Down 3 with 2 to go, would you pull the goalie?

[Taro T]

Didn't watch the 3rd period - had another commitment & already had heard the final score so didn't bother with it. So, due to not having any background on the momentum of the game at that point, avoided the poll.

 

But, in general, in a regular season game, there isn't a point to pulling the goalie down by 3. (Obviously, there could be exceptions to that, but generally, try to make it a 2 goal game & then pull the goalie if you get there.)

 

With a 3 goal cushion, there is no reason at all for the leading team to not try for the empty netter, and in general to that point they've been the better team (how'd you expect they got that 3 goal lead ;)), so the odds are better you'll be down by 4 rather than down by 2.

 

Like LTS said, nobody (except possibly schmucks like us posting on a message board on a Friday afternoon) is going to remember the particulars of the game but seeing the results and a 4 or 5 goal loss will look (& feel) a lot worse than losing by 2 or 3 would.

 

The players actually would probably prefer playing for a coach that doesn't go out of his way to make them look bad than one who sets them up for failure. Which is what playing for 3-6v5 goals is - a recipe for failure.

 

Have them pinch a D into the zone ES to create the overload, double shift your top offensive performers, but don't look for 3 6v5 goals in 2 minutes. For every 1 time it would happen, there'd have to be at least 19 failures.

[Thorny]

Didn't watch the 3rd period - had another commitment & already had heard the final score so didn't bother with it. So, due to not having any background on the momentum of the game at that point, avoided the poll.

But, in general, in a regular season game, there isn't a point to pulling the goalie down by 3. (Obviously, there could be exceptions to that, but generally, try to make it a 2 goal game & then pull the goalie if you get there.)

With a 3 goal cushion, there is no reason at all for the leading team to not try for the empty netter, and in general to that point they've been the better team (how'd you expect they got that 3 goal lead ;)), so the odds are better you'll be down by 4 rather than down by 2.

 

Like LTS said, nobody (except possibly schmucks like us posting on a message board on a Friday afternoon) is going to remember the particulars of the game but seeing the results and a 4 or 5 goal loss will look (& feel) a lot worse than losing by 2 or 3 would.

The players actually would probably prefer playing for a coach that doesn't go out of his way to make them look bad than one who sets them up for failure. Which is what playing for 3-6v5 goals is - a recipe for failure.

Have them pinch a D into the zone ES to create the overload, double shift your top offensive performers, but don't look for 3 6v5 goals in 2 minutes. For every 1 time it would happen, there'd have to be at least 19 failures.

This is all fine and dandy, but to the bolded: that's definitely true, but it's still more likely to achieve the winning scenario by pulling down 3 than not pulling. Even while admitting it's much more likely you'll go down 4 than nudging it down to 2.

 

It can certainly be debated if it's even worth it when down 3, but still.

 

And I'm agreeing with the last line as well. But trying to get 3 ES goals in that time would fail 20/20.

Edited October 27, 2017 by Thorny

[Taro T]

This is all fine and dandy, but to the bolded: that's definitely true, but it's still more likely to achieve the winning scenario by pulling down 3 than not pulling. Even while admitting it's much more likely you'll go down 4 than nudging it down to 2.

It can certainly be debated if it's even worth it when down 3, but still.

And I'm agreeing with the last line as well. But trying to get 3 ES goals in that time would fail 20/20.

Why would you not expect the goalie to be pulled sometime after a goal had been scored to cut the deficit to 2? Pretty sure that nobody arguing against pulling the goalie down by 3 in a generic regular season game would argue to not pull the goalie after the deficit has become manageable.

 

And, I'd expect that if all the data is culled, there is a better chance to win staying ES w/ a 3 goal deficit than damning the torpedos at that point.

[Thorny]

Why would you not expect the goalie to be pulled sometime after a goal had been scored to cut the deficit to 2? Pretty sure that nobody arguing against pulling the goalie down by 3 in a generic regular season game would argue to not pull the goalie after the deficit has become manageable.

