This is a review of the probabilistic physics-based model based on player tracking data research conducted by William Spearman.
Soccer is more a game of strategy than scoring as there are relatively few goals made in a soccer game. What are the other players on the field doing when they do not have the ball and what value do they contribute?
The development of tracking data allows us to look into these questions. A probabilistic physics-based model is built to determine the probability that a player who does not currently have control of the ball will score a goal. The data comes from a 14-team professional soccer league during the 2017-2018 season which included 58 matches. Events were determined as any on-ball actions that occurred during a match. Each event was labeled with the time the event occurred, the player in possession of the ball, and the type of event (pass, shot, goal, etc.). The position of all players on the field were included in the data.
The purpose of the probabilistic physics-based model is to determine the probability that the team in possession of the ball will score with their next on-ball event. In order to do this the probability that the team successfully passes to each point on the field and scores is determined. Three individual probabilities are taken into account – the probability that the ball is passed to one point on the field, the probability that the ball will be controlled at that point by the passing team, and finally the probability of scoring a point from that point on the field. These three probabilities are then combined to represent the probability that a goal will be scored with the next on-ball event at one specific spot on the field.
There are several possible applications of this data for analysts and coaches. Analysts will be able to quickly determine which were the key moments in the game – which moves ultimately lead to the score or offensive chance.
Are players passing to the points on the field where their teammates have the best chance of scoring? If not, then strategies need to be reviewed in order to maximize scoring chances. Coaches can look at how their opponents use the space on the field in order to develop the best possible defensive strategy. Coaches can examine the data to determine player and team preferences. This can be used to strengthen their own team by adapting offensive strategies to maximize these preferences. Defensive strategies can also be developed to minimize the chances of the opponent scoring.
After the data was analyzed for all matches it was found that teams who score more goals also generated more opportunities to score. Coaches and players need to be aware of where the best opportunities to score exist in order to maximize their chances of scoring.
When teams are looking to make trades they can analyze the data of the players they are looking at to determine which one will be the best fit for how the team uses space on the field.
How mature is your team’s analytics program? Take the Sports Analytics Maturity Assessment.
Learn about the Groundbreaking Sports Analytics Model coaches and sports analysts are talking about!
Learn all about Sports Analytics here.