Quality in the Caps Crease

Ask any hockey coach at any level what he wants out of his goaltender and the response is likely to be some variation of "I want someone who will give the team a chance to win every time out." Simple enough.

Perhaps less simple is quantifying "giving the team a chance to win," but the good folks over at Puck Prospectus have given it their best shot by borrowing a baseball stat - the quality start - and adapting it for hockey. Basically, a Quality Start (QS) for a goaltender is one in which he has a save percentage of at least .920 or one in which he has a save percentage of between .885 and .919 and allows two or fewer goals. Now, given the Caps' prolific offense (lest you forget that it was just that prior to the last three games of the playoffs) a goalie can probably "give the team a chance to win" with a performance that doesn't quite meet those parameters (which may be a bit tougher than what you or I would use to describe "quality"). But we're going to stick with them nonetheless.

The number that becomes important, then, is a goalie's Quality Start Percentage (QS%), which is simply his quality starts divided by total starts. This past season, Tuukka Rask led the League (minimum 20 starts) with a 69.2 QS%, Jaroslav Halak rounded out the top ten with a 58.1% rate, and Chris Osgood's 38.1% was best among the bottom five (i.e. fifth-worst overall).

Other related stats are worth looking at as well. For example, Quality Start Wins (QSW) and Quality Start Win Percentage (QSW%) tell us how often a netminder is winning the games in which he's delivering quality starts. Wasted Quality Starts (WQS) and Wasted Quality Start Percentage (WQS%) tell us how often the team is losing when their goalie has given them a quality start (note - regular season overtime losses of quality starts aren't considered "wasted," as the team gets a point; we do, however, consider post-season overtime losses of quality starts to be wasted). Finally, Bail-Outs (BO) and Bail-Out Percentage (BO%) provide the frequency with which a goaltender fails to deliver a quality start, but he ends up with a win anyway.

With all of that in mind, after the jump we'll see how the Caps' three backstops have done over the past two seasons.

In his extremely limited time in the NHL, Michal Neuvirth has produced quality starts at a rate that would land him just outside of this season's top-10 goalies. He's also won an outstanding 90.9% of his quality starts, with the Caps only wasting his solid effort out in Hollywood just after New Year's this year. Neuvy has been bailed-out of only one of his eight non-quality starts (his second NHL start, in which he was actually quite solid), but then again he wasn't tagged with the loss in either of those rough games in Florida back in January of this year, so there's that. And while Neuvirth has sparkled in the AHL playoffs - twice - he has yet to test his mettle at the NHL post-season level.

Semyon Varlamov's NHL goals against average (regular- and post-season) is nearly three-tenths of a point better than Neuvirth's (2.51 to 2.80), yet Neuvirth has turned in quality starts at a better rate - 57.9% to 53.6%. What does that mean? That while Neuvirth has had a few games where he's been blown out of the water (see the eight goals he gave up in fifty minutes to the Panthers and Bolts), he's been more consistently "quality" than Varly. Varlamov has had a quartet of shutouts in his young career (Neuvirth has yet to register his first whitewashing), but also has had more than a few games in which he surrendered three goals or more (two dozen, to be exact, in those 46 NHL starts; Neuvy has seven in 19 starts). Neuvy right now is that relief pitcher who has one abysmal outing in April and spends half his summer bringing his earned run average back to respectability - and whose numbers don't accurately reflect how good he's been.

Back to Varlamov. 54.3% quality starts is a solid number - even if the Caps have wasted 16% of those starts, all in the playoffs (including, most recently, Game 5 against the Habs). Another 12% of the time, during this past regular season, the Caps have lost quality Varly starts after regulation, leaving his QSW% at 72.0%. But lest you feel bad for Varly and those seven quality starts in which the team didn't win or earn a point, he's managed to win eight times without turning in a "quality" effort, which is a generous bail-out percentage of 38.1% - a category in which the Russian netminder led the NHL in 2009-10.

Theodore improved considerably on his QS% in 2009-10 and even turned one in in Game 1 of the playoffs. But he still failed to deliver quality in more than half his starts and got bailed out pretty regularly by his teammates. Even during his fantastic regular season run in 2010 - when he went 20-0-2 in 23 starts - Theo was only "quality" 56.5% of the time (lower than Neuvirth's career rate). These numbers add some supporting evidence to what you may have already suspected - that JT60 might not have been as good as his record would seem to indicate.

Whether or not the Caps feel comfortable with the two kids in net heading into 2010-11 remains to be seen (and will be a dominant storyline in the months ahead), but what each has done in his brief NHL career so far would seem to indicate that they should, at least until proven otherwise. Semyon Varlamov and Michal Neuvirth, on most nights, are already quality netminders.

Now, we all know that what actually qualifies as a quality start varies from game to game. Was Varlamov's Game 7 loss in which he stopped 14 of the 16 shots he faced "quality"? How about Theo's 34-for-37 against the Wings? Neither was by this definition. But this is still a useful metric in terms of demonstrating which netminders are most consistently giving their teams a chance to win. And, after all, that's really all you can ask of a goalie.