The most unbelievable goalie wins

It's not David vs. Goliath like some of the other examples posted here, but this one came to mind:

  • 4/18/2010, Gm3 of the WCQF - Colorado beats San Jose 1-0 in OT off the back of a 51-save shut out by Craig Anderson. The Avalanche only mustered 17 shots on Sharks goalie Evgeni Nabokov. Colorado won the game after San Jose defenseman Dan Boyle shot the puck into his own net in OT.
 
Craig Anderson probably has a lot of claims here relative to the amount of wins he's had. Anderson basically has two modes: Wayne Gretzky's 3D Hockey '98 brick wall mode, where he's literally impossible to score on. And then the ol': 25% rule, where he's giving up 4 on 16 shots through 2 periods and gets first dibs on the hot water...

I don't know if this is calculated out there...but since the big sleep, I can't think of another goalie who is just so unreal in one game and so poor in another. Andy's average game, I think is below average. But he has so many just outright thefts that seem to end up as 0 or 1 GA games, it's bizarre...
 
Craig Anderson probably has a lot of claims here relative to the amount of wins he's had. Anderson basically has two modes: Wayne Gretzky's 3D Hockey '98 brick wall mode, where he's literally impossible to score on. And then the ol': 25% rule, where he's giving up 4 on 16 shots through 2 periods and gets first dibs on the hot water...

I don't know if this is calculated out there...but since the big sleep, I can't think of another goalie who is just so unreal in one game and so poor in another. Andy's average game, I think is below average. But he has so many just outright thefts that seem to end up as 0 or 1 GA games, it's bizarre...

I do something like this; caveat that it's based on save percentage - I compare each goaltender's game against what an "average" goaltender would be expected to allow given the specific opponent and shots faced, and then give the goaltender a score based on how well they actually did against that expectation (standard deviation or z-score).

I then calculate a consistency measure which looks at how consistent those z-scores are during a season - for instance, if a goaltender is exactly average every game, they're very consistent, but if a goaltender is exactly two standard deviations above average, they're just as consistent.

Anyhow, by this measure, Anderson is about as consistent as an average NHL goaltender (not an average starter, but an average goaltender) - average is 1.00 here:

1654015319379.png


Values to the right are a different way to get at the question - how many "average" performances does a goaltender have (between -0.5 standard deviations and +0.5 standard deviations of the expectation), "below average", and "above average" performances.

Save percentage caveat goes here one more (although I do adjust mine for the specific opponent and season faced).
 
  • Like
Reactions: mrhockey193195

Users who are viewing this thread

Ad

Ad