of
cricket matches using advanced statistical models. In this project, we aim to build a web application that uses machine
learning algorithms to predict the results of IPL and ODI matches. Our web application will provide real-time
predictions for upcoming matches based on various factors such as team composition, player statistics, pitch conditions,
and weather. To build our web application, we will be using Python, a popular programming language for data science
and machine learning. Python provides a wide range of libraries and tools for data analysis and machine learning,
making it an ideal choice for building predictive models. We will be using popular machine learning algorithms such as
regression analysis, decision trees, and support vector machines to predict the outcome of matches. Our application will
be trained using historical data from previous matches and will use these algorithms to make accurate predictions for
upcoming matches. One of the challenges of predicting cricket matches is the number of variables that can affect the
outcome of a match. Factors such as the form of the players, the pitch conditions, and the weather can all have a
significant impact on the outcome of a match. To address this challenge, we will be using a combination of statistical
analysis and machine learning algorithms to build our predictive models. Our models will take into account multiple
factors and will use a combination of historical data and real-time information to make accurate predictions.

II. LITERATURE SURVEY

We studied papers based on the area of Cricket prediction. Following papers are related to our area i.e. Cricket
prediction.
1] Monoj Ishi explained “WINNER PREDICTION IN ONE DAY INTERNATIONAL CRICKET MATCHES USING
MACHINE LEARNING”. In this paper explained the focus is given to the prediction of victory in one day international
game of cricket using machine learning. The concept of ensemble algorithm using voting and stacking classifier is used
Copyright to IJARSCT DOI: 10.48175/568 507
www.ijarsct.co.in
 ISSN (Online) 2581-9429
IJARSCT
International Journal of Advanced Research in Science, Communication and Technology (IJARSCT)

Volume 3, Issue 1, April 2023

Impact Factor: 7.301
for prediction with machine learning algorithms. 2] I.P. Wickramasinghe (2015) explained "Guessing the performance
of batsmen in test cricket," in this paper explained how the performance of batsmen in a series of tests can be predicted
using long-distance and long-range methods. In this paper they collect sample data from test cricket batsmen who
played during the 2006-2010 season in nine overseas teams and show that these sample data shows long and high-level
formations of three levels. 3] Jhanwar and Paudi explained about "Predicting the Outcome of ODI Cricket Matches: A
Team Composition Based Approach,” It predicts the outcome of a cricket match by comparing the strengths of the two
teams. For this, they measured the performances of individual players of each team. They developed algorithms to
model the performances of batsmen and bowlers where they determine the potential of a player by examining his career
performance and then his recent performances.4] The paper “Predicting Outcome of ODI Cricket Games” predict the
outcome of One Day International games in cricket by proposing a numerical model. This paper is published by Kevin
Desai, Siddhant Doshi, and Surekha Dholay. The conclusion of their project is in the form of win-loss percentage for
both the participating teams at a given venue. 5] In the paper “Cricket Score Prediction Using Machine Learning”
utilising machine learning algorithms such as SVM, Random Forest Classifier (RFC), Logistic Regression, and they
have suggested a model for predicting the results of the IPL matches.6] The paper “Sport analytics for cricket game
results using machine learning” published by Kumash Kapadia , Hussein Abdel-Jaber investigates machine learning
technology to deal with the problem of predicting cricket match results based on historical match data of the IPL.
Influential features of the dataset have been identified using filter-based methods including Correlation-based Feature
Selection, Information Gain (IG), ReliefF and Wrapper.7] Prasad Thorat, Vighnesh Buddhivant and Yash Sahane
predicted the cricket score. So, to get an accurate score prediction they should have a system that can predict the first
innings score more effectively. Lots of people like watching cricket and they also like to predict the final score. This
research paper focuses on an accurate prediction of cricket scores for live IPL matches considering the previous dataset
available and also considers the various factors that play an important role in the score prediction.

III. DATASET FEATURE

The approach used here is based on ML. Therefore, the basic requirements of any ML algorithm are a dataset, training
on that dataset using the algorithm, and testing the model. So we imported some of the records from Kaggle. After that,
we compute the accuracy and use the random forest classifier for win probability prediction to improve the accuracy.
Use linear regression for score prediction.