And, I'd expect that if all the data is culled, there is a better chance to win staying ES w/ a 3 goal deficit than damning the torpedos at that point.

I can't see that being the case. Scoring 3 ES goals in a minute would be damn near impossible. You'd at least have an exceptionally slim chance at doing it with the extra man. You actually see teams score twice with the man pulled quite often.

 

As for scoring the first ES in a 3 goal deficit before pulling the goalie: ya, maybe that would work better, but scoring that first one is the problem. You are pulling the goalie because you need that goal(s) right away.

 

The hypothetical scenario involves being in a goalie pull situation down three goals with only 2 minutes left. Obviously it's better if at some point a goal is scored to decrease the deficit to 2 before you pull, but we are talking specifically about a situation with 2:00 minutes left. You can't wait around at that point for an ES goal to go in. You need 3 and only have 2 minutes.

 

The most likely way to score 3 goals is by pulling your tender at that time. It's just not very likely at all that you'll get that ES goal, within an already slim period of time remaining, where it leaves a team with enough time to score 2 more after that.

 

The same logic still applies: if you need a quick strike goal, and goals against aren't a detriment (you are going to lose anyways) a team's best bet is with a man advantage.

Edited October 27, 2017 by Thorny

[Taro T]

I can't see that being the case. Scoring 3 ES goals in a minute would be damn near impossible. You'd at least have an exceptionally slim chance at doing it with the extra man. You actually see teams score twice with the man pulled quite often.

As for scoring the first ES in a 3 goal deficit before pulling the goalie: ya, maybe that would work better, but scoring that first one is the problem. You are pulling the goalie because you need that goal(s) right away.

The hypothetical scenario involves being in a goalie pull situation. Therefore there isn't going to be much time left on the clock. Obviously it's better if at some point a goal is scored to decrease the deficit to 2 before you pull, but we are talking specifically about a situation with 2:00 minutes left. You can't wait around at that point for an ES goal to go in. You need 3 and only have 2 minutes.

The most likely way to score 3 goals is by pulling your tender at that time.

For the record, when I refer to staying at ES down by 3, I expect that at a 2 goal deficit that the goalie will in all likelihood be pulled.

 

And, I'd counter to the bolded that THE most likely outcome pulling the goalie down by 3 is losing by 4. ;)

[Thorny]

For the record, when I refer to staying at ES down by 3, I expect that at a 2 goal deficit that the goalie will in all likelihood be pulled.

And, I'd counter to the bolded that THE most likely outcome pulling the goalie down by 3 is losing by 4. ;)

Yes I understand that all would be in favour of pulling the goalie once down by 2. But my argument is that we are dealing with a hypothetical scenario here with specific parameters. If there is only 2 minutes left, you don't have enough time left to hope for that ES goal to go in, that also leaves enough time on the clock to get 2 more.

 

And yes I also agree that most likely outcome when pulling in that situation is going down 4, and I'm not arguing that. I'm saying that pulling is the scenario most likely to result in a win, not that it's more likely to succeed than fail. Two different things.

Edited October 27, 2017 by Thorny

[Taro T]

Yes I understand that all would be in favour of pulling the goalie once down by 2. But my argument is that we are dealing with a hypothetical scenario here with specific parameters. If there is only 2 minutes left, you don't have enough time left really to hope for the ES goal.

And yes I also agree that most likely outcome when pulling in that situation is going down 4, and I'm not arguing that. I'm saying that pulling is the scenario most likely to result in a win, not that it's more likely to succeed than fail. Two different things.

Pretty sure we'll have to agree to disagree. I expect that there is a better chance to win pinching in w/ the weak side D man to create an outmanned situation in the low slot while still at 3 goals down will give a better chance of winning than will getting that outmanned situation by pulling the goalie and allowing the opportunity for the EN goal.

 

It really shouldn't be discounted IMHO that trying for the EN goal has NO realistic downside to the team in the lead as even if they give up a goal on the ensuing faceoff (hardly a given) they STILL have a 2 goal lead w/ less than 120 seconds remaining. And if they get that EN goal, the game IS over.

