While there was a ton to be learned in Day 1 of the Sloan Sports Analytics Conference, it actually built to Day 2, the day in which the Basketball Analytics, Ownership, and Sports Gambling panels were presented. Like Day 1, a lot of interesting research was presented as well. Here are the bullet points of what I learned in Day 2.
Adam "Sexy" Silver
- Adam Silver, the NBA's current deputy commissioner and soon to be actual commissioner, appeared on the first panel of the day regarding ownership in sports. What impressed me most about Silver was the extent to which he dominated the conversation: he acted like a man fit to be commissioner, albeit probably not in the same way as David Stern. It was a big change from the sheepish grins he gives as he announces round 2 of the NBA draft when the crowd chants his nickname.
- That being said, there wasn't anything new or revealing about what Silver said, he was very much speaking from a ownership point of view. He called the last CBA negotiations "a draw", when most observers called it a win for the owners, and insisted on the need for year-to-year profitability for NBA franchises, even as franchise values skyrocket.
- Silver explained the NBA's current revenue sharing system as a "federalist" system, in which teams have an incentive to make huge money, but still put a percentage of those revenues into a league pool to help out the smaller market teams.
- This presentation, at first glance, is really out there: you're telling me that we can measure team chemistry (previously an intangible) through how well players' heart rates sync up? But by the end of the talk, it sounded less crazy.
- Basically, people who are in close proximity to each other for a long time end up displaying synced heart rates; when my friend's heart beats, mine probably will at the same time. Scientists aren't sure why this happens, but it's a real phenomenon. In the lab, people with synced heart rates turn out to perform better at all sorts of teamwork-oriented tasks.
- The study's authors, Daniel McCaffrey and Kevin Bickart, placed heart rate monitors on a Division 1 women's soccer team, and a Division 1 men's basketball team. They then created network graphs of the team's collective heart rate; basically, visual representations of how synced any two players' heart rates were.
- Then, they showed us video of two in-game situations. I'll use the basketball one: two players working together in a give and go. Those players, as the give and go unfolded, had their heart rates become more and more in sync, even as they did completely different tasks (one player cutting, one player passing from the post).
- Of course, is an anecdotal instance of this happening statistical evidence that this happens all of the time? No. But it's at least interesting to imagine what we previously considered unmeasurable as surprisingly so. More on this topic can be found at two posts on Hardwood Paroxysm: here and here.
- The next presentation was by Michael Mauboussin, author of the book "The Success Equation: Untangling Skill and Luck in Business, Sports, and Investing". By now, this shouldn't surprise anyone, but there's a lot of luck involved in sports. Whether or not a ball hits the rim and goes in, time runs out a split-second before leaving a player's hand, and injuries are all example of the impact of luck.
- However, humans have a tendency to ignore the role of luck in their endeavors, instead looking back and fitting an explanation that fits the outcome. Very often, we have outcome bias in sports: we celebrate Al Jefferson when he makes a game-winning three, but criticize him when he misses. Instead, we need to look more at the process, and whether or not the right actions were taken to maximize the chance of success.
- One final note: according to Mauboussin's studies, basketball is actually the least luck dependent of the big 5 team sports. Football has a short season and 1-game elimination playoffs, baseball has relatively low scores and a huge luck factor with hitting, and hockey and soccer both have extremely low scores, in which one score often makes the difference.
- This was easily the most entertaining panel of the conference, largely due to the conflict between Haralabos Voulgaris, everyone's favorite NBA sports bettor, and Matthew Holt, Director of Sports data for Cantor Gaming.
