clock menu more-arrow no yes mobile

Filed under:

Of Quality Starts and Stuff

How Optimus was Optimus Reim? How often did Reim Time occur? If we check the OR, will we like it so far (pardon the use of a line from a mid-90's Ottawa Rap Group featuring Tom Green... the letters just fell into place). Well, using the Quality Start stat proposed by Robert Vollman at Hockey Prospectus, we get a decent representation of how regularly James Reimer was posting decent numbers. Follow up after the jump to see how he compared to his peers this season.

In calculating his Quality Start statistic, Vollman only included goalies that made 41 starts this past season. Reimer only made 35 starts this season, so he didn't qualify for the list of NHL leaders. Despite finishing 12th in the NHL in Save Percentage, and posting a 2.60 GAA, along with a stellar 20-10-5 record, what other evidence can we provide to support the idea that he's a solid netminder? Do we even need to bother? \

Well, wins is a bizarre stat to use for goalies frankly. A win is a team statistic, and the data actually indicates that a starting goalie is MORE likely to win a game where he gives up over 3 goals than he is if he gives up exactly 3 goals. Obviously there's more to a team's win numbers than just the goalie, so it's not a personal stat worthy of notice.

Yes the team won with Reimer in net, but using that as evidence of his skill is a circular argument, and one I don't think is particularly worth making. Good goalies lose games on bad teams, and bad goalies win games on good teams... let's leave it at that.

So what do we use as a measure of goaltending? SV% is the best measure we have on a large scale aggregate level, but trusting it on a game by game basis isn't necessarily recommended either. So grabbing at an idea that largely gained traction in baseball statistics for starting pitchers, Vollman has described the Quality Start in goaltending terms.

To quickly summarize, a quality start is awarded to any goalie that posts a save percentage of .913 or better in a single game, OR if the goalie registers a save percentage between .913 and .885 but only allows 2 or fewer goals. This is to compensate for goalies that play in games where they face very few shots, and goals against skew their SV% down despite their relatively rare occurrence.

Using the total number of starts, and the number of Quality Starts, a goalie's QS% can then be calculated (QS/GS). Here is the ranking of NHL goalies from 2010-11, based on Vollman's calculations:

Goaltender GS QS QS%
Tim Thomas 55 40 72.70%
Roberto Luongo 60 41 68.30%
Pekka Rinne 64 43 67.20%
Sergei Bobrovsky 52 34 65.40%
Antti Niemi 60 39 65.00%
Jonathan Quick 60 39 65.00%
Jonas Hiller 46 29 63.00%
Marc-Andre Fleury 62 39 62.90%
Corey Crawford 55 34 61.80%
Ilya Bryzgalov 67 41 61.20%
Carey Price 70 42 60.00%
Ryan Miller 65 39 60.00%
James Reimer 35 21 60.00%
Cam Ward 74 44 59.50%
Craig Anderson 49 29 59.20%
Niklas Backstrom 50 29 58.00%
Miikka Kiprusoff 71 41 57.70%
Dwayne Roloson 54 31 57.40%
Kari Lehtonen 68 39 57.40%
Jaroslav Halak 57 32 56.10%
Martin Brodeur 54 30 55.60%
Michael Neuvirth 45 25 55.60%
Henrik Lundqvist 67 37 55.20%
Tomas Vokoun 57 31 54.40%
Steve Mason 53 27 50.90%
Jimmy Howard 63 31 49.20%
Ondrej Pavelec 54 26 48.10%
Jonas Gustavsson (09-10) 39 15 38.46%
Brian Elliott 51 17 33.30%
Nikolai Khabibulin 46 15 32.60%
Jonas Gustavsson (10-11) 21 6 28.57%

Just for sake of discussion, I've included the numbers for the Leafs other (formerly) future starter candidate, Jonas Gustavsson. Also note worthy is the fact that if Reimer had been given the Leafs starter's role earlier, the Leafs may have gotten the benefit of his 60% quality start rate far earlier.

If Reimer can produce the way he did this year over the next few seasons, the Leafs will be set pretty well in net from a starting perspective.