Goalies and Garbage Time

Top Goalies Should Play More Games, But Fewer Complete Games (And Their Backups Should Be Better at Playing the Puck Than They Are)

NHL coaches treat their goalies like baseball pitchers from the 1800s: so long as they do not mess up, they get to play a complete game. Yet these same coaches also sit their goalies about one in four games on average, in order to give them rest. No goalie started in more than 64 games last year; no one has cracked 73 in a decade. Goalies sit out games even though, for some, the number of starts they get might be the difference between making or missing the playoffs.

Couldn’t these goalies start more games and get more rest by simply coming out of games once their team has built up an unassailable lead? An NHL team that is up by three goals going into the third period has a roughly 98% chance of winning*. Wouldn’t then, for example, be a better time to rest? Or what about a two-goal lead after the first period – giving you an estimated 80% winning probability – in order to allow the top goalie to start both games of a back-to-back ? Coaches would still have the option of putting the starting goalie back in the game if the score were to narrow. Backup goalies might even benefit too, since they would play on a more frequent basis (if only in relatively short bursts) rather than sit for a week or two between games.

*[Obviously, this is just a generalization and rough estimate. In reality it depends on many variables beyond just the score, the time remaining, and whether or not it is a home game. Good teams, especially good offensive teams, will have a better chance of coming back than bad teams, etc].

The counterargument, or prevailing wisdom, would, I guess, be that goalies would be thrown off their rhythm if they were to be used in this way – perhaps especially backup goalies, who would be getting fewer starts than they currently receive. Or, that a goalie resting on the bench for a period or two at a time is not nearly as rejuvenating as is taking an entire game off. There may be a lot of truth to these arguments, but I am still skeptical that they justify the current system wherein top goalies will play hundreds of minutes of “garbage time” every season (to steal a phrase from the NBA), while also sitting out for more than a thousand minutes of close games.

The current system is perhaps especially questionable when viewed in comparison to the number of games that forwards and defenseman are allowed to play consecutively. Though goalies obviously play many more minutes than their teammates, still it might seem wrong that 39-year-old forward Patrick Marleau has been able to play in each of his teams’ past 777 games (taking an ice bath after the second period of each game in order to physically do so), whereas one of the most esteemed goalies, Carey Price, has already sat out 17 starts this season, even though his Montreal Canadians are in a tight wild card race. And Marleau is not the only iron man now at large: Phil Kessel and Keith Yandle are both also at 750-plus consecutive games and counting.

Of course, goalies still play more minutes in total than their teammates. Last season, the league leader in minutes among non-goalies, the perfectly-named defenseman Drew Doughty of the LA Kings, played all 82 games while also leading the league in average minutes per game, at 26 minutes and 50 seconds. Doughty was on the ice for 2201 minutes in total, 1476 fewer than his goalie, the also-well-named Jonathan Quick, who ranked sixth in overall minutes despite only playing 64 games. Another one of their teammates, Anze Kopitar, led the league in minutes by a forward (trailing 31 defensemen) and also played 82 games, for a total of 1811 minutes on the ice.

The best player in the league, Connor McDavid, played more minutes (1767) than any forward other than Kopitar last year. He too played all 82 games. McDavid’s goalie, Cam Talbot, led the league in starts, with 67, but only ranked fifth in total minutes, as their team’s struggles meant he was often pulled from games. Neither of their efforts was enough for the Oilers to make the playoffs.

The goalie who led the league in minutes, Connor Hellebuyck of the Winnipeg Jets, with 64 starts, 3 backup appearances, and 3966 total minutes played, played almost twice as many minutes as Doughty’s 2201. He also had the most wins (44), and the best save percentage of goalies who played at least 60 games. (The Vezina Trophy winner Pekka Rinne, of the Nashville Predators, only played 59 games). Hellebuyck is again on pace to lead the league in minutes this season.

Out of curiosity, how does Hellebuyck compare with superstars in the NBA? Last season Lebron James, who has already played more career minutes than anyone else in the league, led the league in minutes per game and was one of only eight players to play all 82 games. Lebron was on the court for a total of 3025 minutes in the regular season, compared to Hellebuyck’s 3966. Since NHL games are more than 60 minutes long on average, whereas NBA games are shorter than 50 minutes, this means Lebron played close to as high a share of his team’s total minutes as did the NHL’s busiest goalie. By doing this he was able to carry a very bad team to the fourth seed in the East, on the way to the Finals.