- Voulgaris was upset at Holt for a comment he made in last year's conference, in which he asserted that any player could place a $100,000 to $1 Million bet with his company. Unfortunately, they've limited Voulgaris to $5,000 since, and there was a ton of conflict. Some quotes:
Voulgaris on inefficiencies: "Back when I started, bookmakers were even more clueless than they are now." #SSAC13— Gill Alexander (@beatingthebook) March 2, 2013
Voulgaris: "In reality, lines are created offshore. Bettors (then) are making the line. Bookmakers (here) aren't." #SSAC13— Gill Alexander (@beatingthebook) March 2, 2013
Voulgaris: "In reality, bookmaking is pretty easy. Some of the dumbest people I know are bookmakers. Not you, Matthew." #SSAC13— Gill Alexander (@beatingthebook) March 2, 2013
Voulgaris: "How much will you take, Matthew? Let's go to Vegas right now." Holt: " I understand you're not having a great NBA season."— Gill Alexander (@beatingthebook) March 2, 2013
Voulgaris: "Bookmaking is a parasitic business. The least they can do is take a minimum of $5000 from every person." #SSAC13— Gill Alexander (@beatingthebook) March 2, 2013
Voulgaris on challenging Holt on validity of statements last year on Cantor taking all bets: "This is the only reason I came here." #SSAC13— Gill Alexander (@beatingthebook) March 2, 2013
Voulgaris: "I was so excited when I heard that Cantor was gonna take $100,000 on NBA bets only to find out they wouldn't take $5000" #SSAC13— Gill Alexander (@beatingthebook) March 2, 2013
Millman to Voulgaris: "If it's so hard to get $$ down, why do it?" Voulgaris: "What else am I gonna do. Be a professional tweeter?" #SSAC13— Gill Alexander (@beatingthebook) March 2, 2013
Voulgaris to Holt: "I think when u first started out, u had this idea u were going to take on sharps. Then u realized how difficult that is"— Gill Alexander (@beatingthebook) March 2, 2013
Holt: "I don't want to tell people how to build a model." Voulgaris: " You dont know how to build a model."— Gill Alexander (@beatingthebook) March 2, 2013
- In short, the panel was awesome. That being said, Voulgaris specifically declined to reveal any inefficiencies he had found, so that he could continue to "put food on my table".
- The Basketball Analytics panel had some big names: Kevin Pritchard (GM of the Pacers), Zach Lowe (Grantland), Mike Zarren (assistant GM of the Celtics), and R.C. Buford (GM of the Spurs). However, no one was more talkative or interesting than Stan Van Gundy, the former Magic and Heat head coach.
- Van Gundy was generally in favor of analytics, but had plenty of examples in which the use of analytics could actually hurt a team. For example, he says he was once presented with a 5-man lineup that had done extremely well together, but in only 39 minutes. While he didn't actually use the word sample size, he was unwilling to trust data with a lineup that had been together for only that long. The rest of the panel complimented him on his statistical intuition.
- Van Gundy also took on the idea of 2 for 1s: while they certainly make sense analytically, Van Gundy decried that they chipped away at a system. In short, if you say "you shouldn't force up bad shots... unless there is between 30 and 35 seconds left on the clock", you start to introduce uncertainty into the system. Now, Van Gundy says, a player sees that exception, and tries to widen it by taking bad shots afterwards.
- Most importantly for the Jazz, Van Gundy took exception to the idea that young guys should get more playing time to better their development. If a player is just given playing time, Van Gundy argued, there's no teaching mechanism in place, and no incentive for a player to play smart basketball. He used the example of the Washington Wizards, and in particular, Andray Blatche and Javale McGee: two players who were immediately given playing time, but as a result, haven't bettered themselves to be intelligent players. Without the ability to change playing time, the players don't do as their coach instructs.
- On the other hand, when a coach can use playing time to his advantage, a system can be run as designed. Van Gundy used the example of J.J. Redick, who started out as a poor defensive matchup player, owing to his size and speed. However, he implemented the defensive system so well (Van Gundy asserted that Redick never missed a rotation) that the team's defensive rating was excellent. Zarren and others agreed that defensive stats are impossible to measure well without a knowledge of the system that the individual team is using.
- This obviously has a whole load of implications for the player development for the young guys on the Jazz, I'll put up a more thorough post on my thoughts on this subject later in the week.