Arturo avatar

The Geeks at Sloan, Part 2: The Data Should be Free

For the second year in a row, the Box Score Geeks went to Sloan! Click here to read part one.

My concern is that we’re at the end of a golden age; there’s real value to having the crowdsourcing model. You have this large community that’s passionate about sports. It’s important for leagues to tap into this energy. [SSAC] shows there’s demand for this.” - Arturo Galletti

That's a quote from an interview I had with Peter Dizikes for MIT News (click here to read the article), hitting on what I see as the key issue facing sports analytics going forward: who controls the data? Seven of the eight research papers being presented at the conference were prepared using propietary data.

Why is this a problem? If you think about some of the seminal figures in the sports analytics movement (Bill James being the classic example), the outsider working on publicly available data is a key player in the process. By not having a stake, the outsider is able to go where the insider wouldn't. There are also the advantages that come with crowdsourcing (which is something that we take advantage of constantly). I understand why teams want to put up some walls, but I feel like it significantly hinders the progress to be made. It also leads to another problem that teams have: they are simply not prepared to pay the market rate for work that they've been getting for free. Once they restrict the flow of information to those highly motivated, highly prepared, and highly passionate analysts that are out there doing analytics pro-bono, they will be forced to rely on their own in-house teams and academia. To illustrate the problem with this, allow me to paraphrase Bryan Colangelo: "Teams should be able to find $250 000 to spend on analytics." Mr Colangelo: $250 000 will not get you very far.

The other concern is a journalistic and scientific one. As the data goes into silos, we won't be able to test and replicate results. Transparency and accountability will be harder to come by. I take my journalistic and scientific work very seriously. I take the time to evaluate and analyze what teams do, and the barriers going up make this work harder. There is a lesson to be learned from the NFL here: the NFL has actually made as much of its information public as is possible. If you want to have access to every single shred of tape the NFL has, you simply need to pony up and pay for it. I'm not arguing that professional sports leagues or teams need to give away their information – if they sold access at reasonable prices I'd be willing to pay for it – but there needs to be a way for outsiders to gain access to it.

The first day of the conference started with an all-star panel featuring Bill James, Daryl Morey, Nate Silver, Kevin Kelley (coach of no punting pulaski academy), and George Karl. This was a fun and informative panel. Even if some of the panelists looked a little like they got up on the wrong side of the bed (I won't name names, but one of them had bed hair). There were so many nice little tidbits.

For one, everyone seemed to be bewildered by the state of analytics in the NFL. The feeling was that the NFL was moving one step forward and two steps back. Nate Silver summed it up nicely: NFL has low incentive to innovate because they have such a profitable and popular product. When it comes to football statistics, High Schools and colleges are more likely to innovate than the NFL Brian Burke even built them a nice, publicly available tool for working down fourth down probabilities and no one seems to be using it. Bill James and Daryl Morey don't get playcalling on fourth down either. Kevin Kelley loves the fourth down bot. And I quote: "Going for it on fourth down makes us more likely to score". Simple logic, no? Charming, smart, and innovative, Kevin Kelley was the star of a panel with Daryl Morey, Bill James, and Nate Silver. Seriously, why is this guy not coaching a major college or an NFL team again?

Bill James hit on some crucial points. "A gimmick's simply an innovation so ahead of its time that there's no foundation of knowing whether it's going to work." In essence, Bill hit on the key fact that, in general, innovation is an outsider's game. He had nothing to lose when he started and no stake in the establishment, so he was willing to take the road less-traveled and arrive at a novel conclusion. Teams and institutions are always pulled to dwell in the past, and this is doubly so in baseball. Major league baseball needs to innovate and fix the perception that baseball is a boring sport. This is supremely hard for baseball because of who they are ("people still upset about the DH rule. It's been 41 years – get over it"). Oh, and Bill James wants to measure player potential statistically. I love Bill James.

George Karl was curmudgeonly. He said that, in the NBA, the best team and the worst team aren't that far apart, and then devolved into coach speak ("teams that know how to win") while making googly eyes at GMs in the room (Hollinger: protect your maidenly virtue!). Yeah, GK, that's not right. I did like his suggestion of a single-elimination tournament after All-Star break. And his impassioned plea to the statheads to look for a formula to explain love had me picturing a basketball Dumbledore ("it's love Andres...."). There is something sublime about sitting next to Andres Alvarez while you hear George Karl say that he was just trying to play his best players.

Nate Silver did get over his bad hair day and make some very nice points as well. Like how statistics show that, of all the major North American sports leagues, the NBA's champ is the closest to being the best team in the league (I've done the math and agree wholeheartedly). Oh, and Silver also mentioned on of my favourite observations about pro sports: that the typically capitalist US has a socialist sports model, whereas typically socialist Europe has a capitalist sports model. Looking forward to the new website, Nate (call me :-) ).

