Beware of small sample sizes.
It’s something that we’ve probably heard ad nauseum by now, watching and covering hockey. We’re constantly reminded how wildly things can fluctuate in smaller groupings of game and then equalize over a larger sample, and warned not to jump to conclusions without taking the full picture into account. We’ve been reminded time and again, but perhaps we need to be reminded once more.
It’s old news by now, but earlier this week, the Flyers made a couple of cuts, and some of them came as a bit of a surprise. The roster, as it’s shaking out, looks like it will be featuring Carsen Twarynski and Connor Bunnaman, while players like German Rubtsov and Nicolas Aube-Kubel have been sent back to Lehigh Valley, and this wasn’t an outcome that many would have predicted at the outset of training camp.
And this, if nothing else, has us wondering about recency bias.
If you had asked me a month ago, as in fact someone did, if Twarynski had a chance of making this team out of camp, I would have listed his chances at very slim. Because, despite a strong training camp in 2018, his season with the Phantoms left much to be desired. Placed in a depth role, he struggled some to really put it all together on a consistent basis. He came out of the season with just 24 points in 69 games, and a 42.46 CF% (47 games tracked) and a 45.33 controlled entry% (17 games tracked).
He started to pull things together a bit more down the stretch, but nothing about that stat line screams “this player has mastered playing at the AHL level and is ready to make the jump.” More often, rather, he looked like Just A Guy on team that, greatly diminished by injury and callup, at times wasn’t terribly difficult to stand out on. In short, he still had some figuring out to do. This is a reasonable and understandable happening for a first year professional, but we do have to acknowledge that it happened.
Bunnaman, on the other hand, did have a much flashier season with the Phantoms, or, a flashy second half or so of the season. He was a key contributor and effective in his role down the stretch, with 17 of his 19 goals on the season coming in his last 49 games. But context is also important here. Because things started to come together for Bunnaman when he was moved up in the lineup and placed next to perhaps the team’s best purely offensive player in Greg Carey, while also being placed in the netfront position on a very effective first power play unit. And this isn’t to suggest that Bunnaman was purely a leech in these settings, but he was placed in the most favorable position, the best for him to succeed. And then, in the beginning of the season, when he was playing in a depth role more akin to what will be asked of him at the NHL level, he couldn’t put it together. So much so that the coaches weren’t confident enough in his play to make him a mainstay in the lineup.
We really respect the work he’s done this offseason with his skating and conditioning to really expediate his development, but all of that still leaves the open question of if Bunnaman would actually be able to drive a line, while also producing when not stapled to very strong offensive players. And that matters.
And, we should pause for a moment to acknowledge that Twarynski and Bunnaman are having good camps. Great, lights out, stellar camps? That’s debatable. But they’re doing well. And we can’t take that away from them, and it would be unfair to try.
But there’s a difference between being good in three or four exhibition games and being able to be effective at the NHL level over, if not a complete season, at least a larger sample of games.
What I keep coming back to is this quote from Vigneault, that after the second preseason game, that Fletcher had said that this was the best game he and Flahr had seen Aube-Kubel play. And, on the surface, this is a nice thing to say. They were pleased with his play. That’s good.
But the subtext there, to me, also screams “we didn’t see very many of his games in Lehigh.”
Because, while this was a very good game, he’s played better. Having watched all of his games with the Phantoms for the last season plus, I feel confident saying that. Best can be subjective, that’s important to note, but there have been several more games where his play was more complete, and he looked even more dominant. And these also came, in a larger context, mixed in with a two season sample of him being one of the team’s most consistent 5-on-5 scorers, while driving play well, per mine and Brad’s manually tracked data. It was a good camp for him as well—backed by a history of being productive at the pro level—that saw him doing everything but putting the puck in the net. And somehow it wasn’t enough.
And this, then, brings into question this whole evaluation process. How much of these players did management see last season? And how heavily are they weighing these performances against a brief preseason stint in, frankly, meaningless games?
To our eyes, the process seems to heavily favor performance in exhibition games, and that means that the process is flawed. The nature of the beast is that you do have to pay close attention to how players are faring in training camp, but you simply cannot erase a whole season’s worth of play because Reasons, in favor of letting the preseason tunnel vision win out. At best, it’s myopic, and it’s a good way to ensure that your on-ice product is not the very best that it could be.