Behind The Numbers: Luongo Vs. Schneider

LAS VEGAS, NV - JUNE 22: (L-R) Roberto Luongo and Cory Schneider of the Vancouver Canucks accept the William M. Jennings Trophy during the 2011 NHL Awards at The Pearl concert theater at the Palms Casino Resort June 22, 2011 in Las Vegas, Nevada. (Photo by Ethan Miller/Getty Images)

...Everybody's Talking 'Bout the New Kid in Town"

"Cory's on a roll....!---""He's so much better than Lui...!"---"He should start in the playoffs!"---"The team plays better in front of him"----

Okay, you are the Canucks coach for the playoffs....Who do you start? And, how long do you stay with either goalie? For a detailed and revealing summary of Luongo's performance and Playoff Goaltending using advanced hockey stats check here. It's time to vote and to look 'behind the numbers' after the jump.

You would think that having two 'elite' goalies would be an advantage for any team. Yet in Vancouver, once dubbed by former g.m. Brian Burke a goalie graveyard, it is an ongoing simmering controversy. Cory Schneider's stellar play this year, combined with the harsh judgment of Roberto Luongo's playoff performances, and a coaching staff that still appears apprehensive to simply go with the hot goalie, leaves us wondering how things will play out?. In addition, even though most of the numerical evidence clearly shows that great goaltending is not as important to cup success as other areas (i.e. goal scoring), many are still convinced that the Canucks must have a 'supernatural hero' between the pipes to lead us to the Holy Grail. Thus, the decision on who to play is still what everybody's talking. so lets use an advanced hockey stat approach to decide this question.

Evaluating Goalie performance has proved to be a difficult task. First, it is hard to separate goalie talent from the defensive skill of the team in front of him. Second, performance of goalies on the penalty kill is known to fluctuate wildly season to season and not to be repeatable, so we normally use Even Strength Save% as the foundation for evaluating keepers. Also, there is an issue of 'sample size'. It turns out that the talent of of NHL goaltenders cannot be clearly determined in a small number of games. For example, You would think that a goalie that holds the consecutive shutout record of 5 games, the best performance in 70 years, would be a sure fire all-star and hall of famer. The holder of this record is --, 'Brian Boucher?' , a career average goalie at best. So, we accept it takes time for the best goalies to establish themselves, but how long? When can we know that a new goalie's performance is repeatable? Gabriel Desjardins, a well recognized leader in advanced stats posted this simple example to differentiate between the luck/randomness ans skill of new goaltenders.

For a goalie with an even strength save% of .925 (.920 is league average) it takes 2500 shots or ~110 starts to say with 85% confidence that he is an above average goalie. We will refer back to this graph as we move forward. Steve Mason of the Columbus Blue Jackets is a recent example. He is a former world junior champion in 2008 and in his first 1266 shots (through 61 games) he had a stellar .925 save %. He was the winner of the prestigious Calder Trophy for the NHL best rookie that year, a certain star of the future and many were touting him as a potential Canadian Olympic goalie for 2010. Using the previous graph, at that time, Mason had a ~75% chance of being an above average goalie.In his next 1200 shots his save% plummeted to .911. Now, 3 years later, having faced over 4800 shots in his career Mason's save % is well below average at a bleak .914. So let's start by at looking at:

Even Strength Save % of goalies (since the lockout) who have faced a minimum of 3000 shots.

Goalies Sve.% (Even) Shots (min.3000)
League Average .920 ------
T. Thomas .930 8822
T. Vokoun .930 9797
R. Luongo .929 10035
H. Lundqvist .928 10033
P. Rinne .928 5771
J. Hiller .928 5544
K. Lehtonen .926 7812
J. Halak .926 4584
N. Backstrom .925 7281
J Giguere .925 6449
I. Bryzgalov .925 8438
M. Kiprusoff .924 10900
C. Price 924 6324
J. Quick .924 5178
R. Miller .923 10083
C. Anderson .923 6117
A. Niemi .923 3577
J. Howard .923 4254
E. Nabokov .920 7085
R. Emery .920 4156
M. Fleury .919 8679
M. Smith .919 4939
O. Pavelec .919 4334
M. Brodeur .918 5685
C. Ward .918 9332
M.Turco .916 6939
J. Theodore .916 6646
C. Mason .916 5885
R. Dipetro .916 5072
T. Conklin .916 4095
S.Mason .915 4824
B. Elliot .915 3689
B. Johnson .915 3161
S. Clemensen .915 3097
J. Hedberg .914 4445
N. Khabibulin
.913 6613
C. Osgood .909 3355