3.1 IPL Score Prediction

Dataset which we have used for this is ipl.csv. Dataset contains 76015 rows and 15 columns. This dataset has data from
2008 to 2017. We have selected 8 features from the dataset. Out of the eight features, one is target value and others are
used to predict the target value.
Features are as follows
1. Batting team
2. Bowling team
3. Over (Greater than 5)
4. Runs
5. Wickets
6. Runs scored in previous 5 overs
7. Wickets taken in previous 5 overs
8. Total runs

3.2 IPL Winner Prediction

Datasets used for the winner prediction are matches.csv and deleveries.csv. To predict the winner of IPL match, we
have used the dataset from Kaggle. First dataset contains 257 rows and 18 columns. Second dataset contains 179079
rows and 21 columns. These two dataset has been combined to form a one final dataset. From the dataset, we have
selected some features which help to give better prediction. We have selected 8 features from the dataset. Out of the

Copyright to IJARSCT DOI: 10.48175/568 508

www.ijarsct.co.in
 ISSN (Online) 2581-9429
IJARSCT
International Journal of Advanced Research in Science, Communication and Technology (IJARSCT)

Volume 3, Issue 1, April 2023

Impact Factor: 7.301
eight features, one is target value and others are used to predict the target value. Some attributes have been created
using given attribute such as current run rate, required run rate, runs left, balls left.
Features are as follows
1. Batting team
2. Bowling team
3. City
4. Total runs
5. Wickets
6. Overs finished
7. Ball
8. Result

3.3 ODI Winner Prediction

Dataset which we have used is .csv. It contains 128212 rows and 16 columns. Some features have been selected which
helps to give better prediction. We have also removed some columns from the dataset which are redundant. For the
final prediction, we have selected 10 attributes. Among 10 attribute, one is target value and others are used to predict
the outcome.
Features are as follows
1. Batting team
2. Bowling team
3. Batsman
4. Bowler
5. Current runs
6. Wickets
7. Overs finished
8. Striker runs
9. Non- striker runs
10. Result

IV. FEATURE SELECTION

Feature Selection is a process in which we select an optimal set of features from the input features set by using feature
selection techniques. By removing redundant features, we reduce the dimension of data and we can improve the time
and space complexity of data. Feature selection improves the performance of the model and saves time and space.

V. BLOCK DIAGRAM AND METHODOLOGY

The first step is to import the cricket dataset. After importing dataset, data has been pre-processed. We converted
categorical features using one Hot Encoding method. We further pre-process data by checking null value and replace it
with the mean, median values of respective column. As the dataset has some redundant columns, so we decided to do
feature selection. Features which we have selected will be helpful to give better prediction. After that, Visualization of
data has been done. So that we can give better prediction. Then, we have splitted data into train set and test set. For
training set, 80% of data has been taken. For testing set, 20% of data from dataset has been taken. Then, prediction
model is created. Machine Learning algorithms are applied. Then, we check accuracy of algorithms by using taking the
ratio of the predicted testing data and the actual testing data. After that, we decided to go for one algorithm which has
good accuracy.

Copyright to IJARSCT DOI: 10.48175/568 509

www.ijarsct.co.in
 ISSN (Online) 2581-9429
IJARSCT
International Journal of Advanced Research in Science, Communication and Technology (IJARSCT)

Volume 3, Issue 1, April 2023

Impact Factor: 7.301

Fig. 1.Block Diagram and Methodology

VI. ALGORITHM
6.1 IPL Score Prediction
1. Regression: Regression is a method for understanding the relationship between independent variables or
features and a dependent variable or outcome. Outcomes can then be predicted once the relationship between
independent and dependent variables has been estimated. Regression is a field of study in statistics which
forms a key part of forecast models in machine learning. It’s used as an approach to predict continuous
outcomes in predictive modelling, so has utility in forecasting and predicting outcomes from data. Machine
learning regression generally involves plotting a line of best fit through the data points. The distance between
each point and the line is minimised to achieve the best fit line.
2. Ridge Regression: Ridge regression is a method of estimating the coefficients of multiple regression models
in scenarios where the independent variables are highly correlated.This method performs L2 regularization.
Multicollinearity problems cause the least squares method to become unbiased, have high variance, and have
predicted values that deviate significantly from the actual values.
3. Lasso Regression:-LASSO regression is a regularization technique. This is used via regression strategies for
accurate predictions. This model uses shrinkage. Shrink shrinks the data values to a center point as the mean.
The lasso method promotes simple, straight and thin models (that is, models with few parameters). This
particular style of regression is suitable for models with a high level of multiple regressions and for automating
certain aspects of the model like select variables/remove parameters.
4. Linear Regression:-Linear regression is used to predict continuous values. Certain known parameters are
given to a machine learning algorithm to predict continuous values as output. The proposed model uses linear
regression to predict scores.

6.2 IPL Win Prediction

1. Logistic Regression: The basic structure of logistic regression is: Get a dataset of N points. Each point i is a
set of m input variables x1,i ... xm,i (also called independent variables, explanatory variables, predictor
variables, characteristics or attributes) and a binary outcome variable Yi (also called dependent variable). ).

