A couple of nights ago I gave a preview of the SSAC Research Paper competition. I’ve meant to give a preview of the Research Poster competition in previous years but didn’t follow through. I’m following through this year.
The Research Posters presented at SSAC are works that advanced to the second round of the Research Paper competition but did not advance to the final eight presentations. As such, every poster has a full paper associated with it and you can read those papers at the Research Paper competition link. The posters aren’t being judged like, say, a Science Fair competition (although that would be interesting), but attendees can vote for the best poster which will receive a USD 500 prize.
There are fifteen posters that will be presented and they break down accordingly: soccer (4), ice hockey (3), basketball (2), baseball (1), American football (1), volleyball (1), gymnastics (1), golf (1), auto racing (1), and e-sports (1). I’ve been impressed by the quality of soccer analytics research on display at SSAC, but I’m particularly impressed by the posters if the full papers are any indication. Here’s a description of those papers:
- “Physics-based modeling of pass probabilities in soccer” – This was performed by William Spearman and his data science group at Hudl (and also presented at the most recent OptaPro Analytics Forum). It’s a really clever use of physics-based modeling of ball tracking data to develop a measure of passing quality and probability in soccer matches.
- “Flow motifs in soccer: What can passing behavior tell us? – An application of network analysis and expected goals to assess the quality of attacking styles in football.
- “The market for English Premier League odds” – By a group of economists at the University of Chicago business school, a model of the evolution of in-match odds of Premier League full-time scores using a combination of Skellam and Black-Scholes-Merton models.
- “A PageRank model for player performance assessment in basketball, soccer and hockey” – Application of the PageRank algorithm to create a player performance metric based on match flow in field invasion sports
There are some other posters that I am interested in checking out, from the application of machine learning to RFID data to identify NFL receiver routes, to evaluating defensive ability in hockey using crowdsourced passing data, to ranking professional golfers using network analysis and gradient boosting.
It looks like a poster exhibition of very high quality, and I encourage attendees to check them out on Friday and Saturday and vote on the SSAC App.