Over the weekend I had a discussion on Twitter about whether the Predators should have played Carter Hutton in Winnipeg and Pekka Rinne in Chicago. Rinne got the shutout win Friday night in Winnipeg while Hutton played an impressive game in Chicago that ultimately ended in a 2-1 overtime loss. Rinne posted a 31 save shutout while Hutton finished Saturday's game with a .943 save percentage on 37 shots. In Hockey Abstract, Rob Vollman defines a quality start as any game in which a goalie posts at least a .917 save percentage when facing greater than 20 shots. Both goalies, then, posted quality starts and that's all we can ask of them.
The discussion on Twitter however focused on which goalie should start which game. I made the argument that Rinne should face the tougher opponent in Chicago while Hutton should be put in against a weaker Winnipeg team. The opposing opinion, lead by fellow On the Forecheck writers Anish Patel and Dan Bradley, was that Rinne should play in Winnipeg to virtually guarantee a win against an easy opponent. The thought process here is that you take the two points against a weaker opponent rather than risk losing both games by playing Hutton against Winnipeg. Essentially, playing Hutton in Chicago makes sense because that's a hard game to win even with Rinne in. Take the guaranteed points when they present themselves. While this idea makes sense in theory, looking at goalie numbers shows this may not be the correct way to view goaltender performance, especially in single games.
Save percentage, although simple, is the most accepted metric for goalie evaluation and comparison. As I have pointed out before, @Hawerchuk wrote on Arctic Ice Hockey that we don't truly know a goalie's skill until he faces 3000-5000 even strength shots. Knowing this, it's safe to say we don't truly know what Hutton's skill level is. The sample size is much too small to say that he is significantly worse than Rinne. But still, the comparison continues.
Ignoring the fact that Hutton has only faced 877 shots against at even strength (before Saturday's game), the difference between Hutton's and Rinne's save percentages aren't too far apart. At all situations, Hutton has a save percentage of .910 while Rinne's is at .919. Think about that for a second. That's a difference of 9 goals every 1000 shots. That's the difference between being a backup scrub and Vezina worthy. Eric Tulsky at Broad Street Hockey found that 6 goals equates to one win. So, every 1000 shots, Rinne is worth about 1.5 more wins than Hutton. Keep in mind that we are using all situations data rather than 5on5 or even strength. Generally, all situations data isn't a great measurement because it includes penalty kill and power play totals, but it is a good evaluator for single game comparison. It's hard to say that in a 30 shot game we would be able to actually come to the conclusion that Hutton is significantly worse than Rinne.
Now that we know the difference between Hutton and Rinne isn't as large as it appears, there is still the argument that Hutton has the tendency to give up high goal totals every once in a while when Rinne tends to be much more reliable. By simply looking at the variance of each goalie's save percentage and goals against for every game they started, we can see if this is true. Based on this assumption, Hutton's career variance should be larger in both cases, especially when it comes to goals against per game.
|Goalie||SV% Variance||Goals Against Variance|
As we can see, Hutton not only doesn't have a greater variance on SV% or goals against, his is slightly lower, however insignificant. So it's not really plausible that Hutton is a "riskier" play.
Finally, we come back to this weekend's games. Rinne played outstanding and won the game for Nashville. It's hard to say that Hutton didn't play just well though. Sure, the goals he allowed were not pretty, but Rinne has also given up his share of bad goals. No goalie is immune to giving up bad goals. As previously mentioned, both goalies played well enough to earn quality starts, and that's really all that matters.
When it comes down to it, the argument that Rinne is a guaranteed win over a lesser opponent is wrong, but so is assuming the Predators don't stand a chance against a top opponent with Hutton in net like I was arguing. When we talk about average level goaltending compared to above average goaltending, we are talking about a difference of about 5 to 10 goals over 1000 shots or about 1 to 2 wins. It's obvious the Predators have a better chance of winning with Rinne in net, but that difference is generally negligible in a one game sample. Goaltender play is important, and stellar play from Rinne may just be the catalyst to get Nashville back into the playoffs, but let's think about the numbers before we throw Hutton under the bus.
*Shoutout to Jake Sundstrom at Fear The Fin for writing an article that was an inspiration to this post.
*Career SV% courtesy of Hockey Reference
*All other stats come from War On Ice