Jury still out on value of analytics
Bettman says NHL not about to tweak its playoff system
March 11, 2016, was a momentous occasion: the day on which I unequivocally agreed with Gary Bettman on something.
The NHL commissioner, speaking at the MIT Sloan Sports Analytics Conference — the annual gathering on the Boston waterfront that is like Woodstock for data nerds — was asked about the possibility of tweaking the NHL playoff system to add one- game “play- in” games similar to what now happens with baseball’s wild- card teams.
“To me, it doesn’t make any sense,” Bettman said, shaking his head. Why go through the grind of the 82-game regular season just to be punted from the playoffs in one day?
Moments later, after acknowledging the great success of the new 3- on- 3 overtime format, Bettman was asked if could see the NHL introducing it, or the shootout, in the playoffs. He didn’t hesitate: no. TWO agreements with the NHL commissioner in the space of five minutes? Was this the end times?
Thankfully, no. For the rest of his chat on stage, the NHL commissioner was his usual self, dismissive of all criticism and firmly ensconced in a world in which the league has never made a poor decision and hockey always manages to be better today than it was yesterday.
Was he concerned that Calgary’s Dennis Wideman had already missed 19 games while he was appealing his 20- game suspension? Bettman said he was “perfectly comfortable with the process.” (Literally an hour later, an independent arbitrator reduced it to 10 games).
What was his assessment of the much- maligned overhaul of the league’s website? “It wasn’t perfect, but it was as close to perfect as you could do.”
Fan voting for the All- Star Game? “We are evaluating it,” he said, “but in the final analysis, we were happy” with John Scott’s inclusion.
And were the struggling Canadian teams a problem for the league? “If you are Rogers, you may be a little less than thrilled.”
Make that three times I agreed with him. Night is day, etc.
To note the 10th year of the Sloan conference, the organizers convened a Moneyball reunion: author Michael Lewis, who wrote the seminal 2003 book that introduced the concept of advanced stats to a wide audience; Bill James, the researcher who was a pioneer of the analytics movement; and Paul DePodesta, the former Oakland Athletics executive who was played by Jonah Hill in the movie. (Billy Beane, the Athletics GM, was a late cancellation.) All these years later, the essence of analytics, or advanced stats, or whatever you want to call it, remains the idea of discovering information about sports that isn’t immediately apparent on the surface.
Lewis, in 2002, found that his experience with the Athletics was particularly literal in this regard: they looked like a collection of accountants.
“One guy had fat ankles!,” Lewis exclaimed, telling the story about seeing the Oakland players coming out of the shower and being shocked that few of them were particularly chiselled.
That made them ideal Athletics. A scout for another team might see a fat guy, but DePodesta and Beane saw hidden value.
DePodesta, who was recently hired to run the operations of the Cleveland Browns — yes, a foot- ball team — said overcoming those natural biases remains a big part of analytics. He said that 20 years ago, scouts would watch a guy they didn’t want to draft, and if he waggled the bat a lot, they would note that he had far too much pre-pitch movement. That would never work in the majors, they would write. Then they would see a guy they liked, and if someone pointed out that the guy waggled his bat a lot, they would say, yeah, he’s just getting himself locked in: “That guy’s going to get after it!”
Thus the challenge with scouting: “We see what we want to see,” he said.
Sometimes, the resistance to analytics simply comes down to terminology. In soccer, one of the popular new metrics is Expected Goals, which takes into account shots, shot location, proximity of defenders, and other factors to spit out a number.
It’s a better predictor of future success than goals and points. Blake Wooster, who runs a consultancy called The 21st Club, recalled a Premier League manager who called the idea of Expected Goals “nonsense.”
It was like he couldn’t wrap his head around it: goals aren’t to be expected, they must be EARNED. But then, after a losing game, he would give an interview and say, like every manager since the dawn of time, that they created good chances and deserved a better result.
“That’s like, literally, the definition of expected goals,” Wooster said.
The manager was later sacked. His team’s Expected Goals, at the time, was pretty good.