Projecting Ryan Miller

62206930

Like Gabriel Desjardins, I’ll be posting a series of pieces originally intended to go into an NHL annual this year.  Unfortunately we ran out of time before it could be launched, but most of the pieces are still relevant.  In this first one, I look at the odds of Ryan Miller repeating his Vezina-winning year.

When Ryan Miller was selected from the Soo Indians of the little known North American Hockey League late in the 1999 Draft, there was little reason to think he would ever play an NHL game, much less eventually win the Vezina Trophy. Most of his teammates would top out as NCAA players, with a few going on the lower tier professional leagues like the ECHL and UHL. Of the other players selected in the fifth round that year, only one – Niko Dimitrakos, now in Europe after two AHL seasons – would go on to play more than 100 NHL games. Fully three-quarters of the other fifth rounders have yet to appear in even one game.

Yet, Miller did win the Vezina, and his accomplishments from last season are impressive. His 41-18-8 record and his 0.929 SV% put him among the league’s statistical leaders, though it may be that his Olympic performance was even more impressive. Miller recorded a 0.946 save percentage for the silver medal-winning Americans, and was named both the tournament’s best goaltender and its most valuable player.

The question now is whether Miller can repeat his Vezina-worthy performance. At just 30 years old, there’s little reason to believe his physical ability is on the decline, but on the other hand, 2009-10 was far and away his best season.

To try and answer that question, we will break down Miller’s season in a few different ways and try to isolate items that might not be sustainable.

Month By Month

One problem with looking at overall numbers is that sometimes a hot start or a hot finish can make an average goaltender look like a world-beater. To use a recent example, in his Calder Trophy season, Steve Mason enjoyed a brilliant stretch from November through December (12-7-1, 0.938 SV%) which was followed by a long run of mediocrity to close out the regular season (21-13-6, 0.905 SV%). Did Miller have a similar hot streak?

Table 1: Statistical Results By Month

Month

Record

SV%

Shutouts

October

8-1-1

0.936

1

November

7-3-1

0.938

2

December

6-4-1

0.927

1

January

8-3-2

0.938

1

February

1-3-2

0.901

0

March

9-2-1

0.921

0

April

2-2-0

0.925

0

Miller did indeed have a hot stretch to start the season, but given that the hot stretch covered four months and the majority of his games, this isn’t a comparable situation to Steve Mason. With the exception of a slight lull over six February games, Miller was quality for the entire season.

By Game Situation

Another tactic we can use to get a better look at a goaltender’s overall season is breaking his performance down by game state. The vast majority of shots a goaltender faces come either at even-strength or while his team is killing a penalty, and as a rule the penalty-killing save percentage is far lower than the even-strength one: after all, on the power play the opposition has a better chance to convert on their shots.

valign="bottom" width="85">

2006-07

Table 2: Save Percentage Broken Down By Game Situation

Season

Games

EVSV%

PKSV%

OVSV%

2002-03

15

0.901

0.920

0.902

2003-04

3

0.804

0.667

0.795

2005-06

48

0.919

0.899

0.914

63

0.928

0.846

0.911

2007-08

76

0.915

0.866

0.906

2008-09

59

0.927

0.879

0.918

2009-10

69

0.928

0.919

0.929

Career

333

0.922

0.880

0.914

It doesn’t take much mathematical knowledge to discover where Miller was outperforming last season. At even-strength, he was just slightly better than his career rate – allowing 7.2 goals per 100 shots rather than 7.8. But on the penalty kill he was vastly better, allowing just 8.1 goals per 100 shots, where historically he has allowed 12 goals per 100 shots. To paraphrase a sports cliché, when killing penalties Miller was giving 150%.

To put it another way, in 2009-10 Miller led all NHL starters in penalty-killing save percentage. In 2008-09 he was 11th, in 2007-08 he was 25th, and he was worse than that in the two previous seasons. But in 2005-06, Miller’s 0.899 SV% tied for the league lead (along with Rangers goaltender Henrik Lundqvist).

Part of the reason for the incredible fluctuation from season to season is the small shot totals faced. At even-strength, Miller routinely faces around 1500 shots against (1690 last season) but on the penalty kill the number fluctuates between 300 and 400 – meaning that a goal against on the penalty kill has almost five times the impact on the PKSV% number as a goal at even-strength does on the EVSV% number.

Another possible reason for the fluctuation is that there’s often a wide disparity between the quality of special teams units, both from season to season and from team to team. Breaking down Miller’s appearances by the team he played against, we see something interesting: he played half of his games against teams in the bottom third of the league’s power-play rankings, one-third of his games against teams with a top-10 power-play, and the rest against the other nine teams in the middle.

But as fun as it is to speculate on why Miller’s penalty-killing numbers were abnormally high this season, it seems sane to wager that he is in for some kind of regression next season. The last time Miller led the league in penalty-killing save percentage (in 2005-06) he followed that up with the worst performance of his career (one that narrowly beat out Andrew Raycroft, Fredrik Norrena, and the ghost of Ed Belfour to avoid being the worst number for any starter).

There is little reason to believe that Miller will continue to post such phenomenal numbers on the penalty-kill (his 0.919 SV% this season was the best number of any starter since 2001-02) and significant reason – based on his career numbers – to suspect regression. Obviously, Miller’s still a phenomenal goaltender, but he’s more likely to finish around the 0.915 mark next season than to post another 0.929.

Comments (2)

  1. [...] Canucks: Engine And Anchor Posted by Jonathan Willis under Analysis on Oct 31, 2010 As with the Ryan Miller article yesterday, this article was originally intended to go into an NHL annual.In forty years of NHL play, the play [...]

  2. [...] coach.  The loss of Jason Pominville to injury and inconsistencies across the board have stung, Ryan Miller has not looked like the Vezina Trophy winning goaltender he was a season ago, although he’s [...]

Leave a Reply

Your email address will not be published. Required fields are marked *