Sports Analytics and Wagering

We have worked on a variety of analytics for wagering syndicates, fantasy sports providers, online Sportsbooks, broadcast networks, and sports teams. We have also worked with many private individuals who have independently creating wagering models or fantasy sports tools, and with various start-up businesses in these industries.

Analytics for Professional Sports Teams

We have consulted with baseball team GM’s on sabermetric modeling for arbitration cases, analysis of over- and under-valued players, and win probability modeling. In addition to our expertise in baseball, we also have experience with statistical modeling for NBA basketball, NCAA football, and for MMA fighting, where we operationalized the only accurate ranking system in the sport.

Fantasy Sports Analytics and Modeling

Fantasy sports providers require constantly improving and complex analytics to draw players to their games, as the landscape for fantasy sports is getting more and more competitive. Remaining in front of the competition in terms of analytical complexity is key to attracting the type of knoweldgeable players who provide the greatest revenue. Given the pace at which methods of player evaluation are evolving, it is literally a continuous process to remain at the current edge. We have created player selection tools and tournament bracket optimizers for several established large players in the industry, as well as with owners of startup firms catering to fantasy sports players.

Recently, we developed a game theoretic fantasy football drafting tool for one of our clients. Our statisticians, consulting with a team of fantasy football enthusiasts, created a unique modeling strategy to optimize draft selection strategy. Our model improves upon existing methods of drafting in two ways. First, it calculates a scaled three-factor risk measure for each player, to indicate the statistical confidence surrounding his projection for the year. Second, our application uses artificial intelligence to simulate likely opponent strategies. The most sophisticated team owner knows that selecting the best team is not simply the accumulation of the best players available, but involves forecasting and reacting to opponent owners’ drafting strategies. Precision Consulting developed an artificial intelligence algorithm that recommends choosing the player that leads to the strongest team relative to the other owners.

Sports Wagering Analytics and Modeling

Our team develops wagering models for many types of clients. Much of our work has regarded direct methods of wagering such as for sports wagering syndicates, large individual sports bettors, and horse racing syndicates. However, we have also worked with large media outlets such as ESPN on tournament selection strategies, such as to generate an optimal March Madness bracket. We also can help Sportsbooks distinguish between +EV ‘sharp’ players who are unprofitable to the casino in the long-run, and ‘squares’ who have been on a lucky streak, and whom the Sportsbook should welcome gladly.

Among our wagering projects, we constructed a baseball model for the largest sports wagering syndicate in the world. The player and game projections apply a WPA-based model as well as a standard model utilizing well-known sabermetric measures such as Baseruns and UZR.

We also developed a profitable tennis wagering model using Monte Carlo simulations. Using historical data from tens of thousands of professional matches, the innovative model assigned a level of strength to each player based upon various factors including his strength of competition, performance on different court types, average margin of victory, and also his trend in performance in recent matches. Then, based upon a player’s and his/her opponent’s strength levels, the model predicted the probability a player would win the match.

Proposition betting has become extremely popular, with bettors enjoying wagers such as being able to bet on how many points a basketball player scores in a given night. While allowing these types of bets can be very profitable for Sportsbooks, they are oftentimes hard to set odds for. We have worked on setting the optimal line for proposition bets of hundreds of types, for virtually any situation.

Case Studies

Game Theoretic Fantasy Football Drafting Tool

We developed a Fantasy Football drafting tool that determines the optimal player selection at each decision point in a draft. The tool is based on a two-pronged approach that combines (a) proprietary mathematical models for point projections, and (b) qualitative adjustments based on content analysis to account for factors such as players changing teams or the projected performance of rookies. Using the inputs of our projections, the drafting engine uses an artificial intelligence algorithm to simulate the draft given the available information about the drafting styles and propensities of other players in the pool. The tool then borrows concepts and formulas from Modern Portfolio Theory to compute a drafting strategy that maximizes risk-adjusted points earned, making the drafting engine absolutely unique. To read the full report on our model, please click here.

ESPN Bracket Optimizer

ESPN

Our Bracket Optimizer uses a game theoretic algorithm to assist players in building the ideal March Madness bracket for their particular pool. To do so, we account not only for the probabilities that any given team will beat its opponent, but also for common behavior patterns of other March Madness fans in the same pool (such as geographic bias, or team popularity as evidenced in pubic pools). Our Bracket Optimizer is used as part of the ESPN Tournament Challenge. Read about the Challenge here.

Please click here to read our Data Security Policy.