RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

[Ralph Baer]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

The goalie should be pulled to optimize results. Plain and simple. A one goal loss or a two goal loss is the same thing, and having more skaters at your disposal with the puck in the opposition's end is supposed to be a net benefit, that's why power plays exist. This is a no brainer.

Are you then advocating pulling the goalie as soon as you fall behind, even if it is in the first period? I am sure that you are not. If you pull the goalie too early, you are also more likely to give up a goal, and I am pretty sure that there are more empty-net goals scored than extra-attacker goals. There is an optimal time to pull the goalie based upon scoring probabilities. I do think that the optimum time is before the one-minute left that most coaches seem to like.

[FlagDUDE08]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

You're being intentionally fatuous here. It's obviously a late-game ploy, and you know why.

Yes, but I believe it fully explains the point to which Ralph is trying to make. When does "late-game" take effect?

[Ralph Baer]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

Yes, but I believe it fully explains the point to which Ralph is trying to make. When does "late-game" take effect?

Exactly.

[hrbrendan]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

It's a game where losing by two or three goals is no worse than losing by one goal. Unless ranking systems start taking into account margin of victory it will stay that way. Time decreases, the odds of scoring 5 on 5 decrease and the negative impact of the opponent scoring decreases. As that negative impact of the opponent scoring decreases, the necessity of a goalie decreases, so why not toss a guy out there on offense? Traditionally its done at the 1:15-1:30 mark for a 1 goal game, and 2:15-2:30 for a two goal game. I doubt there have been many studies on this done, as to do so would require a group of coaches to go against the status quo for an extended period of time.

[Red Cloud]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

I'm not going any further with this. Ralph, you ought to be ashamed.

[lugnut92]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

There is a big long mathematical study (PDF here) that basically says that pulling the goalie will average more points in the long run than not. It also claims that when down by a goal, coaches should pull the netminder at 3:00 or maybe even earlier, as opposed to the current (1:00 remaining, or ~1:30 for 2 goals) strategy. Note that this study uses NHL point values, so overtime losses are valued at a point.

I'm pretty sure this got posted last year when we gave up a lot of ENGs in a row.

[ibanezist00]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

Are you then advocating pulling the goalie as soon as you fall behind, even if it is in the first period? I am sure that you are not. If you pull the goalie too early, you are also more likely to give up a goal, and I am pretty sure that there are more empty-net goals scored than extra-attacker goals. There is an optimal time to pull the goalie based upon scoring probabilities. I do think that the optimum time is before the one-minute left that most coaches seem to like.

This is exactly what I was getting at, and what I intended with my original statement. Couldn't have said it better.

[burgie12]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

It's a game where losing by two or three goals is no worse than losing by one goal. Unless ranking systems start taking into account margin of victory it will stay that way. Time decreases, the odds of scoring 5 on 5 decrease and the negative impact of the opponent scoring decreases. As that negative impact of the opponent scoring decreases, the necessity of a goalie decreases, so why not toss a guy out there on offense? Traditionally its done at the 1:15-1:30 mark for a 1 goal game, and 2:15-2:30 for a two goal game. I doubt there have been many studies on this done, as to do so would require a group of coaches to go against the status quo for an extended period of time.

There have not been in-game studies (for obvious reasons), but there have been computer simulations. The simulations established in this paper (PDF) are only valid for NHL scoring and penalty rates. The conclusions drawn can only be said to be true in a general sense, not in a specific sense, for NCAA hockey.

With that disclaimer, their simulations show that pulling the goalie with three minutes left (or earlier if trailing by more than one goal) would earn an NHL team approximately an extra 1.0-1.5 points per season.

PS This same argument and this same paper was put in one of last season's threads. As long as we continue to trail late in games, Appert is going to continue to pull the goalie and we will continue to have ENGs scored against us. Consider it another fact of life.

[FlagDUDE08]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

It's a game where losing by two or three goals is no worse than losing by one goal. Unless ranking systems start taking into account margin of victory it will stay that way. Time decreases, the odds of scoring 5 on 5 decrease and the negative impact of the opponent scoring decreases. As that negative impact of the opponent scoring decreases, the necessity of a goalie decreases, so why not toss a guy out there on offense? Traditionally its done at the 1:15-1:30 mark for a 1 goal game, and 2:15-2:30 for a two goal game. I doubt there have been many studies on this done, as to do so would require a group of coaches to go against the status quo for an extended period of time.

...you mean like the UND women's team from 2007-08 playing without a goalie for practically the entire third period?

Actually, RC made a good point in that pulling the goalie could be easily equated to the existence of a power play. The two biggest differences are that the defending team may not clear the puck without running the risk of icing, and that the attacking team's net remains wide open. Perhaps, knowing this information, we could make some inferences.

Here's one place where we could start the research: What is the average time it takes for a team to score a goal whilst on a power play? For the sake of argument, let's throw out any major penalties where a goal is not scored within 2 minutes and count it similar to an unsuccessful minor (unless we're trying to look at a small subset of teams, specifically ours), and restart the count on a major after a goal is scored. I believe this would give an idea as to the latest time you should expect to pull the goalie and, in the long run, score the necessary goal(s) to force overtime.

Because we are in a league where the number of goals is only considered in relevance to the other team's number of goals, it would seem mathematically prevalent to pull the goalie once you fall behind and are in the attacking zone. Obviously you then have to take possession time, as well as time spent per zone, into consideration.

[Ralph Baer]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

Thanks lugnut and burgie. I don't recall that we discussed this last season, but it doesn't surprise me.

[FlagDUDE08]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

Thanks lugnut and burgie. I don't recall that we discussed this last season, but it doesn't surprise me.

I believe we went through 5 threads. Therefore, we discussed a lot.

[burgie12]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

Thanks lugnut and burgie. I don't recall that we discussed this last season, but it doesn't surprise me.

The conversation started on Post 666 of last season's Part II thread and I posted the paper on Post 687 and the conversation continued onward through nearly Post 800 from there.

[FlagDUDE08]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

The conversation started on Post 666 of last season's Part II thread and I posted the paper on Post 687 and the conversation continued onward through nearly Post 800 from there.

Dang, you have a better memory than me. Perhaps it does go with age.

[DrDemento]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

Dang, you have a better memory than me. Perhaps it does go with age.

It is the second thing to go.

[FlagDUDE08]

Re: RPI Hockey 2012-2013 Part II: You Better Starting Posting the Engineer Way

It is the second thing to go.

Dare I ask what the first thing is? ...