Sloan Sports Analytics Conference Day 2 Takeaways

March 13th, 2016 | by Ben Gaines
CdW5X2LWoAE91uD.jpg-large

Photo by Ben Gaines

One of the big themes of the 2016 Sloan Sports Analytics Conference (SSAC) was rest/sleep for athletes. In the NBA, back-to-backs are legendary for draining players of energy and contributing to injury in some cases. Most SSAC attendees are not athletes, and thankfully I don’t think any of us got injured, but this felt like a back-to-back nonetheless; day one was thrilling, and day two started great, but eventually gave way to fatigue despite the tremendous content. In my experience, day two of SSAC is always like that. You love being there and wouldn’t want to be anywhere else, but it’s a grind nonetheless.

The second day of SSAC 2016 had tons of great NBA content (plus a bit of weird NBA content), and that is where I spent most of my time. Here are some of the most interesting tidbits and quotes from day two.

  • Day two was all about Brian Scalabrine. The White Mamba was part of two panels and was in full effect both times. In the session on modern NBA coaching, he jabbed Mike Brown about the Cavaliers’ loss to the Celtics in the 2010 Eastern Conference semifinals, to which Mike Brown responded, “I don’t seem to remember you playing in that series.”
  • The basketball analytics panel, which included Scalabrine, Tom Thibodeau, Mike Zarren, and Brian Kopp (with Zach Lowe moderating), spent much of the time discussing player health, and how to use data and analysis to better manage players’ minutes. The panelists agreed that looking at an individual in a vacuum is highly problematic because it ignores too many factors, such as synergies between players in a lineup. Even lineup data, however, ignores the opponent. As an example, Thibodeau pointed out that even if analysis suggested otherwise, he would put Luol Deng on LeBron James for as long as he could, even if the lineup this created for Chicago wasn’t as effective on the offensive end. Mike Zarren echoed this same principle—considering analytical principles in the context of the players on your team, rather than at face value—by citing Kevin Garnett’s time with the Celtics, when was automatic from 18 feet and was encouraged to shoot at that range, despite the analytical evidence that the long two is a bad shot. It is a bad shot for most players; Kevin Garnett is not most players, and a good analytics team and coaching staff should be cognizant of that.
  • The coaching panel was bizarre. It was supposed to be about modern or innovative approaches to coaching in the NBA, yet the three former NBA head coaches on it—Vinny Del Negro, Mike Brown, and Scott Brooks—were not particularly known for their innovative approaches. The former coaches did touch on one point that I thought was a theme of the entire conference: the folly of trying to borrow another team’s system. This may or may not work at a general manager or ownership level, but at a head coach’s level it simply doesn’t. It can’t, because no two teams have the same personnel. As Mike Brown suggested, you can borrow a principle here and a concept there, but you have to combine and customize them for the group you’ve got in front of you. Nobody can decide to be the Warriors or the Spurs just by copying how the Warriors or Spurs approach the game.
  • Honestly, my favorite session of the entire conference may have been a talk on a new approach to scheduling that the NBA is beginning to employ in 2016. Evan Wasch, SVP of Basketball Strategy and Analytics for the NBA, described how the league is working to resolve players’ and fans’ concerns about the schedule using a branch of mathematics called combinatorics. Each NBA season there are 1230 games for 30 teams in 29 arenas across about 170 days. I’m not sure I believe this, but according to Wasch, this results in “more possible schedule variations than there are atoms in the Universe.” (I assumed he was joking, but he seemed completely serious.) In the past, the NBA essentially compiled the schedule by hand, starting with national TV games and working their way backward. Even building the schedule manually for this season they worked to get back-to-backs under control; for example, they started scheduling more Thursday games, since the limitation of two TNT games meant that most teams only had six days a week for games instead of seven, producing more back-to-backs. As a result, the average number of B2Bs dropped from 19.3 per team in 2014-15 to 17.8 this season, with no team having more than 20. With the combinatorial approach they will use for next season’s schedule, the NBA can take it a step farther by applying a set of logical constraints (e.g., “don’t let any team travel across two time zones for a single away game” or “don’t schedule any Spurs home games while the rodeo is in town”), and within five or ten minutes, the model spits out a schedule.
  • There wasn’t as much discussion of the Golden State Warriors as I expected coming into the conference. Popular team on an historic run—I expected it to be inescapable, and while most sessions mentioned the Warriors in one way or another, only one of them that I attended made them a focal point as a team. That said, the SSAC was obsessed with Draymond Green. Nate Silver used him on the first day of the conference to demonstrate that analytics often supports conventional wisdom, and Brian Scalabrine used him on day two to point out that even with all of the data and predictive models that we’ve got these days, Green still went 35th overall in the 2012 NBA draft. We’re far from perfect in our use of predictive analytics. Scalabrine also “wondered” why Draymond Green came off the bench for his first two seasons, which was an awesome not-so-subtle jab at Marc Jackson, with whom Scalabrine had beef while on the Jackson’s staff in Golden State.
  • The final panel of the conference was a great one, with a trio of current general managers—Daryl Morey (Houston Rockets), Jeff Luhnow (Houston Astros), and Bob Myers (Golden State Warriors)—talking about their approach to building a team. Bob Myers got his start in sports as an agent working with the legendary Arn Tellem (who has participated at SSAC in the past himself), so he feels like he has a unique perspective as a general manager. He says that agents have gotten very good at using data and analytics to “spin” their players’ value. Myers went on to talk about how the most overlooked aspect of analysis is communication. If you are going to tell someone how to do their job, you need to have a strong relationship with them first, and you need to be sure you communicate in a way that doesn’t make the person defensive or discouraged. Daryl Morey also hinted at why his team was so slow getting out of the gate this season, saying that they chose to prioritize rest and player health over hard work and preparation during training camp and the preseason. (Apparently that was on Kevin McHale.)

And with that, SSAC 2016 came to an end. By now they will have cleared out of the conference center completely until 2017.

Sometimes when I tell a friend about the conference, they ask how I keep up with all the math. I didn’t attend any of the research presentations this year, so I don’t think I saw a single formula or equation. A lot of the appeal of SSAC for me is in the opportunity to listen to people like Shane Battier or Brian Scalabrine or Jeff Van Gundy or Bob Myers talk about their experiences in the NBA, or to hear the original Moneyball crew give their perspective almost 15 years after the book was published. In my opinion, SSAC is for anyone who cares about understanding the sports they love at a deeper level, from the mouths of participants themselves. There was plenty of that here over there past two days.
So I’ve got my fix of sports analytics and business for the next 12 months, and yet I can’t wait to see what the conference organizers put together for next year. I’ll see you in 2017, Sloan Sports Analytics Conference.

  • 2015-16 Jazz SWOT Analysis
    Analytics
    0
    April 22nd, 2016

    2015-16 Jazz SWOT Analysis

    SWOT analysis (Strengths, Weaknesses, Opportunities, Threats) is a popular planning method whereby an organization assesses its...Read More

Leave a Reply

Your email address will not be published. Required fields are marked *