Reconciling Oscar Dansk’s early season struggles in Erie’s net

The Erie Otters don't have the worst defence in the Ontario Hockey League, by a certain measure, but they're pretty close. Much of this is in an attempt to determine why Oscar Dansk, the highly-touted Columbus Blue Jacket-select who was drafted 3rd overall in the CHL Import Draft, has had such a rough start to his OHL career.

Officially, Dansk is listed as having a 4-9-0-2 record with a 4.04 goals against average, though a more puzzling .894 save percentage, which would put him last in the OHL among starting goaltenders.

Unfortunately, though National Hockey League blogs have have fans dutifully count the number of scoring chances going for and against each team, the amount of quality shots recorded for and against each net is not a luxury many of us statistically-inclined junior hockey writers are not privy to.  All we have to work with are the shots for and against each team, and while the number of shots on goal usually gives us some indication of the amount of quality shots a team gives up, at the NHL volume its shown to be more of an indicator of overall volume and not by rate.

That's why save percentage tends to be the preferred metric of hockey statisticians. By way of example, a goalie who faces 20 shots may face 10 quality shots, and he'll stop 18 of those. He's judged to be the equal of a goalie who faced 40 quality shots, stopping 36. In the long run, it's been found that NHL teams and players will generally give up as quality shots at a similar rate, and this is true for teams as well. One of the common misconceptions about goaltender ability in the NHL is that teams will give up a higher rate of quality shots, making it tough on the goaltender. Using scoring chance data, it's been shown time and time again that over a big stretch of games this isn't really true.

However, in the OHL, it could be, and I'm quite convinced it is. In the three games I saw the London Knights play live at last season's Memorial Cup, I determined that they gave up scoring chances at a very low rate compared to the number of shots they allowed, compared to, say, the Kamloops Blazers, a team I followed extensively when I was covering the Western League.

In our example above, however, while the goalie who stops 36 of 40 shots is considered to be the equal of the goalie who stops 18 of 20, this doesn't mean they have the same value. For a team like the Erie Otters, Peterborough Petes or Sault Ste. Marie Greyhounds, who give up a lot of shots, their goaltenders are inherently more valuable since the outcome of any given game is more likely to be determined by them. When Soo goalie Matt Murray stops 87.5% of shots, the Greyhounds are 9-1-0. When he hasn't, the Greyhounds are 0-5-1. When Matej Machovsky of the shot-reducing Brampton Battalion stops 87.5% of his shots, the Battalion are 5-2-2-2. When he doesn't, the Battalion are a worse off, but not disastrous, 1-2-1.

So how to reconcile this as an analyst? Check out the chart below. Each blue dot represents a team, and other than the outlier Battalion, the more shots against a team faces, generally, the lower the teams' overall save percentage:

At least to me this indicates that teams that are worse off defensively this season in the OHL give up a somewhat higher rate of quality shots than what we've observed at the NHL-level. Alternatively, it could be due to goaltender fatigue, as well, but it's clear that goaltenders such as Murray and Dansk are playing at a clear disadvantage when it comes to shots given up per 60 minutes. (UPDATE - However the trend may also be indicative of a small sample. This graph here suggests that OHL goalies in the last three years have a smaller correlation between shots allowed and save percentage, indicating that the trend could reverse over a large number of games)

Here's a chart of starting goaltenders who are at the initial disadvantage when it comes to shots against. Note that all of them are below the league combined save percentage of .908:

Name Team Min SA SA/60 SV%
Michael Giugovaz Peterborough 633 422 40.0 0.908
Oscar Dansk Erie 892 566 38.1 0.894
Matt Murray Sault Ste. Marie 893 566 38.0 0.896
Garret Sparks Guelph 960 588 36.8 0.903
Mike Morrison Kingston 738 444 36.1 0.903
Jake Paterson Saginaw 868 505 34.9 0.901
Mathias Niederberger Barrie 846 486 34.5 0.926
Jaroslav Pavelka Windsor 854 490 34.4 0.902
Keegan Wilson Ottawa 713 406 34.2 0.904
John Gibson Kitchener 705 383 32.6 0.937

[LEGEND: MIN: Minutes played SA: Shots against SA/60: Shots against per 60 minutes played SV% save percentage]

Only Barrie's Mathias Neiderberger and Kitchener's John Gibson face the shots disadvantage, but have a higher save rate than the OHL average. Does this mean their accomplishments ought to be ranked differently? I'd think so. Gibson was the top goalie in the OHL in overall save percentage last year and is 2nd this season, despite facing the disadvantage given the Rangers' struggles to start the season.

I looked back a little to see some future comparables for Dansk, as in, how did goalies facing his situation (drafted high in the import draft and the NHL draft, came over immediately, struggled in the first season) fared later in their junior careers, but unfortunately, there just weren't enough matches.

As much as I hate to say it, we'll take the "wait and see" approach for Dansk. Hopefully for his sake, the Otters improve in front of him. Even if his save percentages don't improve with a better team on the ice in front of him, it may mask his mistakes.