e-ISSN: 2582-5208

International Research Journal of Modernization in Engineering Technology and Science

Volume:03/Issue:06/June-2021 Impact Factor- 5.354 www.irjmets.com

CRICKET ANALYSIS USING MACHINE LEARNING

B V S Sai Praneeth*1, V Srighan Reddy*2, P Jayanth*3, K Jeevan Reddy*4
*1,2,3Student, Department of Electronics and Communication Engineering, SreeNidhi Institute of Science and
Technology, Hyderabad, Telangana, India.
*4Associate Professor, Department of Electronics and Communication Engineering, SreeNidhi Institute of
Science and Technology, Hyderabad, Telangana, India.
ABSTRACT
Indian Premier League is a franchise cricket based, annual tournament which is organized by Board of Cricket
for Control in India (BCCI) every year. IPL is one of the highly rated cricket tournaments around the world and
several international cricket players go through the auction process where they are sold to franchise owners at
very high prices. So the amount of money spent by each franchise team builds a huge pressure on the owners to
search for victories, which depends on the performance of the team. In general, a franchise signs in players
based on their previous year’s performance. The result of the match is predicted by analyzing their previous
encounters, the venues they played under etc. To predict the outcome of the match, we have used Machine
Learning where the current problem statement comes under Classification type of problem. Different Machine
Learning algorithms like Random Forest and Support Vector Machine (SVM) algorithms are used in the design
of the model. We have used K-Fold cross validation process to use the entire data-set both for training and
testing it before the user inputs are given. The classifier which has highest accuracy is then used to predict the
end result of the cricket match.
Keywords: Machine Learning, Classification, Random Forest Algorithm, Support Vector Machine (SVM)
Algorithm, K-Fold cross validation.
I. INTRODUCTION
CRICKET is one of most sought after games in the world. It is the second most played game in the world after
football. The original form of the game was first invented in Saxon or Norman times by the children who might
be playing in the dense woodlands of Weald and clearings of South-East England back in 1611. Later in the
middle of sixteenth century the game had developed enough to be known as village cricket. The first cricket
match including country names was played in 1709. The game has only been played by more and more countries
ever since.
The game has implemented several modifications since its inception. Several rules have been added and some
rules have also been removed. In the millennial era, it is one of the only games with a large viewership and
generates large revenue in India. The first ever first-class match was played in India in 1864 between Madras
and Calcutta but not many records exist of that period. It’s only after the independence that it had started
developing revenue for the economy. The first-class cricket was the longer format of the game with no limit on
the number of overs (six consecutive legal deliveries) and was played for a minimum of 3 days to a maximum of
5 days. It included 4 innings’. Later, the games was shortened to a limited over format of 50 overs per side. Later,
in 2003 the 20 over format of the game was invented. All the international teams started playing this format in
2005.
After the famous Indian victory in the first 20-20 world cup tournament, this format has only gained more
popularity. Hence, in 2008 a state level league including eight to ten teams was founded by Board of Control for
Cricket in India. It was called Indian Premier League also known as the IPL. Initially not many people were
interested but as years passed it has only gained more and more popularity and generated a large amount of
revenue through sponsorship and viewership. BCCI was eyeing at at least Rs.4000 crore as sponsorship revenue
in the latest 2021 edition of the tournament before it got postponed indefinitely because of the ongoing
pandemic. If the tournament is canceled there is an estimated loss of up to Rs.2500 crore.
