This is a review of the NCAA research conducted by Mark Bashuk applying Win Probability Matrix and Cumulative Win Probability.
Basketball games are often judged on how they end. If our team scores a last second point to win, we feel it was an exciting game. If we were cheering for the other team, we feel the game was depressing. However, the final score does not tell the complete story. In order to combat this, a method is developed that uses cumulative win probabilities to quantify each game, combining a team’s average cumulative win probability with strength of schedule to rank teams and predict future game performance.
Rankings are calculated using a SQL stored procedure. The first step is creating a table, which combines play-by-play data and a Win Probability Matrix. The second step then combines the Cumulative Win Probability of each game with each team’s strength of schedule, which produces the team ranking statistic.
In order to determine the optimal value for each of the three variables used in the calculations, three simulations are run. The first simulation tests 100 combinations of each team’s Cumulative Win Probability and its Strength of Schedule. The second simulation looks at the idea of home court advantage by giving the home team an increased percentage and removing that percentage from the visiting team. The third simulation is designed to determine if the performance of a team at the end of a game is more predictive than their performance at the beginning of the game. Games are split into eights segments of three values. The results are charted in a histogram, which clearly indicate that the model is more accurate as later game segments are weighted more heavily than earlier segments.
In order to predict the margin of outcome in a game the rankings of the home team, rankings of the visiting team, and home court advantage are incorporated into the equation.
Analysts, when looking to make predictions regarding upcoming games or predicting which teams will make it to the playoffs, can use this information. This statistic judges team and player performance more accurately as it takes into account more than just the final score. This provides coaches and analysts with an improved ability to judge the ability of their players and the contribution they are making to the team. This will be useful when the time comes to look at making trades. Coaches can also help their players become better all around athletes by looking at the skill sets of the top ranked players. As Strength of Schedule is incorporated, coaches and analysts can determine what effect this has on their team and their winning probability. Leagues can incorporate this information when designing the schedule for a new season, working to minimize the impact strength of schedule has on a team’s performance.
While fans can use emotions to rank a game, analysts and coaches need to put the emotion aside and look at quantifiable aspects of the game in their determination of the ranking.
Analytics Used: SQL Stored Procedure, Win Probability Matrix, Cumulative Win Probability, Strength of Schedule, Simulations
————————
Find out how Sports Analytics Expert Victor Holman can give your team the competitive advantage.
How mature is your team’s analytics program? Take the Sports Analytics Maturity Assessment.
Discover the Groundbreaking Sports Analytics Application and Framework coaches and sports analysts are talking about!
Learn all about sports analytics in Victor Holman’s Sports Analytics Blog.