Arturo and I just finished attending the MIT Sloan Analytics Conference. The good news is I have many many thoughts. I know as time continues I’ll forget a few. Here are some quick bullet points.
Brian Burke (NFL) was a rockstar
My favorite panel by far was Brian Burke from Advanced NFL Stats, who was kind enough to be on the Wages of Wins podcast a few weeks back, attended a panel where he and several NFL coaches analyzed plays called “”Monday Morning Quarterback: Coaching & In-Game Decisions” This is key. There is so much of the random “quants vs. front office” comments, that getting them into a room together to actually talk about strategy was key. If ESPN, who had a huge presence at Sloan, doesn’t work on a Herm Edwards with Brian Burke feature at some point, then I will be sad.
I didn’t get a chance to see the NHL Brian Burke talk, but based on some of the quotes I’ve seen, Advanced NFL Stats was by far the better Burke at Sloan.
Communicating Better, hmmm.
I attended a few panels on this. One of the key problems we have is communicating our ideas effectively. I agree (it’s also one of the reasons the Burke panel was awesome) That said, there’s an elephant in the room here. If you don’t learn stats, then you’re very vulnerable to just listening to stuff you like. One subject not coming up is how to get better educated on stats if you’re not a quant. More and more teams are hiring stats people. All of them are being told to get better at communicating, to be more accommodating. And this means teams are increasing their risk of hiring people that just tell them what they want to hear. It’s becoming more and more important for the people actually running teams to get versed in stats. It’s a subject I don’t feel is being brought up enough.
Optical Tracking Data
A quick reminder. SportsVU is a company that is installing cameras in NBA stadiums. This lets us track every action on the court. It’s a tremendous amount of data, with some huge possibilities for insight.
Kirk Goldsberry was the star of the basketball papers with this. And the Eagle tool that was presented by Rajiv Maheswaran in the Data Morphing: Dynamic Interaction with Basketball Big Data talk was a huge hit the whole conference. (I saw R.C. Buford in talks, standing right near the front.)
I feel this data is a mixed bag. You see, teams have had lots of data for years. Getting more data doesn’t mean they’ll use it correctly. However, people like Kirk and Rajiv are showing very impressive applications of the data. If used correctly, I believe this will advance the understanding of the sport. Now, I want to actually harken to something Ari Caroline said last week. Knowing something may be good, but if it’s not actionable, then it may not be that useful. How teams will use this data and the amazing new tools remains to be seen.
Stan Van Gundy was awesome
Oh, I will be writing more about this. “How does he not have a job?” is what Arturo kept asking during the Basketball Analytics panel of Sloan. He managed to bash lineup data, playing smallball, coaches playing systems over playing to win, as well as explaining how the NBA is not a video game. He was definitely the star of the panel. He kept saying he wasn’t really an analytics guy, and would procede to then say an analytics statement. I was super lucky to get a chance to say hi and snap a picture with the Wages of Wins shirt. I shouted he needed to come to Denver as he walked off.
Jeremy Britton works at a great company called Zazzle. He helped design the now famous “Average NBA Team” shirt. Don’t worry, we’ll work on getting it on the site if you’re interested in buying one. We’ll definitely make this a regular thing for Sloan as the reception to it was great.
I’ll try and make sure to get more stuff out. I have a few people that agreed to be on the podcast (no GMs or coaches but definitely some good stat folk) Sincere apologies to anything awesome I forgot to mention. I was going right from when I touched down in Boston on Thursday to 11pm Saturday night, so I’m bound to have forgotten something. Stay tuned for more.