For those who might be unfamiliar with the story, here is a quick rundown of how the Leafs came to put together their analytics department when Kyle Dubas was hired in 2014: In that summer, Toronto hired Cam Charron (then-editor of this very website), Rob Pettapiece, and Darryl Metcalfe to create a small group of analysts that were all well-versed in “advanced stats”. Metcalfe’s hiring was particularly interesting because to that point he had created ExtraSkater, probably the most prominent numbers resource at the time. It went offline, and the Leafs presumably brought it in-house.
But that department doesn’t look the same today, it’s grown quite a bit.
As of right now, there are seven people listed in the Leafs’ analytics staff; 4 analysts and 3 developers. And that doesn’t include Metcalfe, who has since been promoted to special assistant to the general manager. For a team that until 2014 looked clueless and fell into every overpriced free agent trap, it’s been quite the turnaround over the last half-decade. To that end, if you were to ask most people who follow this side of the league closely which team is at the front of the so-called ‘analytics movement’, the Leafs, headed by Dubas, would likely get a big shout-out.
But is this group now being left behind now by more effective hires around the league? In the last couple months it’s something I’ve thought about a bit, and I don’t think it’s unfair to question whether the Leafs have let their numbers department get bloated and somewhat stalled. It’s great to throw a lot of money around and funnel in as much information as possible, but I think it’s worth asking if eventually it’s possible to start out-thinking yourself or, conversely, if this group simply doesn’t have management’s ear the same way others do around the league (ex. Carolina, Colorado, and now New Jersey).
Why? Well it seems there’s simply lack of evidence that the numbers as we understand them are driving much of the Leafs’ decision-making. We know the Leafs would collect and use any and all metrics that ExtraSkater included (Corsi, Fenwick, and so on), and they definitely use an Expected Goals model that is likely very similar to that of MoneyPuck, but it’s hard to see how that’s showing up in their roster makeup. In fact I’ve caught myself, more often that I’d like, making excuses for roster moves and chaulking it up to “The Leafs have a lot of tools, they must see something in the numbers that we don’t“…but that sort of gets old quick. And Toronto’s management group can’t be completely without criticism because, well…no one should be, and quite frankly the results aren’t there yet to justify it. As much as I praise the overall direction of the club under Dubas, there are fair criticisms we can levy on this group too. You might remember the Leafs sputtered for a good portion of last season after starting hot, had an unbelievably difficult time nailing down some stability in goal outside Freddie, and didn’t seem to see any problem in running out a playoff lineup that included boat anchors in Hainsey and Zaitsev again. Sure, we can keep blaming some of these oversights on Babcock, but he doesn’t make trades or signings. He can only play the guys under contract with the Leafs.
This summer has been an improvement, and perhaps it just has to take time before Dubas can fully settle into the role and make the moves he wants to. But even then, it seems other teams are getting better and being bolder, so it’s not hard to wonder if the Leafs are going to end up chasing.
For instance, it isn’t uncommon to see the Hurricanes put together a tidy piece of business and the immediate reaction along the social channels is “Whoa, Tulsky is at it again”, referring to their Vice President of Hockey Management and Strategy Eric Tulsky. For years Tulsky was a writer with a heavy analytics lean whose work is still regularly linked when you need to make a point about player aging curves and the like. Many believe he’s the Canes’ shadow GM pulling strings behind Don Waddell, and outside of him the only two numbers-related hires I see with Carolina are Kevin Kan (creator of the now-defunct DataRink, a great resource in its short time) as a Developer and one other Data Engineer. In the past year or so, the Canes have nabbed analytics darling Dougie Hamilton from the Flames, fleeced the Wild in the Niederreiter-Rask trade, added Jake Gardiner for relatively cheap to an already-stacked blueline, and are apparently now on the verge of roasting the Ducks in a Faulk-for-Kase swap.
And it isn’t just the Canes, the Colorado Avalanche have continually made an effort to bolster their analytics group as well, and it shows pretty obviously in their decision-making.
Where it doesn’t quite seem so obvious is in Toronto. While the Leafs have made substantial-to-earth shattering moves in the last year or so with the Muzzin and Tavares additions, they haven’t clearly burned anyone in a trade or brought aboard anyone that would suggest “Wow, they clearly used the numbers to get an edge here”. Again, this is fine, and I’m not suggesting the front office as a whole is underperforming; The team is better and more solidly set-up for the future with some of the moves they’ve made like the Zaitsev swap and the dead money LTIR moves. But those are cap moves, more attributable to Dubas and Pridham’s skill in navigating the CBA than anything else.
I think it’s fair to wonder if someone like Tulsky would bring a team into a Cup contention year with Cody Ceci set to play meaningful minutes due to depth issues down the right side. And yes, I get it, that deal was totally good and much-needed for the Leafs to get out from under future money. But a lot of people have pined for the Leafs to go different avenues to flip Ceci again after that, and it never materialized – they actually gave him a $200K raise for some reason, suggesting they believe he will be a real contributor this season. So with that move, like a few of the Leafs’ moves since they created this bulky analytics group, we’re again left to appeal to authority and say “The Leafs have a lot of tools, they must see something in the numbers that we don’t”. I suppose it will come out in the results whether that’s the approach we should keep taking.