Time on Ice - Goaltenders

Scott Reynolds
November 06 2012 07:12AM


Photo by Michael Miller via Wikimedia Commons

There are a lot of things that can happen to a team over the course of an NHL season that will result in really poor results. As a fan of the Edmonton Oilers, I know that injuries and ineffectiveness can really cripple a team. Every now and again you might end up with Jeff Deslauriers as your regular starting goaltender, Sebastien Bisaillon taking a turn on defense straight out of junior, or Ryan Potulny among your team's leading scorers. There's no doubt some bad luck mixed in when this kind of thing happens, but in some cases, it's probably also bad planning. Over the next several days, I'm going to take a look at how many players NHL teams have used at each position in a given season in order to provide a more concrete idea for what's reasonable as far as adversity. I begin today by checking on goaltender usage.

In order to answer this question, I decided to use each of the last seven seasons, which provides 210 seasons in total, and should give us a pretty good idea of recent trends. Since each team plays a slightly different number of minutes (because of overtime and empty-net situations), I measured each goaltender's playing time both in total regular season minutes and as a percentage of his team's regular season total. Here's a look at the league averages:

In the chart above you can see that one team used at least seven goaltenders, three used at least six, and so on. The average starting goaltender played just 66.2% of his team's minutes, which I found surprisingly low, but the top two goaltenders played 93.5% of the minutes on an average team, which seems about right. The really surprising part for me was the number of teams using three or four goaltenders in one season: about two thirds used three goalies and about a quarter used four.

This isn't, of course, all the result of injuries. Many of the teams using at least four goalies brought one of them in via trade, often because the options available to them internally were underperforming. This would be the case for Dwayne Roloson's arrival in Edmonton in 2005-06 as well as his arrival in Tampa Bay in 2010-11. As such, I think it's reasonable to suggest that you shouldn't need to plan on using your fourth goaltender for an extended period of time. On the other hand, teams probably should be comfortable with their third goaltender seeing significant minutes at some point during the season (or if you're the Capitals, at some point during the playoffs).  

Another question that this analysis raised for me was around team effectiveness based on goaltender usage. Did teams who used one goaltender most of the time have more success? In the chart below, I've shown how various teams performed in terms of standings points depending on what percentage of the team's total minutes their top goaltender (measured by ice time) played.

Unsurprisingly, the most successful teams in the league had one goaltender playing most of the team's minutes. The fall-off is pretty small until you get below 60%, at which point teams take a huge hit. There are, of course, exceptions to the rule (e.g. the 2011-12 St. Louis Blues), but that's the kind of thing that happens to a team because of either injury or ineffectiveness. In that it's rare to have a team begin the season with a commitment to play two goaltenders in roughly equal measure, this isn't the kind of statistic that will help a team with poor goaltending (playing your bad goalie 80% of the time instead of 55% or the time isn't going to help), but it does show the importance of coming into the season with a solid plan between the pipes.

Previously by Scott Reynolds

 

889688bd2efc9140bec34eeba34eacde
Avatar
#1 Eric T.
November 06 2012, 10:32AM
Trash it!
0
trashes
+1
0
props
this isn't the kind of statistic that will help a team with poor goaltending (playing your bad goalie 80% of the time instead of 55% or the time isn't going to help), but it does show the importance of coming into the season with a solid plan between the pipes

And of course, having a solid plan doesn't necessarily correlate perfectly with how many games the goalie actually gets.

A team might try to get more starts for a goalie who has been playing well, so the 70-start-guy label might be a post-facto designation. Consider Phoenix -- their goalie situation was unclear and up for grabs heading into the season last year and they split time early in the season. But then Smith went on a run where he started 11 of 12 games and had a 94.1% save percentage, he ended up getting 67 starts. On the flip side of the Phoenix goalie situation, Bryzgalov was terrible for most of the season and as a direct result ended up playing his fewest games since he became a starter.

Anyway, I think your work does mesh with the idea that a top goalie is worth a couple of standings points relative to an average goalie. If anything, I think people would be surprised at how little difference there is between a 50-start guy and a 70-start guy -- people tend to assume that differences between goalies are much larger than they are.

Comments are closed for this article.