Copyright to IJARSCT DOI: 10.48175/568 510

www.ijarsct.co.in
 ISSN (Online) 2581-9429
IJARSCT
International Journal of Advanced Research in Science, Communication and Technology (IJARSCT)

Volume 3, Issue 1, April 2023

Impact Factor: 7.301
Variable, response variable, output variable or class), i.e. can have only two possible values 0 (often 'no' or
'error') or 1 (often 'yes' or 'success') can. The goal of logistic regression is to use a data set to create a predictive
model of the outcome variable.
2. Random Forest: Random Forest (Random Decision Forest) is an ensemble learning method for classification,
regression, and other tasks that works by building different decision trees during training. For classification
tasks, the output of a random forest is the classes chosen by most trees. For regression tasks, the mean or mean
prediction for each tree is returned.

6.3 ODI Match Prediction

1. Decision Tree Classifier:-Decision tree learning is a supervised learning approach used in statistics, data
mining, and machine learning. This format uses a classification or regression decision tree as a predictive
model to draw conclusions about a set of observations.
2. Logistic Regression:-The basic structure of logistic regression is: Get a dataset of N points. Each point i is a
set of m input variables x1,i ... xm,i (also called independent variables, explanatory variables, predictor
variables, characteristics or attributes) and a binary outcome variable Yi (also called dependent variable). ).
Variable, response variable, output variable or class), i.e. can have only two possible values 0 (often 'no' or
'error') or 1 (often 'yes' or 'success') can. The goal of logistic regression is to use a data set to create a predictive
model of the outcome variable.
3. Gaussian NB:-Gaussian Naive Bayes is an extension of Naive Bayes. Other functions are used to estimate the
data distribution, but Gaussian or normal distributions are the easiest to implement, as they require computing
the mean and standard deviation of the training data.
4. Gradient Boosting Classifier:-Gradient boosting is a machine learning technique used in regression and
classification tasks, among others. It produces predictive models in the form of an ensemble of weak predictive
models (usually decision trees). If the decision tree is a weak learner, the resulting algorithm is called a
gradient boosted tree generally better than random forest.

6.4 IPL Score Prediction Algorithm Analysis

Linear regression was found to be more accurate in predicting scores compared to ridge and LASSO regressions.
Table 1. Accuracy of the Score Prediction Models (IPL)
Algorithm MAE MSE RMSE
Linear Regression 12.118699658080088 251.04770095081787 15.844484874896308
Ridge Regression 12.117294527004992 251.03172964112588 15.84398086470461
Lasso Regression 12.213583996827493 262.36538279606964 16.19769683615759
Linear regression was found to be more accurate in predicting scores compared to ridge and LASSO regressions.
y = Co + C1 * x (Linear regression equation)
Where y is the dependent variable, x is the independent variable,
C0 is the bias coefficient, and C1 is the coefficient of x.

6.5 IPL Win Prediction Algorithm Analysis

Logistic regression was found to be more accurate in predicting win probability compared to Random Forest Classifier.
Although, Random Forest Classifier gives more accuracy, but we want the value in the form of percentage. So, we
decided to go for Logistic Regression.
Table 2. Accuracy of IPL Win prediction
Algorithms Accuracy
Logistic Regression 80.31
Random Forest Classifier 99.92

Copyright to IJARSCT DOI: 10.48175/568 511

www.ijarsct.co.in
 ISSN (Online) 2581-9429
IJARSCT
International Journal of Advanced Research in Science, Communication and Technology (IJARSCT)

Volume 3, Issue 1, April 2023

Impact Factor: 7.301
6.6 ODI Win Prediction Algorithm
As the Gradient Boosting Classifier has was found to be more accurate in predicting win probability compared to
Decision Tree Classifier, Logistic Regression, and Gaussian Classifier.
Table 3. Accuracy of ODI Win prediction
Algorithms Accuracy on training data Accuracy on testing data
Decision Tree Classifier 77.10 77.30
Logistic Regression 74.7 74.60
Gaussian Classifier 76.6 76.8
Gradient Boosting Classifier 100 99.9

VII. IMPLEMENTATION OF GUI

A graphical user interface is developed for machine learning models using the Flask Framework. The backend for pages
uses Python. This site can be used to predict his IPL score using the last 5 over data. The model is provided with all the
input information it needs to make a prediction. All calculations are performed in real time.

