An increase in aggressive play and misconduct across European football leagues has led to excessive delays and interruptions owing to the frequent use of Video Assistant Referee (VAR) by on-field referees. This investigation aims to highlight the tactical and misconduct-related offences that evoke VAR interference using data mining and machine learning based SVM algorithm. The data set of referee activities concerning cautions and dismissals of players containing 6232 matches from five consecutive seasons in the five top European leagues was used in this study. The SVM model demonstrated an accuracy of 98% considering most essential indicators. The findings indicate that certain actions, such as misconduct, attack wrecking, dangerous play, ground challenges, and challenges off the ball, play a significant role in determining the likelihood of a referee consulting the VAR during a match. Referee training should focus on players’ misconduct and offensive tactics during gameplay.