The Myth of Shelter Island

For reasons I can’t quite figure out, the myth that Nikolai Khabibulin played tough opponents while Devan Dubnyk got carefully selected starts still has advocates around these parts. I don’t understand it.

Back at the start of February, I took a comprehensive look at the opposition Khabibulin and Dubnyk faced, and determined there was a marginal difference in the difficulty of the teams they were playing. I summed up that article this way:

In conclusion, Nikolai Khabibulin has faced very slightly tougher sledding than Devan Dubnyk, even when we allow for his greater percentage of home games, when both were healthy. This may show a tendency on the part of the coaching staff to shield Dubnyk, but the difference is so minimal that I’d argue they either aren’t doing it intentionally or they’re very, very bad at it.

And yet, the myth persists, so I decided to look at the problem from a few other angles.

Where Khabibulin Got His wins


  •  Home game vs. MTL
  • Home game vs. DAL
  • Home game vs. CBJ
  • Home game vs. TB
  • Home game vs. STL
  • Home game vs. FLA
  • Home game vs. CGY
  • Road game vs. TOR
  • Road game vs. CHI
  • Road game vs. CHI

The Stats

  • 3 of Khabibulin’s 10 wins (30.0%) came on the road
  • 2 of Khabibulin’s 10 wins (20.0%) came against Western Conference playoff teams
  • 2 of Khabibulin’s 10 wins (20.0%) came on the road against Western Conference playoff teams
  • 3 of Khabibulin’s 10 wins (30.0%) came when he allowed 3+ goals
  • 1 of Khabibulin’s 10 wins (10.0%) came when he allowed 4+ goals

Where Dubnyk Got His wins


  • Home game vs. VAN
  • Home game vs. CBJ
  • Home game vs. ATL
  • Home game vs. NYI
  • Road game vs. VAN
  • Road game vs. COL
  • Road game vs. NSH
  • Road game vs. PHX
  • Road game vs. SJ
  • Road game vs. MTL
  • Road game vs. ANA

The Stats

  • 8 of Dubnyk’s 12 wins (66.7%) came on the road
  • 6 of Dubnyk’s 12 wins (50.0%) came against Western Conference playoff teams
  • 5 of Dubnyk’s 12 wins (41.7%) came on the road against Western Conference playoff teams
  • 3 of Dubnyk’s 12 wins (25.0%) came when he allowed 3+ goals
  • 0 of Dubnyk’s 12 wins (0.0%) came when he allowed 4+ goals

The Summary

Dubnyk was more likely to win on the road, more likely to win against good teams, and more likely to win on the road against good teams. Khabibulin had a higher percentage of wins in games where he allowed 3 or more goals, and unlike Dubnyk he managed a win in a game where he allowed four goals.

Being “The Man”


Then there are those who argue that the shelter is more of a psychological issue related to games played than it is to the actual strength of Dubnyk’s opponents. The theory goes that Khabibulin was the veteran starter, with all the expectations, while Dubnyk was the young backup able to come in and play with less stress.

Logically, it’s an argument that makes no sense to me. At the start of this season Khabibulin had a three year contract for a lot of money, and there was no way the Oilers were going to demote him. Meanwhile, Dubnyk had to battle from Day 1 of training camp with Jeff Deslauriers and Martin Gerber for the backup job. Once Deslauriers was finally demoted, Dubnyk’s spot still wasn’t secure: the coach showed an obvious preference for Khabibulin, and any slip-up on Dubnyk’s part would have led to questions about whether he deserved an NHL job in Edmonton, and the possibility of following Jeff Deslauriers down to the AHL. I don’t see how that can possibly be less stressful than being signed for the next here seasons.

But let’s ignore that logic for a moment. Let’s say the extra games in front of a bad team really did shatter Khabibulin’s confidence, and that Dubnyk would have had the same problem. It’s a laughable concept because Dubnyk saw over 1000 shots, just 287 fewer than Khabibulin. But even so – how bad would Dubnyk have to be over those 287 shots to match Khabibulin’s save percentage?

The answer: atrocious. Worse than Vesa Toskala at his worst. Dubnyk wuld have needed to post a 0.798 save percentage over the next 287 shots to get down to Khabibulin’s level.

What This Is & What This Isn’t

First, what this isn’t.

This isn’t an excuse to beat up on Khabibulin. Yes, he was brutal, but it’s difficult to imagine that he isn’t better next year. As I wrote yesterday:

Barring further deterioration due to age/health, this season represents pretty much the floor of Khabibulin’s ability. In six seasons since the NHL lockout, Khabibulin’s save percentage has ranged from a low of 0.886 to a high of 0.919, with an average save percentage of 0.902. While not impressive, that’s better than 10 points above his save percentage this season, and it is at least possible that Khabibulin will play above his post-lockout average next year.

Secondly, this isn’t a wholesale endorsement of Dubnyk. Yes, he was excellent this season, but he’s a young goaltender with an uneven track record. As I noted at the end of February, I need to see more sustained success before I trust him.

So, what is my aim with this piece? Simple: to take the blinders off. Khabibulin had a bad year, Dubnyk had a good year, and the evidence suggests that the coaching staff did Dubnyk no favours – he earned his extra wins by being the better goaltender. As far as I’m concerned, denigrating Dubnyk’s performance because he wasn’t “the man” or based on a misguided notion that he got spotted against lightweight opponents is flat-out wrong. Dubnyk had a good year; he may not be able to repeat it but let’s at least acknowledge that his play was an unmitigated bright spot in a too-gloomy season.

  • Quicksilver ballet

    Devans one and only must win game this year was two weeks ago, the back end of the HNIC Saturday night special. 16000 strong, looking for blood that night, his teamates gave him a 4-1 lead five minutes into the third, Dubnyk let his guard down and thought he could coast through the rest of the game. When Calgary made it 4-3 we all knew it was over.

    One must win game (from the fans point of veiw) and he showed he wasn’t ready. He looks fine when the games don’t matter, but the ones that do to Oiler fans, well, he won’t last long sending the fans home disheartened like he did two weeks ago. The jury is still out on him and we still don’t know if he can win when it matters. He may have been better off playing 60 games in Oklahoma.

  • Sorensenator

    Kill the fist already. Its like buying wearing acid wash jeans from Saan store 10 years after they went outta style, it ain’t hip anymore. It had its day a year ago and it was cool and refreshing and everone loved it. Its like saying “Not, or “I am going to crush your head.” We need something new…

  • All right JW, you made your point. Doobie is good, Khabby is baddy. Can’t argue with all those numbers I didn’t bother to read.

    I just happen to think that if Doobie and Khabby have a good relationship, and Doobie feels comfortable with the process and is playing well, why take a chance on buying out Khabby, bringing in another b-rated goalie backup who may be no better than Khabby, and taking our chances that the goalies will perform well together. I’m not saying this is what you’re saying to do, just expressing my opinion on the matter.

    Good article. Appreciate you looking up all those numbers because I am wayyy too lazy to do it.

    • OB1 Team Yakopov - F.S.T.N.F

      Because if the team actually wants to start winning games next year it needs better then a .890 SV% from it’s back-up (let alone starter).

  • master of my domain

    it wasn’t that long ago that we were making similar arguements for Garon over Roli based on one year. I for one am not quite ready to run Khabi out of town and crown Dubnyk the new saviour.

  • Bucknuck

    Excellent article JW. I wonder what our record looks like if Khabibulin played more like DD this year. Still lottery of course, but I am not sure the Oil would be last.