As you may have heard, there’s been a whole hullabaloo recently in the hockey world about the Toronto Maple Leafs. Specifically, they had a good run last year and in the beginning of this season that the more numerically-inclined NHL people believed was due to an unsustainably high shooting percentage that covered up their very weak possession metrics. Accordingly, the stats folk predicted substantial regression, which was met with derision by many Leafs fans and most of the team’s brass. The Leafs have played very poorly since that hot streak and have been eliminated from the playoffs; just a few weeks back, they had an 84% chance of making it. (See Sports Club Standings for the fancy chart.)

Unsurprisingly, this has lead to much saying of “I told you so” by the stats folk and a lot of grumblings about the many flaws of the current Leafs administration. Deadspin has a great write up of the whole situation, but one part in particular stood out. This is a quotation from the Leafs’ general manager, Dave Nonis:

“We’re constantly trying to find solid uses for [our analytics budget,” Nonis said. “The last six, seven years, we’ve had a significant dollar amount in our budget for analytics and most of those years we didn’t use it. We couldn’t find a system or a group we felt we could rely on to help us make reasonable decisions.”

[…]

“People run with these stats like they’re something we should pay attention to and make decisions on, and as of right now, very few of them are worth anything to us,” he said at one point during the panel, blaming media and fans for overhyping the analytics currently available.

This represents a mind-boggling lack of imagination on their part. Let’s say they honestly don’t think there’s a good system currently out there that could help them—that’s entirely irrelevant. They should drop the cash and try to build a system from scratch if they don’t like what’s out there.

There are four factors that determine how good the analysis of a given problem is going to be: 1) the analysts’ knowledge of the problem, 2) their knowledge of the tools needed to solve the problem (basically, stats and critical thinking), 3) the availability of the data, and 4) the amount of time the analysts have to work on the problem. People who know about both hockey and data are available in spades; I imagine you can find a few people in every university statistics department and financial firm in Canada that could rise to the task, to name only two places these people might cluster. (They might not know about hockey stats, but the “advanced” hockey stats aren’t terribly complex, so I have faith that anyone who knows both stats and hockey can figure out their metrics.)

For #3: the data aren’t great for hockey, but they exist and will get better with a minimal investment in infrastructure. Analysts’ having sufficient time is the most important factor in progress, though, and the hardest one to substitute; conveniently, time is an easy thing for the team to buy (via salary, which they even get a discount on because of the non-monetary benefits of working in hockey). If they post some jobs at a decent salary, they basically have their pick of statistically-oriented hockey fans. If a team gets a couple of smart people and has them working 40-60 hours a week thinking about hockey and bouncing ideas off of each other, they’re going to get some worthwhile stuff no matter what.

Let’s say that budget is $200,000 per year, or a fraction of the minimum player salary. At that level, one good idea from the wonks and they’ve paid for themselves many times over. Even if they don’t find a grand unified theory of hockey, they can help with more discrete analyses and provide a slightly different perspective on decisions, and they’re so low cost that it’s hard to see how they’d hurt a team. (After all, if the team thinks the new ideas are garbage it can ignore them—it’s what was happening in the first place, so no harm done.) The only way Toronto’s decision makes sense is if they think that analytics not only are currently useless but can’t become useful in the next decade or so, and it’s hard to believe that anyone really thinks that way. (The alternative is that they’re scared that the analysts would con the current brass into a faulty decision, but given their skepticism that seems like an unlikely problem.)

Is this perspective a bit self-serving? Yeah, to the extent that I like sports and data and I’d like to work for a team eventually. Regardless, it seems to me that the only ways to justify the Leafs’ attitude are penny-pinching and the belief that non-traditional stats are useless, and if either of those is the case, something has gone very wrong in Toronto.