As far as I can tell, all the research on pulling the goalie early only compares pulling the goalie with a minute left, or earlier. I haven’t located any research that compares pulling the goalie with not pulling the goalie. Presumably, teams pressing for a tying goal, might actually occasionally score at even strength if they didn’t pull their goalie. But pulling your goalie will result in your opponent scoring about three times as often as it results in your team scoring, and your chances are only about 10% anyway. The trouble is, everybody pulls their goalie, so there’s no benchmark to compare it to. (Except in Russia, apparently, where it’s not done.) The infamous study that leads to the misconception.
So I am sticking to my theory that it makes just as much sense to pull your goalie right from the start of the game (or any time you have a face off in your opponent’s end) as it does in the last minute. In other words, it does not make sense.
I wish someone would try it.
And we all know why no one will: because the fundamental equation does not work.
The same logic applies to “no-doubles defense” in baseball. This idea is simply silly. There is presumably an optimal location in the outfield for all the fielders during normal play. This position has been arrived at through years of experience and analysis: where is the ball most likely to be hit. The goal is to maximize the chance of catching it and making an out. The essential goal is to stop the other team from advancing a runner to home plate. If you put the fielders in the wrong place, you might prevent more doubles but you would allow more singles, which will result in more runs being scored. For eight innings of baseball, everybody agrees with that wisdom.
Suddenly, in the 9th inning (and sometimes sooner), all this analysis and wisdom is out the window and, instead, we move everyone back to make sure that nobody gets a double. This obviously increases the chances of a hitter getting a single, because you’ve opened up space between the infielders and the outfielders. And the idea is that decreasing the chances of making an out on a hit that might now be a single is worth the possibility that you will hold the batter to one base, instead of two. So, you get something like 1.3 singles instead of one double. So you now have runners on first and third when you might have had two outs instead.
I haven’t seen a good statistical analysis of this idea yet but it really isn’t necessary. It’s a logical problem. Is it logical to increase the chances of a hit and reduce the chance of an out in order to reduce the chance that the hit will be a double? No. Is it logical to believe that moving the outfielders back from their optimal positions on the field is an advantage in the late innings? If it is, then why is it not an advantage in the first inning?
Think about it– is there any logical reason why it would not also be an advantage early in the game? It doesn’t really matter when the other team scores their runs, as long as they score more than you.
What is understandable is this: the manager has to manage. What else is he going to do in the late innings of a close game? Get a glove and join the outfielders?
Just listen to the commentators! I just saw one last night on the issue of pulling the goalie in the last minute. Mike Babcock, after the Leafs already gave up an empty net goal, making it 4-2 for the Carolina Hurricanes, hesitated to pull the goalie a second time. One of the “analysts” said, “I wondered, why the hell is he not pulling the goalie!”.
Mike Babcock is certainly smart enough to know that pulling the goalie only marginally increases your chances of tying the game if you are one goal behind. The chances that you would score 2 goals is ridiculously small.
But it would take a truly remarkable coach to defy commentary like that and do something different.
So for the foreseeable future, we will be stuck with an NHL that refuses to entertain the idea that pulling your goalie is a bad strategy.
Incidentally, last year, Tampa Bay Devil Rays tried the novel approach of having a reliever start the game, pitch an inning or two, and then bring in a “starter”. I love the fact that somebody has the guts to at least try something different. How did they do? Very, very well.