Goalie stats for playing in relief

[SotasicA]

I've seen goalies come in relief and play great. Anywhere I can find collected stats for goalies when playing in relief?

I am especially interested in start vs relief stats.

1. How a goalie's stat line in relief compares to his "games started" stat line.

2. League-wide, how does games started compare to played in relief?

3. Who is the best goalie when playing in relief?

4. Overall, would it be feasible to use multiple goalies per game as a strategy? Only hard stats for or against, no opinions on practicality.

 

[NHL WAR]

I've seen goalies come in relief and play great. Anywhere I can find collected stats for goalies when playing in relief?

I am especially interested in start vs relief stats.

1. How a goalie's stat line in relief compares to his "games started" stat line.

2. League-wide, how does games started compare to played in relief?

3. Who is the best goalie when playing in relief?

4. Overall, would it be feasible to use multiple goalies per game as a strategy? Only hard stats for or against, no opinions on practicality.

Download Statistics - Hockey Abstract

Download the goalie data, it has relief stats. I am interested to see what results you get.

 

[SotasicA]

Thanks. I filtered a basic comparison. 2017-18 goalies with their "games started" save percentage compared to their relief save percentage. A positive difference means they've played better statistically in relief than when starting a game.

I used a minimum of 40 shots faced, but the data is still quite limited. Only 30 goalies faced that many shots in relief.

You can see there is a slight negative correlation. 17 out of 30 goalies had better stats when starting a game.

Craig Anderson faced 53 shots and only allowed 1 goal in relief, making him the best performer. Carter Hutton allowed 11 goals on 69 shots, making him the worst of everyone included.

 

[SotasicA]

I guess this puts away my planned super system of swapping in a rested, focused goalie every TV timeout.

 

[Michael Farkas]

Meh, maybe not, at a glance, there's probably a good eight or so goalies that are not NHL caliber and a few more that are fringe guys based on their talent level...

Though, changing goalies in a game has been done...I think Rangers did it with Chuck Rayner and Sugar Jim Henry, IIRC...

 

[Doctor No]

Yep, a few teams have tried shuffling goaltenders in the NHL; most recently, I believe, was with Ron Low and Bernie Wolfe on January 8, 1977. The Capitals lost, 7-2, but the 1976-77 Capitals against Montreal were probably going to lose anyhow, no?

Anyhow, one thing to consider when doing this is that you'll get a skewed distribution for goalies playing in relief, since the vast majority of times goalies appear in relief it's because the starter got drilled (and that happens disproportionally against stronger opponents).

I have this data (back to 1967-68) and could split it, including the opponent strength, although it will take some time since work is busy.

 

[SotasicA]

Anyhow, one thing to consider when doing this is that you'll get a skewed distribution for goalies playing in relief, since the vast majority of times goalies appear in relief it's because the starter got drilled (and that happens disproportionally against stronger opponents).

This is true, and also why GAA is probably a bad stat to use in this case.

Why my eye test makes it seem a goalie in relief is playing better, is because they are replacing a goalie who probably wasn't having a good game. So it's all relative.

Findings here don't support a better performance coming in mid-game, but it's true quality of shots is higher in blowout games.

Funny timing for this "study". The Hawks - Sens game just saw all four goalies already in the first period, including the Official Relief Champ Craig Anderson.

 

[morehockeystats]

This is true, and also why GAA is probably a bad stat to use in this case.

Why my eye test makes it seem a goalie in relief is playing better, is because they are replacing a goalie who probably wasn't having a good game. So it's all relative.

Findings here don't support a better performance coming in mid-game, but it's true quality of shots is higher in blowout games.

Funny timing for this "study". The Hawks - Sens game just saw all four goalies already in the first period, including the Official Relief Champ Craig Anderson.

Often a relief goalie performs well because the game is already decided, and both teams are sort of mailing it in.