Available Technology

Dynamic Assessment of Player Performance

Traditional statistical methods to analyze sports data, such as season cumulative averages, typically do not account for the volatility of an individual player and of team play over time. This variability contains invaluable information about the instantaneous skill and performance level of players and teams not captured in traditional methods of analysis. The inventors demonstrate a method for creating new descriptive statistics to describe real-time performance of sports players for use by team managers, fantasy players, sports commentators, and various agents that rely on instantaneous indices of performance for decision-making.
Patent Abstract: 
Dynamic Analysis of Player Performance (DAPP) is a non-patented, web-based, real-time software that serves as a sports performance analytics engine and provides an instantaneous assessment of a sports team. DAPP tracks sports performance metrics in real-time through use of novel multinomial, binomial, and Poisson filtering algorithms. These filtering procedures rely on the Chapman-Kolmogorov-Bayes' Rule system of equations to produce instantaneous estimates of team performance based on all available data at a given time. DAPP also includes new descriptive statistics based on this dynamic Bayesian state-space paradigm. DAPP uses well-established, open-source technology standards to power the analytics engine. In particular, the implementation of DAPP for baseball features a database engine powered by PostgreSGL, a Python-based back-end, and an interactive visualization front-end powered by d3.js.
Easily integrated with existing web- and mobile based sports statistics products - Application to any sport that reports performance as time-series e.g. baseball, basketball, American football, and cricket
Emery Brown
Lab Representatives
Share to Facebook Share to Twitter Share to Google Plus Share to Linkedin