In a tournament where every match contributes to large amount of revenue, it is very important to predict the
outcome of a match based on the facts from the past. This information can include the history of each individual
teams, performance in the past tournaments, performance in the present tournaments, matches won by each
teams in head to head competitions of the past, the venue, team captains, highest scorer etc. Now, in the 21 st
century, with the innumerable advancements in technology, collection of data has become comparatively easier.
www.irjmets.com @International Research Journal of Modernization in Engineering, Technology and Science
[981]
 e-ISSN: 2582-5208
International Research Journal of Modernization in Engineering Technology and Science
Volume:03/Issue:06/June-2021 Impact Factor- 5.354 www.irjmets.com
Now, with the data being available very easily it has become an important task to develop models that predict
the outcome of the match.
Hence, we need to develop a model that predicts the outcome of a match played among the two teams with the
help of some standard Machine learning algorithms along with some necessary modifications so that the models
can make use of all the factors available to obtain a prediction with a good accuracy. The model should be able to
predict the outcome dynamically, and not just for certain types of data. That is a very important factor for any
model.
So, now after understanding the importance of the league, some basic understanding of mathematics, statistics,
and some standard machine learning tools we will begin developing a program to predict the outcome of the
match.
II. METHODOLOGY
We have followed general Machine Learning Cycle to build our project which includes the following steps:
1) Data Collection: We had collected the dataset from www.kaggle.com
2) Hypothesis Generation: Here entire dataset and its features ae analyzed by studying various publications
and came to know about he features that will have a impact in predicting the result of the match.
3) Data-Preprocessing: The dataset we had consists of missing values, outliers and duplicate records. So, in
order to build a model , we had to first correct these values in the necessary features to improve accuracy and
checked these features using univariate and bivariate analysis.
4) Data-Encoding: As our project is about a Classification Problem, to analyse some categorical variables we
have done encoding of those feature values for better results.
5) Model Selection and Evaluation: We had built our model using Random forest and Support Vector machine
(SVM) and by using K-Fold cross validation technique entire data-set is used for both training and testing of the
model.
a) Random Forest:
Random Forest is basically a tool that builds as many decision trees as possible to the dataset and makes a more
accurate and stable prediction. The number of hyperparameters in a Random Forest is similar to that of simple
decision tree hence making it very simple to implement and also the Random Forest contributes to more
randomness to the model making it more dynamic.
b) Support Vector Machine:
Support Vector Machine (SVM) algorithm, is used for both the Classification and Regression tasks. However, it
is primarily used for Classification tasks of Machine Learning. SVM algorithm creates the best decision
boundary (also known as Hyperplane) which splits the N-dimensional space into distinct classes such that we
can easily enter the new data point into the correct category in future.
6) Real Time Testing: Once the model is ready, inputs given by the user are taken and analyzed to predict the
winner of the match.
III. MODELING AND ANALYSIS
We have used Jupyter Notebook and Anaconda Navigator as tools for building the project. After importing
necessary libraries like Scikit learn, Matplotlib, Numpy, Pandas into the notebook and the required dataset after
Data preprocessing and Encoding, we have created a function called classification( ) in which we have
described our model. The architecture of the project is shown as below:

