Predictive Analysis of Chess Player Performance: An Analysis of Factors Influencing Competitive Success Using Machine Learning Techniques

The world of competitive chess has long been a captivating arena for intellectual competition, where human intelligence, strategic thinking, and long-term planning converge. This study delves into the intricate web of factors that influence a chess player’s competitive success through the lens of predictive modeling and machine learning techniques. Some of the factors influencing a chess player’s performance include experience, relative skill level, current form, opening preferences, playing style, and even psychological aspects like confidence and concentration. One of the objectives of this research paper is to represent many of these factors as features in a dataset and use them as input to a prediction model. Data collection and cleaning are conducted by the authors and are thoroughly elucidated in this paper. Gathered games were played by players of various skill levels, spanning from beginners to seasoned masters, and predominantly feature shorter time controls such as bullet, blitz, and rapid which are more popular in online chess. The research shows that while simple classification techniques are unable to use additional features for better predictions, ensemble techniques are. Our most successful model was constructed using a Random Forest classifier in conjunction with an award streak form evaluation function. This model demonstrated a peak prediction accuracy of 68.67% and maintained an average accuracy of ~60.4%, surpassing the performance of previously employed models in this area of research. The significance of different factors affecting game outcomes is highlighted, confirming that the relative difference in playing strength holds the most substantial influence. Additionally, we show that using a larger number of recent games can help better evaluate the current form and confidence of the player, leading to more accurate predictions.