How bad are Corey Crawford's playoff problems? (Trending Topics)

On Sunday, Corey Crawford had the ignominious job of sitting there with a cap on and maybe sometimes opening the door for his teammates, while a relative unknown got the start and the win for Chicago to push the team to a 2-1 series lead over Nashville. Then he had it again Tuesday. Then last night as well.

Not an ideal turn of events, to be sure, but one you might have seen coming for a while here.

This is very much Joel Quenneville making a decision based on recent play, but rather than going with the “hot hand” as they say, he's simply going with the not-cold one. And look, he's been 100 percent awful in the two games he started, allowing nine goals on 47 shots (.809) which is and should be good enough for a coach of Quenneville's quality to go to the bullpen.

This is, as noted by James Mirtle, part of a trend this year. Cheaper, less experienced backups are supplanting the highly paid playoff veterans with plenty of success in their playoff pasts. Maybe you could expect that, statistically speaking, because the difference between good and bad goaltending at this level is so small to begin with (the average elite goaltender will only allow about 15 or 20 fewer goals on a few thousand shots versus one who's abjectly poor), and because in the playoffs you deal with such a small sample to begin with. Especially when you're losing.

Ahead of last night's games, 23 goalies had played postseason minutes for the 16 teams that qualified for the playoffs. The teams that have used two goalies are Washington (Braden Holtby was sick, not pulled; a .943 save percentage is worth keeping around), Ottawa (the wheels inevitably coming off for Andrew Hammond), Calgary (Karri Ramo got mop-up duty in Game 2), Vancouver (Eddie Lack was Eddie Lacking), Detroit (Jimmy Howard flat-out lost his job), Minnesota (Devan Dubnyk crashing back to earth), and obviously Chicago.

What's interesting to me about Crawford, though, is that he is a slightly above-average goaltender in this league, demonstrably so, over a period of about five years. Career .917 in the regular season, career .917 in the postseason. That's reliability to some extent. He saves a lot of goals versus the league average, and yet he has a reputation as a not-great goaltender. It's a view I share, but there's not a lot of fact-based evidence to support the argument, so I found myself wondering why this is.

What has Crawford done that was so bad to warrant a lot of people thinking, “Well this is a waste of $6 million a year,” other than play on a good team? After thinking about it for a bit, it seemed to me that it's a similar problem which plagued Antti Niemi both in Chicago and San Jose. Here we have a goaltender who is consistently very good in the postseason (career .917, just like Crawford) and at least in Niemi's case there's the .907 postseason number to give you doubts. Not so with Crawford.

Here is a chart of his regular-season save percentage versus the approximate league average over the course of his career.

That looks to me like a pretty good goaltender who had a rough go of things a few years ago but has mostly spent the majority of his career well above water. Chicago may be good and everything, but for me in the regular season he's in a tier not too far below the league's best.

Put another way: He'd have to turn in a great season by his standards to even warrant Vezina consideration (it's not out of the question but it would take some serious work), but this is a guy who, by this merit, should be worth about $6 million annually.

But then the playoffs roll around, and things are a little less rosy. Part of the reason for that is that in the postseason, there are fewer bad goaltenders, and thus the league average rises sharply. Guys like Crawford who are perfectly good in the regular season quickly become a little below-average.

Okay that's a lot more worrisome.

Now, I'd argue he's still spending most of his time north of that line, of course. And the season in which Chicago won the Cup, Crawford went an incredible .932 in 23 games. That alone earned him his current contract (by making his bosses so enamored of his play in much the same way Marc-Andre Fleury did so many years ago for the Penguins). Besides that, though, the performances have been less impressive, with only one postseason breaking .915, and it was a series in which he went .927 but Chicago was eliminated in the first round in 2010-11.

The quote-unquote Book On Crawford is that his glove hand is weak, but if you shoot high glove most goaltenders are going to give up goals. If the Book On Crawford was to shoot low, that would be a different story entirely, because goalies who give up low goals in addition to high goals are just flat-out bad, but it's not, so that's not an issue.

What you have to also keep in mind is that Crawford has played just 58 playoff games, compared to 268 in the regular season, and so when he's bad, which any goalie occasionally is, it stands out more. If you have trouble for a few days and try to play through it, you might blow two games, and all of a sudden you're down 2-1 in the series and someone else has your job.

A stat I think is interesting is Rob Vollman's RBS —Really Bad Starts — which looks at the, like, truly and deeply bad starts goaltenders churn out. To qualify for such an awful stat, you need to have an individual game save percentage of .850 or less. Crawford has suffered 30 of them in his NHL regular-season career (almost half of which came in his disastrous post-Cup season; 13 RBS in 57 games). That's 30 out of 268 (11.2 percent), about 1 every 9 starts. By contrast, Henrik Lundqvist's career number is 56 in 620 games (9 percent) or 1 in every 11 starts. This is, again, what separates elite middling in this league.

But again, once the playoffs roll around, Crawford has difficulties, perhaps due to improved quality of competition, or something else. His RBS rate jumps to 9 in 57 (15.7 percent), or 1 per 6.5 starts or so. But Lundqvist's, too, goes up, to 10 in 83 (12 percent), or about 1 in 8.

Goalies are going to have bad starts, and apparently at increased rates in the playoffs. It's inevitable. The big newsflash here is that Scott Darling isn't a .940 goaltender. The numbers he's posting here and now dramatically outstrip his career highs at almost any level in his career. And you have to wonder if this is just a case of Chicago not being able to pick up Crawford when he stumbles as they have in the past.

Building on the RBS stat, I also went back and looked at times when he posted save percentages of .900 or less but Chicago still won (Kinda Bad Starts?). It's happened 23 times out of 92 in the regular season (25 percent), and just 3 of 17 (17.6 percent) in the playoffs. Again, I'd relate that back to quality of opponent more than anything else — it's harder to overcome bad goaltending against good teams — but you have to do it if you want to win the Cup.

Here's what you can expect from Crawford in any given game — regular-season or playoffs — across his career. As you can see, he gives you roughly league-average to stellar play more than half the time. Factoring in the nights when he's a little bit bad, you're looking at two-thirds of every appearance. You might even get some winnable games in that orange “Kinda Bad” range. Only about 1 in 7 starts is truly and deeply awful.

In the playoffs, things are a little different. The percentages of red and orange games actually shrinks overall, to 29.3 percent (based on 15.5 percent and 13.8 percent, respectively) from 33 percent. And for the record, another 13.8 percent each are in the meh-but-winnable yellow and good green categories, leaving — hey, whaddaya know? — 41.4 percent for the blue games in which he's great.

The point of this is not to excuse Crawford, because his his play has been inexcusably poor this postseason. He earned the benching he got this year. But relying on Darling as any sort of goalie of the future is foolish, and you'd be much wiser to play Crawford for the balance of his contract than try your hand with Darling, who's getting the Andrew Hammond love-in treatment right now. (And who gave up four goals last night for a .857 save percentage -- not a Really Bad Start, but a bad one.)

If Darling wins them a series or two, that's great. But Crawford generally gives them a better chance to win a lot more than that over the next few years. Gotta play the odds there.

Ryan Lambert is a Puck Daddy columnist. His email is here and his Twitter is here.