NHL teams generally look at three factors to determine when to pull their goalie: the score of the game, the amount of time left remaining the game, and the location of the puck (i.e. if it is in the defensive zone, the goalie will not usually be pulled). It seems to me that two extra factors are needed:
- the exhaustion level of the opposing team’s five on-ice players
- the purpose of pulling your goalie
1. Exhaustion Level of Opposing Team’s Five On-Ice Players
Here’s a riddle: if your team was trailing by one goal, would you rather have the goalie pulled with 2 minutes left against a relatively well-rested defense or, instead, pulled with 3 minutes against a defense that is utterly exhausted as it is being caught on the ice during a really long shift?
There is no empirical evidence by which we can attempt to answer this riddle, because coaches almost never pull their goalies when down one goal with 3 minutes left. My guess, however, is that playing 6-on-5 against exhausted defenders with 3 minutes left may be better than playing 5-on-5 against exhausted defenders with 3 minutes left and then waiting until around the 2 minute mark to pull your goalie. Here’s why:
1) an exhausted defense is less likely to clear the zone and/or score an empty net goal
2) if an exhausted defense tries to score a long empty net goal and misses, resulting in an icing, then they will pay a big price for it: the other team will be able to bring on fresh players, which will make the difference in tiredness between the two teams even greater.
3) an exhausted defense playing 5-on-6 is less likely to get a lucky bounce or turnover that would allow them to clear the zone (or, if they do clear the zone, to clear it enough to get many of its players to reach the bench)
4) an exhausted defense playing 5-on-6 is more likely to have its goalie screened, so the odds of the goalie making a save to stop play and allow a line change is reduced.
5) by bringing a 6th attacker on the ice, you have the opportunity to seamlessly bring on a top player on who is fully rested himself.
6) pulling your goalie early means that the exhausted defense has less of a chance of winning the game by simply running down the clock. From a psychological perspective also, it may be more difficult for an exhausted player to muster his remaining energy when he knows he is not closed to being ‘saved by the bell’.
7) the exhausted players may not be that team’s best defenders; whereas with 1 or 2 minutes left in the game to play, a team normally has their best defenders on the ice. Moreover, if they cause an icing, you can bring on your own team’s best players
8) If the opposing team knows you might employ this strategy at some point during the game, they will be less willing to use their ‘coach’s challenge’ and so risk losing their time out. They will also be less willing to use their time out earlier in the game, even at times when they may need it. Your team gains an advantage by them being less willing to use their time out or coach’s challenge.
9) If the other team does manage to clear the zone and change lines, you can then use your own time out in order to rest your top line so that it can stay out on the ice for the rest of the game.
10) If you are playing a division rival or wild-card rival, and would like to deny them the chance of getting a point from an OT loss, this strategy gives you a (small) chance of winning the game in regulation
For all these reasons (some much more than others, obviously), I suspect that if you are facing a scenario where the opposing team’s line is exhausted with 3 minutes left and you are down a goal, you may be better off pulling your goalie then rather than waiting to do so with 2 minutes left against a better(-rested) line. If I were an NHL coach, I would try to simulate this scenario in practice during the offseason in order to try to answer this riddle. The reason I would run such an experiment is this: if it is true with 3 minutes left, what about with 4 minutes left? What about with 10? What if you were down by more than one goal? In other words, how exhausted do the opposing team’s players need to be, and much time left does there need to be, and how many goals down in the game do you need to be, to make this strategy worthwhile? We don’t know, as teams never try it.
We do know, though, that teams get caught out on long shifts fairly frequently. And we know that players’ effectiveness tends to drop dramatically when being caught on a long, tiring shift. So, if the strategy really were to prove effective, whichever team discovers it and implements it first may actually gain a significant advantage. (If it proved really effective, there may even be a case for waiting until the playoffs to deploy the strategy for the first time, in order prevent other teams from adopting the strategy themselves after seeing you use it). If successful, the benefit of simulating these scenarios in practice in the offseason could far outstrip the cost (of time and energy) that will be required to properly simulate the scenarios as required.
2. The Purpose of Pulling Your Goalie
We assume that the purpose of pulling your goalie must be to score a goal playing 6-on-5. But what about pulling your goalie to increase your odds of scoring a goal 5-on-5? Consider the following scenario: your team is trailing by a goal with 3 minutes left in the game, and is in control of the puck in the offensive zone. Some or all of your players on the ice are physically exhausted, and your best offensive player is on the bench. You would like to swap out one of your tired players to bring your star on the ice, but you don’t want to change on the fly because you are worried the other team might take advantage of the brief swap to try to gain control of the puck and clear the zone. Well, maybe you should think about pulling your goalie for a few seconds to bring him in, and then, once he joins his teammates in the offensive zone, have another player exit the game as quickly as possible so that your goalie can reenter the game. (This plan also works better if the players on both teams are tired, as at best they are only likely to get a chance to score an empty net goal from behind centre-ice, so they would be risking an icing). If done smoothly, you might be able to improve your odds of tying the game by trying this move.
Future Economics 