On February 16, Devan Dubnyk was pulled in a game against Colorado after allowing three goals on nine shots. Nikolai Khabibulin replaced him and earned a win, stopping 16 of the 17 shots he faced. Since that game, Dubnyk has started just one contest, while Khabibulin has started three.

Is Dubnyk secure in the starting role, or could Khabibulin be stealing his job?

It looks right now like head coach Ralph Krueger’s going with a ‘win and you’re in’ approach, though we’re dealing with a small enough sample that this may not be the case. Khabibulin won against Colorado, and was rewarded with a start against L.A.; he lost that game so Dubnyk came in and lost to Minnesota; Khabibulin came back, beat Phoenix, and then got the start against Chicago.

The Chicago start was probably Khabibulin’s weakest of the year, and he was still quite good – Krueger said he played an excellent game in his post-game presser. All of his games have been good, as a quick look at his season statistics shows:

If this seems familiar, that’s likely because last season something very similar happened. Khabibulin collapsed down the stretch in 2010-11, and Dubnyk appeared to supplant him as the Oilers starting goaltender. Dubnyk earned the first start of the season, but was given very little rope in the starting role – Khabibulin excelled early, and then-coach Tom Renney turned the job over to him. Renney clearly entered the season with the idea of giving both goalies a shot, and Khabibulin was too good not to play.

From that October 20 win through Christmas, Khabibulin would start 19 games to Dubnyk’s nine, and his play warranted it. The problem was that Renney continued to give Khabibulin starts long after his play had slipped – Dubnyk won the starting job after Christmas and made 30 of the team’s final 48 starts – but Khabibulin still started 18 contests. The Oilers went 1-13-4 in Khabibulin’s starts; 16-11-3 in Dubnyk’s.

The difficulty was knowing when the music would stop for Khabibulin; his start was too good to be true and out of keeping with his post-lockout work. It’s the same problem Ralph Krueger will face if he continues to give Khabibulin starts – we know with relative certainty that Dubnyk is the superior goaltender to Khabibulin, so while riding the hot hand is fine to a point it’s playing with fire if it goes too far.

If Khabibulin had started five games rather than 18 after Christmas last season, and the Oilers kept up their average point percentage with Dubnyk over those 13 additional starts, it would have meant an additional 11 points for Edmonton. Ultimately, that would have worked against the team (no Yakupov, 12th in the West instead of 14th), but it would have put them 10 points out of the playoffs instead of 21. I don’t think it’s crazy to suggest that keeping Khabibulin in the regular rotation probably cost Tom Renney his job.

This marks the second season in a row that Khabibulin’s gotten off to a good start in net – is it possible that he is a legitimately good goalie, but that age and injury catch up to him as the season carries on? It’s a nice idea, because it suggests that Khabibulin still has some utility as a strong backup option as long as he isn’t overused by the coach.

Unfortunately, that seems likely not to be the case. In 2005-06, Khabibulin got lit up like a Christmas tree in October (3-7-0, 0.851 SV%) before settling down a little the rest of the way. In 2006-07, his October save percentage was 0.890. In 2007-08, it was 0.892. In 2008-09, it was 0.918 – this was his best post-lockout season, the year he finished with a 0.919 save percentage and convinced the Oilers he was worth a four-year contract. Moving to Edmonton, he got hurt early in 2009-10, and then in 2010-11 posted a 0.901 SV% in October and a 0.847 SV% in November.

In short: it just so happens that his hot streaks have coincided with the beginning of the season the last two years – it would be a mistake to believe that this makes Khabibulin viable as long as his playing time is kept low.

It’s great that Khabibulin has played so well for the Oilers, but everything he’s done since the 2004-05 lockout suggests that it’s going to stop, it’s going to stop soon, and when it stops he’s going to be bad. As Tyler Dellow points out in that linked piece above, Khabibulin has a 0.905 save percentage since turning 30, and we can probably expect him to be worse than his career average going forward given his age.

With that in mind, I’d play him in back-to-backs and otherwise keep him on the bench. Dubnyk is the much better option, the guy that is more likely to give the team a chance to win every night, and he should be started accordingly. One of the nice side effects of this decision is that it virtually guarantees Khabibulin won’t have a chance to see his save percentage drop – and at the deadline, given that he’s a Cup-winner, by all accounts a good, veteran, professional, and in the final year of his contract maybe, just maybe, he’ll be worth a draft pick from another team.

I certainly wouldn’t give him a chance to do to me what he did to the 2011-12 Oilers, and to Tom Renney.

To answer the title question: if Devan Dubnyk isn’t safe as the Oilers’ starter, somebody isn’t paying attention to recent history.