The 14th Annual MIT Sloan Sports Analytics Conference took place in Boston Massachusetts last weekend. Aimed at those in sports business and analytics, it once again cemented itself as one of the premier forums in sports.
As in previous years, they showcased new ideas and analytics innovation in football, soccer, hockey, and basketball. New industries that received a deeper look included sports media, general sports business (ticketing, audience retention) and sports betting.
Sportlogiq was present with our very own Data Scientist/Machine Learning researcher Michael Horton, presenting on Learning Feature Representations from Football Tracking. Michael’s research paper was on display, and he presented to an engaged audience on Friday.
Michael designed a framework that learns directly from raw tracking data. He implemented networks for predicting passing and tackles. He discussed how trajectory data differs in football and issues that come with that based on position variety. He also explained how the trajectory learning framework addressed these issues. The models’ predictions for pass completion and likely tacklers were very accurate.
Our team attended panels on soccer, hockey, football and betting. In order to highlight key takeaways from the conference and innovations in our markets, we published our developing thoughts on the talks via our Twitter feed. You can catch up on the thread here.
It was a great opportunity to network and talk to those who are pushing limits and shattering boundaries in the sports world. The world of sports analytics is evolving quickly and the advancements are visible from year to year at Sloan.
We’ll be back at the conference again next year and can’t wait to see how the industry has been elevated once again.
Check out our highlights from #SSAC20 below.