Selecting players’ is one of the most crucial tasks for a team in sports like cricket. A player’s performance depends on several factors like playing conditions, venue of matches, climate, etc. The most difficult situation that comes during the match is to find the substitute when a player is injured or ill on the field. The important factor that needs to be considered in this situation is the previous performances of the player with an opposition team in all the matches. To identify the best substitute, we have used the K-nearest neighbor classifier and random forest classifier model. The study concludes that K-nearest neighbor classifier model suits best for decision-making and performance analysis of the players with an accuracy of 94%.