Pulling the Goalie: Hockey and Investment Implications

[troutman]

https://www.tsn.ca/the-math-on-pulli...nders-1.465010

[troutman]

Strategies for Pulling the Goalie in Hockey
http://people.math.sfu.ca/%7Etim/papers/goalie.pdf

[Salt Water Cowboy #10]

What if the team is brutal with a man advantage? Guys squeezing their sticks because they know they should be able to score, but their fragile minds play with their confidence. I’ve seen it happen. Also, what if they dump it in, but can’t retrieve the puck because they lack a decent forecheck?
It’s a serious gamble pulling the goalie. I’d rather see the best players double shifted and try and score. Then with less than a minute left, pull the goalie and go for it. GG pulled the goalie early a few times this year. Didn’t really work out that I can remember.

[CSharp]

Originally Posted by CaptainYooh

2. All NHL coaches are groomed to prefer losing while making proven decisions than winning while making unproven decision.

I can attest to that 100% of the time for the Flames at least for the last 10 years. But the initial rebuild years should be how it's done if you want a shot at a generational once-in-a-lifetime player.

I can guarantee that's what the Oilers were doing for the past 20+ years

[CaptainYooh]

Originally Posted by Salt Water Cowboy #10

What if the team is brutal with a man advantage? Guys squeezing their sticks because they know they should be able to score, but their fragile minds play with their confidence. I’ve seen it happen. Also, what if they dump it in, but can’t retrieve the puck because they lack a decent forecheck?
It’s a serious gamble pulling the goalie. I’d rather see the best players double shifted and try and score. Then with less than a minute left, pull the goalie and go for it. GG pulled the goalie early a few times this year. Didn’t really work out that I can remember.

See, this is the line of thinking that the article is attacking so hard. Coaches are not pulling the goalie earlier, only BECAUSE they are afraid to do. All the data collected from the actual 6-on-5 playing time accounts for the factors you've mentioned and suggests that they should do it sooner anyway, because the probability of the reward is higher than the risk.

[mrkajz44]

Interesting read. I think the whole article can be summed up as follows:

You should be pulling your goalie much earlier than previously thought because losing by 5 goals is the same as losing by 1 goal.

It's counter intuitive to most people's gut feeling as losing 2-1 seems better than 6-1, but its really the same outcome. So you should gamble more and get more blowouts, but you'll be better off as those gambles will result in more games that you actually tie it up.

[Mattman]

Originally Posted by mrkajz44

Interesting read. I think the whole article can be summed up as follows:

You should be pulling your goalie much earlier than previously thought because losing by 5 goals is the same as losing by 1 goal.

It's counter intuitive to most people's gut feeling as losing 2-1 seems better than 6-1, but its really the same outcome. So you should gamble more and get more blowouts, but you'll be better off as those gambles will result in more games that you actually tie it up.

Going off that... how would a 6-1 blow out (for arguments sake let's say 3 games in a row) vs losing 2-1 (3 games in a row) affect the moral of the players, the fans, and the other team? Would you be just as willing to let the Oiler's have a 6-1 blow out over your team just the same as a non-rivalry team like the Florida Panthers?

[R0taryRocket]

Near the end, the paper discusses some of the qualitative factors that go against this model based purely on quantitative analysis. The author goes as far as calling their model shortsighted, "only caring about maximizing the points" per game.

[CaptainYooh]

Originally Posted by R0taryRocket

Near the end, the paper discusses some of the qualitative factors that go against this model based purely on quantitative analysis. The author goes as far as calling their model shortsighted, "only caring about maximizing the points" per game.

Yes. Also, in there, they talk about how pride (player's, owners') and fan's disappointment with a 5:1 loss over 2:1 loss are factors that the mathematical model simply cannot account for.

[mrkajz44]

I honestly don't think moral would change that much. Right now, losing 5-1 means the other team fired 5 shots past your goalie. But if you started to pull the goalie much earlier, losing 5-1 might just mean the other team scored 3 empty net goals. The team would know it was actually a close game and the score is not reflective of that.