7.1 IPL Score Prediction: The GUI required at least 5 overs of the data to predict the score as shown in the Fig.
2.Model require the input data of Batting team, Bowling team, Over, Runs, Wickets, Run Scored in last 5 overs,
Wickets fall in last 5 overs to predict the score of the match as shown in the Fig. 2
Score prediction model

Fig.2.Model of Score Prediction(IPL)

Copyright to IJARSCT DOI: 10.48175/568 512

www.ijarsct.co.in
 ISSN (Online) 2581-9429
IJARSCT
International Journal of Advanced Research in Science, Communication and Technology (IJARSCT)

Volume 3, Issue 1, April 2023

Impact Factor: 7.301

Fig. 3. Input to IPL Score prediction model

The output we get from the model is not exactly the predicted output. Therefore, to improve the accuracy of our model,
we add and subtract 3 to get the score range shown in Fig 4. So our model works in most cases

Fig.4.IPL Score prediction result

Copyright to IJARSCT DOI: 10.48175/568 513

www.ijarsct.co.in
 ISSN (Online) 2581-9429
IJARSCT
International Journal of Advanced Research in Science, Communication and Technology (IJARSCT)

Volume 3, Issue 1, April 2023

Impact Factor: 7.301
7.2 IPL Win Prediction
The GUI required team1, team2, city , total runs, wickets, overs finished to predict the Winner of the IPL match as
shown in the Fig 5. Model requires the input data as shown in the Fig 6.

Fig.5.Winning prediction model of IPL

Fig.6.Input to Winning prediction model of IPL

Copyright to IJARSCT DOI: 10.48175/568 514

www.ijarsct.co.in
 ISSN (Online) 2581-9429
IJARSCT
International Journal of Advanced Research in Science, Communication and Technology (IJARSCT)

Volume 3, Issue 1, April 2023

Impact Factor: 7.301
The final output of this model is shown in Fig 7.

Fig. 7.IPL Winning prediction result

7.3 ODI Win Prediction

The GUI required team1, team2, Batting team, Bowling team, Batsman, Bowler, Current runs, Wickets, Overs finished,
Striker runs, Non- striker runs to predict the Winner of the ODI match as shown in the Fig 8. Model requires the input
data as shown in the Fig 9.

Fig. 8.Winning prediction model of ODI

Copyright to IJARSCT DOI: 10.48175/568 515

www.ijarsct.co.in
 ISSN (Online) 2581-9429
IJARSCT
International Journal of Advanced Research in Science, Communication and Technology (IJARSCT)

Volume 3, Issue 1, April 2023

Impact Factor: 7.301

Fig. 9.Input to Winning model of ODI

Fig. 10.ODI Winning prediction result

VIII. CONCLUSION
Predicting winners in sports, especially cricket, is a challenge and very complex. But you can make this much simpler
and easier by incorporating machine learning. This paper gives information regarding IPL score prediction, IPL win
prediction and ODI win prediction. By using all the information, we developed a website by using flask. In this study of
his, various factors were identified that influence the outcome of Indian Premier League matches. Factors that
significantly affect the outcome of an IPL match include the teams competing, the venue, and the city.We get better
prediction by using linear regression for the IPL score prediction than other algorithms. We got 80% accuracy by using
logistic regression for the IPL win prediction. We got 99% accuracy for the ODI win prediction by using Gradient
Boosting Classifier.

Copyright to IJARSCT DOI: 10.48175/568 516

www.ijarsct.co.in
 ISSN (Online) 2581-9429
IJARSCT
International Journal of Advanced Research in Science, Communication and Technology (IJARSCT)

Volume 3, Issue 1, April 2023

Impact Factor: 7.301
REFERENCES
[1]. Monoj Ishi(2022) “WINNER PREDICTION IN ONE DAY INTERNATIONAL CRICKET MATCHES
USING MACHINE LEARNING FRAMEWORK: AN ENSEMBLE APPROACH” ,Indian Journal of
Computer Science and Engineering (IJCSE) Vol. 13 3 May-Jun 2022.
[2]. I. P. Wickramasinghe, "Predicting the performance of batsmen in test cricket," Journal of Human Sport &
Excercise, vol. 9, no. 4, pp. 744-751, May 2014.
[3]. Madan Gopal Jhanwar and Vikram Pudi, Predicting the Outcome of ODI Cricket Matches: A Team
Composition Based Approach, Published in MLSA@PKDD/ECML 2016,Computer Science.
[4]. ]Kevin Desai, Siddhant Doshi, Surekha Dholay ,“Predicting Outcome of ODI Cricket Games”, International
Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181, Volume 3, Issue 01, Special Issue
– 2015.
[5]. Preetham HK , Prajwal R , Prince Kumar, Naveen Kumar, "Cricket Score Prediction Using Machine
Learning", INTERNATIONAL JOURNAL OF INNOVATIVE RESEARCH IN TECHNOLOGY, Volume 9
Issue 8
[6]. Kumash Kapadia, Hussein Abdel-Jaber, Fadi Thabtah, Wael Hadi,"Sport analytics for cricket game results
using machine learning: An experimental study", Emerald Publishing Limited, Published in Applied
Computing and Informatics
[7]. Prasad Thorat, Vighnesh Buddhivant, Yash Sahane, "CRICKET SCORE PREDICTION", International
Journal of Creative Research Thoughts (IJCRT),Volume 9, Issue 5 May 2021

Copyright to IJARSCT DOI: 10.48175/568 517

www.ijarsct.co.in