 

And, if it DID actually increase the odds of winning, in our entry in the early years of the analytics era, wouldn't somebody, anybody (not named Roy) have tried it regularly by now?

[Thorny]

All we have to look at is the first goal.

 

Both scenarios involve 2 goalie-pulled goals needed to be scored after that first one.

 

What is the most likely way to get that first quick goal? I maintain it's with the extra attacker, and move forward from there.

 

Are the next two goals somehow more difficult to score because one empty net goal for already happened, relative to the difficulty of those 2 same empty netters had the first goal been ES? I don't think so.

Edited October 27, 2017 by Thorny

[Taro T]

All we have to look at is the first goal.

Both scenarios involve 2 goalie-pulled goals needed to be scored after that first one.

What is the most likely way to get that first quick goal? I maintain it's with the extra attacker, and move forward from there.

Are the next two goals somehow more difficult to score because one empty net goal for already happened, relative to the difficulty of those 2 same empty netters had the first goal been ES? I don't think so.

So, then when being down by 4 w/ 3 minutes to go, should the right call be pulling the goalie? What about when down by 5 or more?

 

If pulling the goalie down by 3 w/ 2 minutes to go is the right play, wouldn't pulling the goalie w/ 4 minutes to go be better?

[Thorny]

So, then when being down by 4 w/ 3 minutes to go, should the right call be pulling the goalie? What about when down by 5 or more?

If pulling the goalie down by 3 w/ 2 minutes to go is the right play, wouldn't pulling the goalie w/ 4 minutes to go be better?

The more goals needed in that set two minute time, the more necessary it is to pull the goalie. More need for quick goals, and less chance the ES goals or goals needed have the time to arise. Obviously there is a point of no return. You aren't going to score 5. If people think 3 is that point of no return, that's fine. I don't think it is, but that's not even what I'm arguing.

 

To the bolded: I'm not sure. The amount of time remaining is the key variable. Stretch it to say, 20 mins. Do you pull with 20 mins left? Of course not. Too much regular strength time remaining that may be captitalized on, and pulling the goalie is ALWAYS more likely to result in a goal against, than goal for.

 

But it's a desperate times, desperate measures strategy. It's not likely to work, but that's not the point. The point is it often gives a team it's only chance, no matter how slim. Would a team onside kick all game? No, but you attempt when you need to because it gives you a slim chance of success, where *the punishment for failure* is largely irrelevant: you were going to lose anyways. It's not worth resigning yourself to that likely failure earlier in a game when there is more time.

 

Do you pull with 4 minutes? I'd say no, at that point. I'm not sure where the time line would be.

Edited October 27, 2017 by Thorny

[Neo]

Points > Goals Against.

 

I would’ve.

 

There are probably some wild analytics around time, deficit and probability.

 

My gut in this game? Analytics would say pull with a snowball in hell’s chance that slightly exceeds the chance associated with not pulling.

Edited October 27, 2017 by Neo

[PASabreFan]

Taro, they did lose by 4. Are you saying this loss is going to linger? I'd guess it's already forgotten. Would it linger more if the fifth Columbus goal had been into an empty net instead of past a disinterested Johnson?

 

The psychological argument just doesn't hold water for me. Nor does saying the decision was OK because it was a regular season game, but you'd do something different in a Game 7.

 

I'm trying to think of the basketball or football analogy. It's hard to convert the time. Let's say there are 20 seconds left in a basketball game and one team has just scored a bucket to close to within 9 points. A three possession game. Of course the coach calls a timeout and tries to draw up a way of getting the ball back, or fouling, or whatever (not a big b-ball fan). No coach in that situation ever quits.

 

Or with 1:30 left in a football game, of course you onside kick it after getting to within 21 points. You wouldn't just blithely kick it off.

 

Quitting seems peculiar to the NHL.

[Thorny]

Points > Goals Against.

I would’ve.

There are probably some wild analytics around time, deficit and probability.