Goalies might be wise to follow NBA stars in sitting out more during garbage time in order to play more during crunch time*. Sitting when your team is way ahead in the game is one way of doing this. But there is also the question of when to pull your backup goalie during games your backup goalie is starting. If your backup goalie starts a game and quickly lets in a bunch of goals, should you pull him to put the starting goalie – who is supposed to be getting a night off – back in? Even more interesting, if your backup goalie starts, plays decently, and the game is tied after two periods (for example), should you put in a Vezina-quality starting goalie into the game to play the decisive third period ahead?

*[This would especially be the case if goalies were more likely to get injured late in games when they are tired, or to experience an increased rate of wear and tear late in games when they are tired, or to get injured more in garbage time situations when the opposing team is desperately gambling for offensive chances in order to mount a comeback].

The Comeback Kid

 The most interesting implication of this way of thinking, however, is also perhaps the craziest; namely, the idea of having your backup goalie be just that: a backup-only goalie. More to the point, if your backup goalie is no longer actually starting many, or any, games – if, for example, even in the event of your starting goalie getting injured you rely not on your backup goalie to start, but instead call up your top prospect goalie who is getting regular starts in the AHL rather than languishing on an NHL bench – then what skills might you want for a backup goalie playing this new, more specialized role? One possible answer: have your backup goalie be a comeback specialist.

Since a backup goalie of this sort would be playing mainly, or only, at times when his team has either a solid lead (to let the starting goalie rest) or is behind in the score (because the starting goalie has been pulled), he should ideally be a goalie who is good at helping his team to mount a comeback. In other words, he should – all else considered – be exceptionally good at playing the puck when his team is behind in the score, particularly as his team becomes more desperate towards the end of games.

Such a goalie could be useful even in situations in which backup goalies currently do not play. Consider, for example, a situation in which your team is down by one goal, with five minutes left in the third period and an offensive zone faceoff following an icing by an opposing line that is tired from having just playing a long shift. Putting in your comeback specialist backup goalie at such a moment might be beneficial for a number of reasons. First, it would be more difficult for your tired opponents to dump the puck to get a line change in against a goalie who excels at passing the puck. Second, an aggressive puck-playing goalie might help your team score a goal in general during the remaining few minutes of the game*. Third, you would eliminate the risk of your starting goalie getting injured; a risk which would maybe be increased by your team gambling offensively to catch up, which could lead to more odd-man rushes and so, perhaps, injuries.

*[Also, if your backup goalie were later able to get to the bench for the extra attacker a second or two more quickly – or maybe even a half-second more quickly – than the starting goalie is able to (either by being a faster skater than the starting goalie or by being able to play further from the crease before heading to the bench than the starting goalie), this might help your team to score a 6-on-5 goal by making it more likely that the extra attacker will make it into the action before the opposing team has a chance to dump or clear the puck out of their own defensive zone.

…Indeed, to get even crazier here for a moment, what if this goalie were to sometimes attempt a back-and-forth strategy in late-game situations: for e.g., you pull the goalie with two minutes left when you are in the offensive zone, but, if the opposing team immediately clears the puck out of their zone, then you quickly put your goalie back in so that you don’t give up an empty net goal trying to gain reentry into the zone. If you pull your goalie early enough, you might perhaps be able to try this a number of times before the final minute of the game, so that your net might only be empty when your team is already in the offensive zone. Pulling your goalie early and temporarily might, at least, be a useful strategy to employ at times when the opposing team’s players on the ice are exhausted during an especially long shift].

Another example could be a short 5-on-3 power play when your team is down a goal late in the game. Putting in a goalie who can aggressively and excellently pass the puck could help your power play unit avoid wasting critical time on the 5-on-3, while also making it more difficult for tired penalty killers to get in a line change. To a lesser extent, this may also be useful in desperate 5-on-4’s.

A backup goalie’s puck-playing skills might also be well suited for the times when his team is well ahead in the score. If, for instance, the opposition begins to gamble more to create offensive odd-man rushes, passing opportunities for a skilled-passing goalie might open up. Or if the opposition becomes disheartened and begins trying to dump and chase more often, a puck-playing goalie might be able to help thwart some of these attempts. As such, a goalie who is used primarily or exclusively in situations when his team is either behind or ahead in games could perhaps possess puck-playing skills that would be useful in both of those types of situations.

Of course, this comeback-specialist-backup-only goalie plan might be a terrible idea. But the idea from which it is indirectly derived, namely that certain goalies should start more games than they do now, nevertheless appears to have merit. The question may not be whether some goalies should sit more in order to start more, but rather only who should do this and when should it be done.