Sloan Sports Analytics Conference, Day 1

I finally made it to the Sloan Sports Analytics Conference (#SSAC on Twitter) and it was all I hoped it would be.  After all the powerpoints I watched today I'm only capable of writing in bullet points, so on to the highlights, in chronological order:


  • Houston Rockets GM (and event founder) Daryl Morey had a great one-liner that POTUS would be welcome at the conference so he could be disabused of the notion he expressed on the BS Report that Blake Griffin should shoot more jumpers.*  Even if the joke hadn't been funny, I would have pretended it was--I was sitting next to Daryl's dad!
*A topic close to my heart given my position in this post.
  • The opening panel with NBA Deputy Commissioner Adam Silver, NHL Commissioner Gary Bettman, Scott Boras, Rob Manfred of MLB, and Steve Tisch of the New York Giants was noteworthy mostly for the drool collecting in Silver and Bettman's laps as Tisch discussed the NFL's revenues.
  • Scott Boras had a good point that younger players' "myopic" focus on simply making it to the league results in them getting the short end of the stick in CBA negotiations.
  • What I would have loved to ask Silver was whether BRI or "System issues"/"perceived competitive balance"* was more important to the owners.  To these eyes, they mostly caved on the system issues to get the players to accept a lower percentage of BRI.  Does that mean that the small-market owners, when it comes down to it, care more about making money than winning?
*Research conducted during the lockout concluded that teams that spend more don't really win more, but the key here is that the owners appeared to believe that limiting the high-spending teams would improve competitive balance.  From that standpoint, they gave up perceived competitive balance to make more money.
  • The research highlight for me was this paper on clutch performance.  While this was only one of its many conclusions, what I found most interesting was the finding that offensive rebounding percentage increases in clutch situations--for the home team only. Statheads (including myself) are often quick to dismiss the effect of intangibles such as crowd noise on players, usually with a rationale such as "an NBA player who has played thousands of games in his life wouldn't just decide to try harder because the crowd is cheering."  I found the paper's finding refreshing, as it would appear that there is a genuine positive effect of the crowd exhorting the players.  I brainstormed with Jeremy Conlin of Suitesports.com (@JConlin3789) for another reason other than crowd noise that home team offensive rebounding might rise.  The best we could come up with was the refs calling less loose ball fouls against the home team, which struck me as a rather unlikely explanation compared to the simpler explanation that the crowd noise just makes players try harder.  Anybody have any other explanation for this clutch increase in home team offensive rebounding?
  • My big takeaway from a lot of the research papers was the overarching lesson to account for their conclusions, but only as another data point in making a decision rather than the be-all, end-all arbiter.  James Tarlow's excellent paper on experience and winning in the NBA was an example.  His data indicate that teams should be kept together and that coaches with more playoff experience tend to do better in the playoffs.  But, as I think he would acknowledge, this doesn't necessarily mean management should keep together a lousy team or hire a playoff-experienced coach who isn't a good fit for the roster.  Nor can we say conclusively that continuity wasn't the result of, rather than the cause of, being a good team, since good teams are less likely to be broken up.  He sure held up well under a pretty merciless post-presentation grilling though.  Incredibly, when I asked Tarlow reported that he has neither blog nor Twitter.*  Here's hoping he starts one so we can see more work from him.
*You know, just like me 3 months ago.
  • I didn't have a chance to see the presentation on using optical tracking software to chart rebounds because I accompanied my friend and Sloan '13 student @zhouey to the live taping of the BS Report with Bill James, but I'll have a chance to see it when they present again tomorrow.  This technology probably has the greatest potential to revolutionize NBA analytics of anything that's come along so far.
  • I had a couple nice conversations with Henry Abbott (@truehoop) in which I thanked him for his kind words about The Team Rebound in this post.  He also invited me to play in the "only corner 3s and layups" TrueHoop network pickup game on Sunday, which I sadly deferred until next year because I'm still rehabbing a torn ACL and patellar tendon.  His sage advice: "Next year, don't tear your ACL and patellar tendon."
    • Who was the worst GM of the last year?  Danny and Andres insisted on Otis Smith. I had to nominate the Warriors' "braintrust" for displaying so much confidence in Andris Biedrens' ability to live up to his $11 million a year contract that they used their amnesty on Charlie Bell...so they could sign Biedrins' replacement DeAndre Jordan for another $11 million a year.  If they had confidence that Biedrins would be good, why clear the room to sign Jordan in the first place? When Michael Lewis talked about finding undervalued players, I don't think cornering the market on $11 million a year lefthanded centers who shoot under 50% from the line is what he had in mind.
    • Was the institution of zone defense a good thing?  Yes and yes from Zach and me.
    • How much better is the league now than 20 years ago?  This allowed me to bring up one of my favorite teams, the 1990-91 Bucks, which somehow won 50 games and captured the 4th seed in the East with a starting lineup of Jay Humphries, Alvin Robertson, Fred Roberts, Frank Brickowski, and Brad Lohaus.*  That frontcourt would get completely housed in today's NBA.
*It's worth noting that same Bucks team had also recently employed Randy Breuer and Paul Mokeski. Amazing.
    • In the same vein, what is the most recent NBA champion that the worst team in the league today (currently the Bobcats) could beat?  Zack nominated the '75 Warriors, a pretty good pick since the league was diluted by the amount of talent still in the ABA and the Warriors didn't have much of a post presence to abuse the 'Cats inside.
  • I met a ton of awesome, friendly people today who didn't make me feel stupid for showing up with a 3 month-old, 10-post blog.  Can't wait to do it again tomorrow.

2 comments:

  1. ND,

    I am sure you have seen Kahnman and Teversky's papers on "the hot hand" fallacy. How good is the data to back up a statistical increase in rebounding for the home team when crowd noise is up? Did they decibel testing to see if there was a correlation between increased sound and increased rebounds? (Note: I have a feeling that correlation ends when the noise starts making players' ears bleed.)

    My theory, as usual, is that there is an evolutionary reason. Studies show that winning teams experience a boost in testosterone, while losing teams experience a loss. Turns out this is true even when an invested fan is watching sports, which is why I gorilla mounted that asz so often in 2010.

    I expect this is a similar deal. When clans of monkeys or cavemen were after each other, the clan with the most/biggest combatants was making the most noise. At a subconscious level, this releases chemicals/enzymes/adrenaline or whatever that triggers FIGHT over flight. Hearing the loud noise against you probably triggers flight. Since in essence you have to be assertive and fight for rebounds, my guess is this is why... however why offensive but not defensive rebounds would fall in this category... IDK.

    WW

    ReplyDelete
  2. I didn't pore through his data, but basically what it found was that home team offensive rebounding percentage increased by about 3 percentage points in clutch situations, which means that teams grabbed about 10% more offensive rebounds.

    One plausible reason for why offensive rebounding percentage increased is that offensive rebounding is usually the result of a Herculean individual effort. That is why the paper's authors picked it to contrast with clutch free throw shooting, which is not effort-dependent. In fact, home team free throw percentage declined in clutch situations, while road team percentage did not.

    ReplyDelete

Keep it clean.