My gut in this game? Analytics would say pull with a snowball in hell’s chance that slightly exceeds the chance associated with not pulling.

Much more concise than me, and I agree.

[Taro T]

Taro, they did lose by 4. Are you saying this loss is going to linger? I'd guess it's already forgotten. Would it linger more if the fifth Columbus goal had been into an empty net instead of past a disinterested Johnson?

 

The psychological argument just doesn't hold water for me. Nor does saying the decision was OK because it was a regular season game, but you'd do something different in a Game 7.

 

I'm trying to think of the basketball or football analogy. It's hard to convert the time. Let's say there are 20 seconds left in a basketball game and one team has just scored a bucket to close to within 9 points. A three possession game. Of course the coach calls a timeout and tries to draw up a way of getting the ball back, or fouling, or whatever (not a big b-ball fan). No coach in that situation ever quits.

 

Or with 1:30 left in a football game, of course you onside kick it after getting to within 21 points. You wouldn't just blithely kick it off.

 

Quitting seems peculiar to the NHL.

How the f### is having your top offensive guys out on the ice and pinching a D-man up into the attack "quitting?" That is what I proposed upthread.

 

NOT intentionally converting a 3 goal deficit into what in most all probability is a 4 goal deficit is NOT quitting.

[Neo]

https://fivethirtyeight.com/features/nhl-coaches-are-pulling-goalies-earlier-than-ever/

 

I googled.

 

Oh, and for our True PhD and fledgling PhD!

 

http://people.stat.sfu.ca/~tim/papers/goalie.pdf

 

 

And, my favorite easy read.

 

http://thesportsquotient.com/nhl/2015/10/15/when-why-and-how-of-pulling-the-goalie

Edited October 27, 2017 by Neo

[PASabreFan]

How the f### is having your top offensive guys out on the ice and pinching a D-man up into the attack "quitting?" That is what I proposed upthread.

 

NOT intentionally converting a 3 goal deficit into what in most all probability is a 4 goal deficit is NOT quitting.

Well, Seth Griffith took the faceoff after he scored. So there's that. Defensible? Keep the "hot hand" on the ice? I guess. (That may be the only time "hot hand" and "Seth Griffith" appear in the same sentence.) Housley also didn't take his timeout when down by three, which also adds a whiff of quitting. Don't ruin your momentum? OK, maybe.

 

Your argument seems to rest on the idea that up by three, the Jackets would have fired at will (great Geico commercial) at the empty net without fear of icing. I'm not so sure. The instinct not to do that is pretty strong.

 

Anyway, didn't you suggest the Sabres would have pulled the goalie if it were a playoff game? Why? Wouldn't you always do the logical thing?

[Taro T]

Well, Seth Griffith took the faceoff after he scored. So there's that. Defensible? Keep the "hot hand" on the ice? I guess. (That may be the only time "hot hand" and "Seth Griffith" appear in the same sentence.) Housley also didn't take his timeout when down by three, which also adds a whiff of quitting. Don't ruin your momentum? OK, maybe.

 

Your argument seems to rest on the idea that up by three, the Jackets would have fired at will (great Geico commercial) at the empty net without fear of icing. I'm not so sure. The instinct not to do that is pretty strong.

 

Anyway, didn't you suggest the Sabres would have pulled the goalie if it were a playoff game? Why? Wouldn't you always do the logical thing?

1st off, as stated up thread, did not watch the 3rd.

 

In a non-elimination playoff game, would most likely not pull the goalie down by 3 w/ 2 left. The psychology doesn't work.

 

In an elimination game, almost definitely would (& probably would do it w/ about 2:20 left) to build on & feed off the desperation. The adrenaline would almost definitely be working to advantage at that point.

 

Moments are different; circumstances are different. Different circumstances don't necessarily call for the same solution (though they might; depends on the situations).

[Thorny]

1st off, as stated up thread, did not watch the 3rd.

In a non-elimination playoff game, would most likely not pull the goalie down by 3 w/ 2 left. The psychology doesn't work.

