Pre-match analysis often leans heavily on scouting reports tied to specific teams, which can leave analysts scrambling when facing less familiar opponents. Sim2Win changes the game by stripping away team identities and zeroing in on tactical behaviors extracted from StatsBomb event data. Using data from 178 teams across eleven competitions, it builds rolling tactical profiles based on recent matches, capturing key feature ratios that reveal playing styles.
By clustering these tactical features, Sim2Win creates interpretable profiles that transcend club or country labels, making it easier to identify patterns and anticipate strategies. This approach shifts the focus from just predicting outcomes to supporting tactical decision-making with actionable insights.
For football analysts, this means a powerful tool to decode opponents through data-driven lenses, even when little is known about them. While the model’s effectiveness depends on the quality of event data and evolving tactical trends, Sim2Win highlights how thoughtful feature engineering and clustering can turn raw event data into meaningful pre-match guidance.