End Of Year OHL QualComp Estimates


With the Canucks long since having been eliminated from playoff contention, it’s been a bit of a spring malaise here at Canucks Army as we really haven’t had much to write about (other than the occasional spat of off-ice personnel news, though I’m sure we’ll have plenty of #hot #sports #takes when Jim Benning the next GM is hired). 

Regrettably, the next real thing to look forward to that involves actual hockey players is the draft which is more than a month and a half away. Fortunately, this down time gives us plenty of opportunities to break down the CHL in an effort to see just which 18-year old kid we all want to bear the burden of all of our collective hopes and dreams, and how the most recent drafted crop of Canucks prospects are doing.

As you all know, our own Josh Weissbock does tireless work tracking not only Canucks prospects through the season, but also a whole whack of other stuff for various North American hockey leagues. Among these things is Quality of Competition and Quality of Teammates, which we’ll look at for all OHL players past the jump.

Advertisement - Continue Reading Below

First off, some pretty critical notes and disclaimers:

  1. We’ve looked at three QoC/QoT measures here: Time on Ice (TOI), Goals For% (GF%), and a “Composite” measure. The composite measure is simply [TOI * GF%]. The theory behind TOI QoC is that coaches play their best players the most, and the theory behind using GF% or Corsi% or something similar is that the best players perform the best. The composite measure is simply an easy way of combining deployment and performance.
  2. All measures here are estimates. The Canadian Hockey League doesn’t track TOI or shot data, and only the OHL provides us with any sort of special teams data. Time on Ice can be estimated through the following formula: Average TOI per game = Player on-ice goals per game * Team average minutes per goal
  3. Since we know that goal-based data can be noisy and our TOI estimates also rely on goal data, the actual values of the numbers are less important than where a given player ranks relative to his teammates and/or peers. Also, be careful of small samples in terms of games played. Anything under 36 games played (half of one CHL season) is probably a bit sketchy.
Here’s the data (I believe it’s interactive so you can sort in any way you can in Excel):

Advertisement - Continue Reading Below

Stuff that jumps out immediately:

  • Bo Horvat faced easily the toughest competition among non-bottom feeding teams, which makes his strong draft+1 season more impressive. Not only did he increase his production by nearly 0.5 pts/GP, but he managed to do so while shouldering the heaviest burden on the Memorial Cup host London Knights. We don’t have enough historical data to know if this is at all meaningful, but it’s definitely another point in his favour. (I’ll also point out that I had a Bo Horvat jersey shirt before all of you)
  • The top end of the composite QoC list is dominated by Owen Sound and Plymouth, who both played in the same division as London (who are really really good), Erie (who are really really really good), and Guelph (who are really really really really good).
  • The top draft eligible prospects in terms of QoC are Sarnia’s RW Nikolay Goldobin and D Anthony DeAngelo. Sarnia had its problems just icing a full lineup this season, so it isn’t really surprising that they leaned on their two best players so much. Also, considering DeAngelo’s ridiculous stats, he may be the steal of the draft if he slides due to “character issues.”
  • Brendan Gaunce faced the most difficult TOI QoC and composite QoC on Erie, but this may be influenced by spending 1/4th of the season on a poor Belleville Bulls team.
  • Speaking of Belleville, Jordan Subban appears to have been the Bulls top “tough minutes” guy with overage partner Jake Worrad. Also, Jordan has scored more goals than his oldest brother did in his draft+1 year did while playing on a much worse Belleville Bulls team. Just sayin’… (P.K. exploded to score at just shy of a 100-point pace the very next year though, so it’ll be another full season before we can begin to say if the youngest Subban can even come close to following in his brother’s massive footsteps)
  • Cole Cassels also appears to have faced difficult competition but did so with better linemates, spending most of his draft+1 season riding shotgun with Scott Laughton and Michael Dal Colle.
Overall, this data set provides some context to the performances of Vancouver’s major OHL prospects, and for the most part says good things about them. However, to get drafted you pretty much have to be a standout player, and if you’re a standout player your coach is going to use you in critical situations. As a result, a high QoC in the CHL may not mean the same thing as a high QoC in the NHL.

Still, every bit of information we can get regarding prospects is a good thing as it gives us one more data point from which we can build an informed opinion. This data reflects particularly well on Bo Horvat. As I said earlier, he managed to see a huge jump in scoring this season which is very positive in itself. Given our best estimates, we can also say that he was probably one of the very best (if not the best) two-way forwards in the OHL this season due to the fact that Dale Hunter used him as the hockey equivalent of a meat grinder.

Of course, blog rules dictate that I can’t write a prospect post without a “Rhys craps on the Horvat pick” section, so here comes that much anticipated part: we don’t have enough historical data to be able to determine if estimated QoC at the CHL level means anything. Eric Tulsky concluded here that even in extreme examples of QoC-related deployment, full-season QoC measures had an almost negligible effect on an individual player’s production at the NHL level, so there’s not much reason to conclude that this is different for the CHL.

The moral of the story is to take everything about prospects with a grain of salt because there are so many unknowns and so much uncertainty. Horvat could be anything from Getzlaf to Glennie, from O’Reilly to O’Marra; we just don’t know. The best course of action is to assume he’ll fall somewhere in the middle and then adjust your opinion as new pieces of information become available. While we didn’t exactly learn anything new here, this information acts to confirm the “tough competition” narrative that has surrounded Horvat since his selection. All in all, this is definitely a feather in Horvat’s cap, but not enough to fundamentally change what we should expect from his NHL career.

Advertisement - Continue Reading Below

Josh and I plan to look at a variety of stuff this summer to see how far we (read: Josh) can dig into CHL stats. I also know that we plan to have the usual summer stuff here on Canucks Army (draft coverage, prospects series, and other news items as they happen), so keep a look out for that. 

Until then, you can keep yourself occupied with either the NHL playoffs or the Memorial Cup, which gets going this Friday with coverage on Sportsnet. The first game will see Bo Horvat’s London Knights take on either Anthony Mantha’s Val-d’Or Foreurs or Charles Hudon’s Baie-Comeau Drakkar, depending on who wins the QMJHL title later today. Enjoy the hockey.

  • Awesome post. Again, it’s hard to sometimes read too much or too little into major junior performance because the next stages are so qualitatively different, but then that’s what the assumptions that go into drafting does to some extent anyway. It would be interesting to see this data in terms of player age and to be able to track usage and deployment from year to year given the composition of teams and player development.

  • Cool stuff. The only thing I might have liked to have here is a column for age and/or draft year, so as to compare players against their same-age peers.

    Pretty interesting stuff. It will be cool to see how this plays out over a few/several years if you guys continue to do this.