In an elimination game, almost definitely would (& probably would do it w/ about 2:20 left) to build on & feed off the desperation. The adrenaline would almost definitely be working to advantage at that point.

Moments are different; circumstances are different. Different circumstances don't necessarily call for the same solution (though they might; depends on the situations).

I agree with PA that we most definitely should have pulled, but this is a reasonable take.

[Weave]

Seems like a simple risk-reward choice to me.  And I'm not convinced there is any risk in the closing minutes of the game.

[French Collection]

I can't believe I read this whole thread.

 

There is merit to studying this and some analytics guy will come up with the proper odds of pulling the goalie at all specific times and deficits.

 

To me it boils down to a gut feeling by the coach, supported by some numbers.

 

Do you do it at random times during the game to give your team a boost? Risk the fans going home after the second period.

[PASabreFan]

I will confess to not screaming at my TV for Housley to pull the goalie. I was more in shock that Griffith had scored and pondering if he had celebrated a bit too much given the situation. But I want my coach to save my soul. I don't want him giving up down by three with two to go, even though I give up down by one with 30 to go.

 

I also don't remember if there was a push at all to score at even strength — I don't recall the personnel or if anyone pinched or what the disposition of the team was.

 

There are so many angles to this. I can imagine a coach letting his team stew in the mess it created, for example.

 

The real problem here is that 11 games into another season, yet again the games seem to be absolute must-wins.

[LTS]

https://fivethirtyeight.com/features/nhl-coaches-are-pulling-goalies-earlier-than-ever/

 

I googled.

 

Oh, and for our True PhD and fledgling PhD!

 

http://people.stat.sfu.ca/~tim/papers/goalie.pdf

 

 

And, my favorite easy read.

 

http://thesportsquotient.com/nhl/2015/10/15/when-why-and-how-of-pulling-the-goalie

 

Well, i hope the stats guys jump in after reading the paper.  Thank you for sharing it.

 

My questions that I have regarding the paper are that it seems to use seasonal scoring rates in situations and then apply those rates to a period of time when the players on the ice are not the normal players that are on the ice throughout the season.  So, while teams may score at a certain rate under normal circumstances the real measurement I'd like to see is how often they score when the teams ice their top players respectively.  We all can accept that the 5 guys a team puts on the ice to prevent a goal will be different than the 5 guys they put on the ice under normal shifts in the NHL.  The same holds true for the attacking team. 

 

Additionally, I'd like to see those numbers over multiple seasons.  Using a single season provides numbers are that a single output of a set of rules and tactics used by coaches and the NHL that season.  

 

Finally, there are no analytics for coaches who play a pure attack style in a 5 on 5 mode.  If I were down by 3 I would be more likely to never pull my defender off that blue line and allow the other team a breakaway on a goaltender.  In most 5 on 5 scenarios the team remains somewhat defensive even if its 1 guy who has in his head that "we can't give up a breakaway".  I'd rather see what 5 of my top guys can do when they throw defense out the window rather than skate 6 and still have to have one player who is thinking defensive because if the puck makes it past center ice its a free shot for the opponent.  (It's not a free shot in the breakaway).

[rakish]

PA, did you notice Ottawa scored 2 empty netters to tie last night?

[Sabres Fan in NS]

PA, did you notice Ottawa scored 2 empty netters to tie last night?

 

I was just about to post that.

 

http://www.cbc.ca/sports/hockey/nhl/ottawa-senators-new-jersey-devils-recap-shootout-comeback-1.4376543

 

Not exactly down 3, but they did score 2 goals late with their goalie pulled to earn a valuable point.

[bunomatic]

I was just about to post that.

 

http://www.cbc.ca/sports/hockey/nhl/ottawa-senators-new-jersey-devils-recap-shootout-comeback-1.4376543

 

Not exactly down 3, but they did score 2 goals late with their goalie pulled to earn a valuable point.

And thats something you can't teach in practice. There are situations in every game that players can learn valuable lessons from and gain real game experience. Why not go for it in that situation so the players are always prepared for it in the future ?