Daryl Morey was the first to bring up what would be a recurring theme of the conference: getting rid of the marginal incentive to lose. We've talked tanking to death here, so I won't bore you too much. Suffice it to say, a significant change appears to be coming. Mike Zarren's draft wheel is on its way and it appears like the league is working out the finer points. Be prepared for at least ten thousand words from me on this subject during the summer.

Stay tuned for Part 3: Lies, Damn Lies, and Statistics

Good article.

I hate the draft wheel idea. Sport is more entertaining when it is more competitive. It is more competitive when worse teams draft higher. That is why drafting based on record makes sense.

Sport is also entertainment. Certainty is boring ('here are my team's draft picks for the next three decades'). Uncertainty is exciting ('which pick will we get in the lottery this year'). There is a huge amount of debate and engagement among fans each year on who their team will draft ... because it's an uncertain process.

On the other hand, you also don't want to incent losing.

The financial rewards for making the playoffs give team owners a large incentive to qualify. So the concept of giving the top draft picks to the teams that don't make the playoffs works. People are unlikely to intentionally lose their way out of the playoffs.

The current lottery odds, however, give people a large incentive to lose. The more you lose, the better your pick: guaranteed. This guarantee needs to be removed. You need a chance that the absolute worst team gets pick number 14. Rather than one added loss improving your pick position from 2-5 to 1-4, you need it to improve your odds of a higher pick by a couple of percent.

So the solution is to still have a lottery among non-playoff teams, with changed odds that remove the certainty, and make the reward for losing imperceptible. Forget the wheel!
The move toward proprietary data is a huge disappointment as a fan. I understand, of course, why teams would want to own it - they are trying to get a leg up on their competition, obviously. But, for the field as a whole, it's certainly a bad thing.

Thanks for the posts on Sloan, guys.
I think the draft wheel is bad for the reasons above.

What I haven't really heard is why an unweighted lottery for all the draft spots outside the playoffs is a bad thing for the NBA. If you miss, you have an equal chance at every spot. If you make it in the playoffs, that is its own reward.

What exactly is the problem here? Bad teams always have the illusion of hope, you don't reinforce good teams winning, but there's no marginal incentive to lose. Being worst vs. 15th worst is identical. It's a hard in/out line for playoffs. To benefit from losing, you'd have to deliberately miss the playoffs, which is something most owners won't condone unless you truly are doing a full tear-down (the point of which is cleaning house, the losing being a side-effect, but even in that context, it eliminates the incentive to be an atrocity against man instead of merely normal bad).
@KuH

"Sport is more entertaining when it is more competitive. It is more competitive when worse teams draft higher."

Well, not really ! I am reading Prof Berri's books right now, and he shows that the links between entertainment and competitive balance , competitive balance and a lottery weighted draft are not evident.

I like the draft, and like you, I like to guess which team will draft which player. The wheel does not cancel that. It cancels the uncertainty of the lottery, not the uncertainty of the draft itself.

@Arturo

"I won't name names, but one of them had bed hair"

"Nate Silver did get over his bad hair day"

You made me laugh.
Someone once told me that a draft wheel is just a flat circle.
Ok, I've said it more than once on this forum and continue to not understand how more people don't pick up on it: the league will NEVER change for the better until we disincentive the decision makers. Punish the owners and/or GMs for being crappy over a sustained period. If we are not going to touch that monster in the room issue (crappy owners/decisions makers get rewarded by the current structure) then we will never fix the problems. Screw the wheel, that doesn't fix anything. You'll still have crappy teams that don't try to win because they get money either way. Make the owners AND GMs responsible for consistent horrible play (by force selling teams that have been bad for too long a period and by not allowing GMs with a bad enough record to have an NBA job for a certain period of time) and you'll see improvement. Everything else is just window dressing.
Adam Silver mentioned something in passing to Malcolm gladwell that I clung to. He said "At the same time ā€“- and I know there was some discussion yesterday about so-called tanking ā€“- these are to the extent that teams believe they may have an incentive to perform poorly. Iā€™m not sure analytics bear out that being the optimal way to operate in that situation". Yes, yes, yes, a thousand times yes!
Nate Silver's site will be awesome, and we won't be subjected to his hair.
As for Karl, the one compliment I have after watching 88 games of Nuggets basketball last season is that Karl's team liked transition and shots close to the rim, and this years team has lost interest in those concepts. How much the coaching change has to do with that, I don't know. But if that really was Karl's doing, then despite his playing sfs over faried at the 4 and love of Corey brewer, his scheme was improving all his players. What if a coach had a decent scheme and a decent understanding of when to play players...Arturo, please come coach the Nuggets.

Sign in to write a comment.