Sports prediction has experienced a sharp increase in popularity over the last two decades with the implementation of machine learning (ML) methods. The objective of this paper is to examine the research trends in Sports prediction using machine learning approaches and then to identify effective ones amongst those for assessing the accurate sports predictions especially cricket with key characteristics of such successful applications. In this paper, five machine learning algorithms such as: Support Vector Machine (SVM), Logistic Regression (LR), Multinomial Naïve Bayes (MNB), Random Forest (RF) and Extreme Gradient Boosting (XG Boost) are evaluated for effective sports prediction of Cricket after data pre-processing is done through grid search, mean encoding based feature engineering. The experiments uses team sports data such as cricket which is collected from GitHub and the results shows whether the difficulty level of predicting outcomes varies across sports considering team data only or in combination with player. The experimental research obtained are interesting which may help the sports enthusiasts and analysts in making informed decisions while predicting outcomes of team sports.