This event has ended. Create your own event → Check it out
This event has ended. Create your own
View analytic
Wednesday, April 26 • 11:15am - 11:35am

Sign up or log in to save this to your schedule and see who's attending!

For years, statisticians, experts, and sportsbooks have aimed to create a model that can accurately predict the outcome of a sports game. These models are kept private or sold at high cost to interested parties. In making these generalized models, outside factors are often disregarded as they may not carry a numerical value. Ignoring these non-quantitative circumstances that impact player and team performances leads to prediction errors. In this work, we create a mathematical model that uses historical statistics and a weighted value system to predict the outcome of NBA games. The value system sets a numerical value on these non-quantitative, discounted factors by testing to determine what situations consistently impact the outcome of a game and by what margin. We use the software R to scrape data directly from web pages that publish the individual player and team statistics relevant to the model. We select the significant variables by calculating the p-values using stepwise regression. We create a predictive model using this data. The model uses future games to test and evaluate its accuracy. Mathematicians test the accuracy of the mathematical model to ensure that it is sound. The results are available on a free, open source website in a sortable, table containing the currents day’s slate of games. It is open for collaboration, to create a more accurate model to any interested party.


Wednesday April 26, 2017 11:15am - 11:35am
125 Rhoades-Robinson Hall

Attendees (1)