Predictive Analytics For FIFA Player Prices: An ML Approach

Abstract

Soccer extends beyond being a widely followed sport; it constitutes a flourishing industry. In the realm of player transfers, team managers face critical decisions regarding player valuation, transfer fees, and market values. Market values represent estimated prices that players can command in the football market, playing a crucial Involvement in transfer negotiations. While football connoisseurs. have traditionally been relied upon for market estimations, their judgments often lack accuracy and transparency. Thankfully, data analytics offer a promising as an alternative or supplementary methodology to expert-based player valuation, this research introduces an Impartial numerical approach assessing market worth of players. Our method employs ML algorithms, specifically utilized with performance-related data extracted from sofifa.com. We conducted empirical investigations employing four regression models: multiple linear regression, random forests, linear regression, and decision trees, with the aim of assessing the market values of players. Additionally, our objective extends to data analysis to identify the pivotal factors impacting market value determination. Our empirical results indicate that the random forest algorithm surpasses other models in predicting market values of players. It attained the highest level of accuracy and exhibited the lowest error ratio in comparison to the baseline. These findings underscore efficacy of proposed methodology, outperforming established approaches documented in previous studies. Furthermore, we posit that our outcomes can exert a significant influence on negotiations among football players agents and clubs. Our model serves as robust foundation, streamlining the negotiation process and delivering an objective, quantitative estimation of a player's market value.