2023 SABR Analytics: Watch highlights from Artificial Intelligence and Machine Learning Panel
At the SABR Analytics Conference on Saturday, March 11, 2023, in Phoenix, Arizona, a panel discussion was held on the Future Impacts of Artificial Intelligence and Machine Learning on Baseball
Panelists included Allison R. Levin of Webster University; David Loving of Sage; data scientist Stuart Neilson; and moderator: Brian Hall of Tidal Quantitative Research and New York University.
Here are some highlights from the panel:
On how artificial intelligence is used in different sports
- Hall: “AI is transforming these different sports in a way that you’re forced to learn what the AI thinks if you want to be a champion in these areas. When we talk about artificial intelligence, this is something that will dramatically impact every sport. … Depending on the sport, you learn so many different things from the AI.”
- Loving: “(A) global explanation explains something that can be applied consistently across all teams. (A) local explanation asks questions about specific teams and specific games. ‘Why does this team win this specific game? Why does this lineup end up winning versus this other lineup?’… The local explanation is where we can move from aggregate statistics like OBP to where AI can really get involved and start to teach us things that we would otherwise have a hard time understanding on our own. … AI is really useful for things like lineup optimization and player evaluations.”
On how sports teams are using artificial intelligence
- Neilson: “The prediction engine that we’re using here generates predictive features based on every single one of these players to feed into the game prediction. For example, we have a big boost coming from Jacob deGrom and his pitching relative to his opponent that day. It’s boosting the Mets’ winning percentage by 50 percentage points. … It’s coming from many years’ worth of major league games. It’s all public data that’s been used here. … The AI in this example sides with Jacob deGrom, [but] It may not always side with the pitcher.”
- Hall: “You’re going to see teams use [AI] before games to strategize. They’re going to try to maximize their wins by analyzing games ahead of time and using it accordingly. … I would think that teams would want to use the AI before spending $400 million on a player.”
On how to get involved with the AI/Machine Learning Research Committee at SABR
- Levin: “Research committees are a great way to get involved at SABR to build your resume, to find connections, networking connections. … We are looking for people who want to help by collecting and analyzing data, and learning how this AI machine process works on the real level.”
On what exactly AI does and why it’s so important
- Loving: “[AI] will run thousands or millions of permutations through [it] and it will start to identify ‘What are the aspects that consistently have an impact on that win probability?’ These are the kinds of things the AI tells us to focus on.”
- Hall: “The problem in sports is, and in baseball as well, there’s been an exponential growth in data. We’ve come a long way since Henry Chadwick and there’s a lot of different new information being collected every year in every sport. So the problem is, if we are very easy to make mistakes with very simple decisions, how do we not make mistakes in very complex decisions?”
Transcription assistance from Anthony Remedios.
For more coverage of the 2023 SABR Analytics Conference, visit SABR.org/analytics.
Originally published: May 1, 2023. Last Updated: April 14, 2023.