www.irjmets.com @International Research Journal of Modernization in Engineering, Technology and Science

[982]
 e-ISSN: 2582-5208
International Research Journal of Modernization in Engineering Technology and Science
Volume:03/Issue:06/June-2021 Impact Factor- 5.354 www.irjmets.com

Figure 1: System Architecture

Now that our model is designed, we retrieve the desired results in the necessary form. As we are just predicting
the winner of a match played between two teams, we must ascertain the accuracy at which out model predicts
the output.

Figure 2: Testing Using Support Vector Machine (SVM) Algorithm

Figure 3: Testing Using Random Forest Classifier

www.irjmets.com @International Research Journal of Modernization in Engineering, Technology and Science

[983]
 e-ISSN: 2582-5208
International Research Journal of Modernization in Engineering Technology and Science
Volume:03/Issue:06/June-2021 Impact Factor- 5.354 www.irjmets.com
IV. RESULTS AND DISCUSSION
As Random Forest Algorithm has an accuracy which is far greater than that achieved by Support Vector
Machine (SVM) algorithm, we have used Random Forest in the final model for predicting the winner of the
match. Then the inputs from the user are taken after the toss in predicting the outcome of the match.

Figure 4: Prediction of Winner of the match by the user inputs

So, we can simply say that most of these machine learning problems are optimization problems in which we
minimize a target label into different numerical constraints.
V. CONCLUSION
The tools used for Machine Learning are very powerful and some strong predictions can be made through
efficient use of these tools. Understanding Machine Learning concepts will be very important as the amounts of
data generated by the modern advancements in technology will hugely help in the making decisions that might
effect the future in a positive way. These tools will also play a very important role in making important decisions
in business organizations in which the impact will directly effect the revenue generated by the firm. Random
Forest is one of the accurate machine learning classifier since it takes multiple decision trees in account and
builds an ensemble model of them. As we got an accuracy of 87.037% with Random Forest which is way more
than Support Vector Machine Algorithm, we have used it to build the model using Random Forest in order to
predict the result.
ACKNOWLEDGEMENTS
This project is accomplished based on the knowledge gained during pandemic infused lockdown by the authors
under the guidance of a Mr. K. Jeevan Reddy who is an Associate Professor at Department of Electronics and
Communication Engineering in SreeNidhi Institute of Science and Technology, Hyderabad, Telangana, India.
VI. REFERENCES
[1] Abhishek Naiket. Al, “Winning Prediction Analysis in One-Day-International (ODI) Cricket Using
Machine Learning Techniques”, IJETCS, vol. 3, issue 2, ISSN:2455-9954, April 2018.
[2] Akhil Nimmagadda et. Al, “Cricket score and winning prediction using data mining”, IJARnD Vol.3,
Issue3.
[3] Bunker, Rory & Thabtah, Fadi. (2017) “A Machine Learning Framework for Sport Result Prediction.
Applied Computing and Informatics”, 15. 10.1016/j.aci.2017.09.005.
[4] Esha Goel and Er. Abhilasha, “Random Forest: A Review”, IJARCSSE, Volume 7, Issue 1, DOI:
10.23956/ijarcsse/V7I1/01113, 2017.
[5] I. P. Wickramasinghe et. al, "Predicting the performance of batsmen in test cricket," Journal of Human
Spo
[6] Munir, F., Hasan, M.K., Ahmed, S., Md Quraish, S., 2015. Predicting a T20 cricket match result while the
match is in progress (Doctoral dissertation, BRAC University).rt & Exercise”, vol. 9, no. 4, pp.744-751,
May 2014.
[7] N. Pathak, H. Wadhwa, Applications of modern classification techniques to predict the outcome of ODI
cricket, Procedia Comput. Sci. 87 (2016) 55–60.
[8] Passi, Kalpdrum & Pandey, Niravkumar. (2018) “Predicting Players' Performance in One Day
International Cricket Matches Using Machine Learning” 111-126. 10.5121/csit.2018.80310.
[9] R. P. Schumaker, O. K. Solieman and H. Chen, "Predictive Modeling for Sports and Gaming” in Sports
Data Mining, vol. 26, Boston, Massachusetts: Springer, 2010.

www.irjmets.com @International Research Journal of Modernization in Engineering, Technology and Science

[984]
 e-ISSN: 2582-5208
International Research Journal of Modernization in Engineering Technology and Science
Volume:03/Issue:06/June-2021 Impact Factor- 5.354 www.irjmets.com
[10] Rabindra Lamsal and AyeshaChoudhary, “Predicting Outcome of Indian Premier League (IPL) Matches
Using Machine Learning”
[11] Rameshwari Lokhande and P.M.Chawan, “Live Cricket Score and Winning Prediction”, International
Journal of Trend in Research and Development, Volume 5(1), ISSN: 2394-9333
[12] Ujwal U J et. At, “Predictive Analysis of Sports Data using Google Prediction API” International Journal
of Applied Engineering Research”, ISSN 0973-4562 Volume 13, Number 5 (2018) pp. 2814-2816.

www.irjmets.com @International Research Journal of Modernization in Engineering, Technology and Science

[985]