Is Miller better than Halak?

Kevin Hoffman-USA TODAY Sports

The data suggests that Miller is about the same as Halak. Even if he were better, come playoff time it wouldn't matter.

You don't have to look too hard on NHL.com to find comments like

The Blues are second team in the league in shots against per game and have the second best offense in the game and if they were to acquire ryan miler in a trade they are gonna be as tough as anybody to beat in the playoffs. 

Is Miller really better than Halak?  Would swapping Miller for Halak lead the Blues to the Stanley Cup?


Career, Regular Season, Even Strength

Miller has faced 11648 shots and made 10745 saves for a 0.9225 ESS%.
Halak has faced 4904 shots and made 4535 saves for a 0.9248 ESS%.

Edge Halak, but realistically those save percentages are the same.

Last 3 years, Regular Season, Even Strength

Miller has faced 4056 shots and made 3748 saves for a 0.9241 ESS%.
Halak has faced 2448 shots and made 2262 saves for a 0.9240 ESS%.
Those save percentages are obviously the same.  That's one goal difference in 10,000 shots.

Over the last 3 seasons, Miller has played a lot more than Halak, 167 games to 119.  Some of that is Miller's durability (or Halak's lack thereof).  Some of that is the teams' willingness to play the backup.  Who would you play more, Elliott or Enroth?

This season, Even Strength

Miller has faced 844 shots and made 784 saves for a 0.929 ESS%.
Halak has faced 519 shots and made 477 saves for a 0.919 ESS%.
Edge Miller, but 5 more saves for Halak would have them equal.  The difference is no where near statistical significance.

Career, Playoffs, Overall

Miller has 47 GP, is 25-22 with a 2.46 GAA and a 0.917 SV%.
Miller last won a playoff series in 2007.
Halak has 23 GP, is 10-11 with a 2.42 GAA and a 0.923 SV%.
Halak last won a playoff series in 2010.

Does better goaltending translate into playoff success?


Nope.  And I don't just mean some existential "It's the Blues and they will find some new way to crush my soul again this Spring" answer.  When I looked at PDO and Playoff Success, I found that Regular Season Save Percentage is minimally predictive of Playoff Success.  About 2.5% of the variability in Playoff Win Percentage is explained by Regular Season Save Percentage.

The problem is that a 7 game series is a really small sample.  Miller and Halak have both faced about 30 shots a game in the playoffs.  Roughly 25 of those shots are at Even Strength and 5 are on the PK.  So in a 7 game series, the starting goalie will face around 175 ES shots.  Even if Miller was elite-level (0.930 ES) and Halak was average (0.920 ES) we would expect less than 2 goals difference between the two over 175 shots.  If a 0.920 goalie faces 175 shots, the 95% Confidence Interval for observed save percentage is 0.8694 to 0.9556.  That means the difference between a 0.920 goalie with bad randomness and a 0.920 goalie with good randomness is about 18 goals.

Output_medium


In fact, out of 100 series like this, the 0.930 goalie winds up with a higher observed save percentage about 60% of the time.  33% of the time the 0.920 goalie  winds up with a higher observed save percentage and 7% of the time they tie.

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

Join St. Louis Game Time

You must be a member of St. Louis Game Time to participate.

We have our own Community Guidelines at St. Louis Game Time. You should read them.

Join St. Louis Game Time

You must be a member of St. Louis Game Time to participate.

We have our own Community Guidelines at St. Louis Game Time. You should read them.

Spinner

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9355_tracker