The duller side of the statistical cutting edge
It's not easy to be on the cutting edge of sports. The MIT Sloan Sports Analytics Conference is a gathering of the finest in the quantitative analysis community, and the reputation of the conference has grown exponentially every year. The audiences are bigger than ever, the conference's schedule has expanded from one day to two, and there's no doubt that each year's event will be the biggest and baddest yet.
All of which makes research that doesn't shake the world a bit underwhelming. The "watch the games" vs. advanced stats dynamic has created an adversarial expectation; for a variety of reasons, advanced stats are only really deemed to be conversation-worthy when they challenge the things we think we know. That's not necessarily a fair standard, as empirical data that supports the status quo opinion can be just as valuable as more controversial findings even if they don't specifically instigate perceptual change.
Case in point: Arup Sen's research paper entitled, "Moral Hazard in Long-Term Guaranteed Contracts - Theory and Evidence from the NBA." There are some interesting figures that Sen noted in the presentation of his research, but his conclusion was ultimately quite predictable: NBA players perform better in their contract year than they do otherwise. That's information that most NBA fans already had, even if it was only confirmed by bits of anecdotal evidence.
That said, there’s a necessary distinction between interesting work and important work. Sen’s research doesn’t challenge conventional wisdom, and thus in the context of this conference, it pales in comparison to the draw of some of the other presentations or panels. There’s no gut reaction at all to Sen’s conclusion because it falls so perfectly in line with basketball common sense. It may not generate discussion, but work like this is nonetheless important in setting a precedent for the way that topics in this field are broached analytically.
If basketball analysts seek to disprove the conventional tenets of the sport, they’re inevitably going to test some hypotheses that don’t break new ground. After all, some of those tenets have been established for valid reason; not every headfirst dive into a spreadsheet has to run contrary to the things we think we know, and that’s important to keep in mind with the presentations at the Sloan conference this year and in every year to come.
When analytics were properly introduced to the basketball world, it opened the door for exploration, but also re-examination. Quantitative analysts (and by extension, the rest of us as sports fans, media members, and hopefully team executives) are not merely supposed to be looking at the game in new ways, but constantly reviewing older perspectives. Some aspects of those perspectives will hold up, and the contract year phenomenon is clearly one of them. Affirmation doesn’t make for an exciting story or warrant a place in the accounts of how analytics will change the world, but it’s a part of the same process that grants data-centered approaches their empirical power.