Goalie Performance by Game State and Varlamov's Case for the Hart

[Hammer Time]

There's a thread on the main boards where Avs fans are making a case on Varlamov's behalf for the Hart. Some of them mentioned a "clutch" play from Varlamov. I decided to investigate this further.

In this investigation, I defined "clutch" situations as 3rd period or OT when the game state was either "leading by 1" or "tied" i.e. times where Varlamov was trying to prevent a tying or go-ahead goal. Results may be surprising.

From the beginning of the season through the Rangers game on April 3, 2014, Varlamov has faced 164 shots when tied in the 3rd period or later, and has been beaten 6 times, for a save percentage of 96.3%. When leading by one goal in the 3rd period or later, Varlamov has faced 207 shots and has been beaten 9 times (95.7%). Overall, in "clutch" situations Varlamov has a save percentage of 356/371 = .960.

I just did this quickly (copying from NHL play-by-play data manually) so I didn't do any PP/SH/ES breakdowns or the like. Does anyone know a source where data like this might be found?

Click to read more...

 

[soireeculturelle]

is the relationship persistent from season to season? if not it could be like shooting%, which varies randomly within a certain range of the player's true skill.

 

[Hammer Time]

is the relationship persistent from season to season? if not it could be like shooting%, which varies randomly within a certain range of the player's true skill.

I highly doubt it is. If a goalie really could sustain a .960 sv% in close, late game states, they would be the greatest ever, ahead of Roy and Hasek. It's much more likely this is just an aberration, but from the point of view of awards voting, it satisfies the criteria of "valuable to his team" very well. His play in these situations has probably been responsible for 5-10 points to the Avalanche.

 

[soireeculturelle]

this was an interesting article I saw recently which touches on a similar topic: http://www.sportsnet.ca/hockey/nhl/nhl-goalies-better-with-high-shot-volumes/

we know that score effect is a thing, and apparently goalies tend to post a higher sv% with higher shot volumes, so maybe varlamov's numbers a inflated a bit because the avs tend to have the lead late in the game and allow more shots. seems like he's playing great, in general, so that little bump in the 3rd period could be just due to that. tuuka rask, ryan miller (in STL), marc-andre fleury and corey crawford could be a few goalies to examine too, to see if that same phenomenon exists.

 

[Caeldan]

How does this compare to other netminders in the league?
I mean on its own, it sounds impressive... but perhaps an inflated SV% is the norm in close situations?

Don't teams tend to collapse and force initial perimeter (ie easier to save) shots in these situations -- goals then come off the 3rd or 4th rebound when they do?

I'd posit that sv% especially in this case is going to be highly dependent on team playstyle, are there enough situations with Varly's backup to at least make a comparison within the team?

 

[TheDevilMadeMe]

You mean "Varlamov's case to be Hart runner up." At this point, the only question for the winner is how close it will be to unanimous.

 

[hatterson]

It's certainly a surprisingly high number, but I'd like to see what other goalies do.

We already know from score effects research that save percentages are higher/shooting percentages are lower when teams are up. Presumably due to the trailing team throwing more bad angle shots towards the net in an attempt to score. So the real stat that matters is how much above "average" or other elite goalies is Varly's .960. And how many goals does that actually translate to given his workload.

Ie, if the expected it .950 then a .960 mark on ~350 shots is about 4 goals?

 

[Hammer Time]

Booth's goal tonight was the first game-tying goal conceded by Varlamov in the 3rd period since Jan 16 (New Jersey). He is still a .954 when leading by one and a .966 score tied in the 3rd period or OT. The latter number (score tied means no score effects!) is mind-boggling.

I calculated Rask's numbers for a comparison. So far he's .935 up by one and .929 score tied in the 3rd period or OT. And this is another Vezina candidate we're talking about. It looks like Varlamov is an outlier who has been responsible for a large chunk of Colorado's success this year.

Some of this may be the influence of Patrick Roy. It can't really be isolated because Giguère doesn't really play much.

If I have time after the season I might do a few more of these for different goalies, to get a better idea of what "most" goalies would be expected to do in these situations.

I doubt Varlamov will be able to sustain anything like this in the long term, but even he regresses next season he's still a much improved goalie from the past. Wonder what he would be like if Colorado had a top-notch defence. François Allaire's magic strikes again.

 

[Wintersun]

I don't think he'll win any of those two, but I think he deserves both the Vezina and the Hart.