Squad selection for cricket team using machine learning algorithms
MetadataShow full item record
There are mainly three renowned formats of cricket – ODI (One Day International), Test Match and T20 (Twenty Twenty). Selecting a 15 men perfect squad for a particular cricket match in a particular cricket tournament isn’t an easy task. Coach and captain play a vital role to select the most perfect players for a match to win by keeping many parameters in their head such as – analyzing the past scored runs comparing with the balls faced, strike rate, total 50’s , total 100’s, whether the player is right handed or left handed, against which team she/ he has scored well enough (for batsman). Analyzing the past conceded runs comparing with the overs she/ he had taken, economy rate, strike rate, wickets, against which team she/ he has performed well, pitch condition, venue etc. In this paper, we have taken maximum number of parameters in consideration for selecting 1 captain (for captaincy issue), 6 batsmen (top-order, middle-order and finisher), 2 all-rounders, 1 wicket keeper and 5 bowlers (fast bowlers, spinners). Our model can be extended for the team selection in other formats of cricket too. We have used k-means clustering, Linear Regression, Naive Bayes and Page Rank algorithms for selecting batsmen and all-rounders. Support Vector Machine, Naive Bayes, Linear Regression, Decision Tree and RankSVM have been used for selecting bowlers. We have used Bar Graph to show the statistics of different parameters for both captain and wicket-keeper. Our motive is to recommend 15 men squad for a cricket team to the selectors.