{all stats from NHL.com}

With this large sample size we can be more confident in the talent level of these goalies. The usual top performers,and high salary earners (Thomas, Luongo, Lundquvist, Rinne) lead the way. Yes, Vancouver fans, Luongo is one of the best. Kari Lehtonen who played a large part of his career in the obscure Atlanta market is a real surprise, could he steal a series for Dallas this year? And Jonas Hiller who already has had playoff success with Anaheim are both underrated. Again, Marc-Andre Fleury is surprisingly below average and I contend, overrated. In fact, the usual sharp Elliot Friedman from HNIC has repeatedly touted Fleury as the starting goalie for the 2014 Canadian Olympic team, most recently on the team 1040, and in doing so,clearly disrespecting Lui's tremendous performance in International Hockey.{64 ga in 33 games, or 1.94 ga.} And, Martin Brodeur, who many agree is on the downside of his career, appears to have hung around too long. Brian Elliot appears to be really benefiting from coach Ken Hitchcock's defensive system and, last but not least, it is Stanley Cup Winner Chris Osgood.- more proof that Nikolas Lidstrom and the Detroit puck possession system has been so dominant that goaltending was practically irrelevant. Now, lets look at Cory Schneider and,

The Even Strength Save % for goalies under 3000 shots

Goalies Save.%(Even) Shots(under 3000)/ [~% of Confidence]
T.Rask .932 2308 [ 99%]
J.Harding .926 2404 [84%]
C. Schneider .926 1383 [78%]
S. Varlarmov .925 2512 [85%]
J. Riemer .925 1728 [82%]
S. Bobrovsky .922 1749 [~70%]
C.Crawford .921 2544 [~73%]
D. Dubnyk .920 2308 [~70%]
S.Bernier .915 961 [?unknown]
B.Boucher .914 2627 - a below average goalie
J.Macdonald .914 2281 - a below average goalie
J.Gustavsson .909 2366 - a below average goalie

Yes, Cory has been really good and well above the league average. However, the title of 'the best goalie not yet in a starting role' goes to, Boston's Tuukka Rask . Minnesota's long time back-up Josh Harding has faced almost twice as many shots as Cory, however, Cory is younger with more chance to improve. And, what do you think about Cory having a 22% chance that he is only an average goalie? I know I know its very hard to believe.! Colorado's Varlarmov looks like he is a solid bet to be good.. And, Toronto's much maligned James Reimer is third! I had to rerun the numbers! I couldn't believe it! I mean who would say that Reimer is as strong a prospect as Schneider? Once again we see that objective analysis disregards myth-building. Still, I wouldn't give up on Reimer too quickly. And, we will see shortly just how bad Ron Wilson Maple Leafs have been on the PK. So we need to examine the numbers further to understand why we feel Cory's play has been so stellar.

The PK Save % of goalies who have faced a minimum of 1300 shots.

Goalie Sve.%(PK) Shots
League Average .870 ---
N. Backstrom .893 1566
C. Anderson .885 1527
H. Lundqvist .884 2415
C. Mason .883 1448
R. Miller .882 2290
J. Hiller .881 1329
T. Thomas .880 1983
R. Luongo .880 2835
T. Vokoun .880 2750
M. Brodeur .880 1417
C. Price .879 1449
M. Fleury .877 2195
M. Kiprusoff .874 2916
C. Ward .873 2122
E. Nabokov .873 1746
J. Giguere .870 1886
I. Bryzgalov .870 1967
a. Nittikmaki .870 1396
N. Khabibulin
.860 1797
J. Theodore .859 1714

Again we have the best goalies near the top with little to choose among them. However, because of the smaller number of shots we do have a few surprises such as N. Backstrom of the Minnesota Wild, and C. Andersen of Ottawa, and C. Mason now with Winnipeg, but there is a strong chance that they will all regress down to the group with additional shots. Now let's see how Cory measures up? :

The PK Save % of Goalies facing under 1300 shots

Goalie Sve. % (PP) Under (1300 shots)
C. Schneider .921 341
S. Bernier .899 237
J. Quick .893 1194
J.Howard .890 872
T. Rask .888 436
J. Halak .883 973
J LaBarbera .882 836
P. Rinne .880 984
M.Smith .877 1077
B.Elliot .874 831
A. Niemi .874 697
J.Harding .872 539
S. Clemmensen .870 618
B.Johnson .869 961
R. Emery .864 1044
S. Mason .861 1099
D.Dubynk .861 571
O. Pavelec .861 924
B. Boucher .860 671
s. Bobrovsky .856 363
T. Conklin .855 995
S. Varlarmov .855 471
C. Crawford .855 470
J. Gustavsson .842 414
J. Reimer .831 315

Wow! We clearly see that Cory has been unworldly on the PK. (And, boy has the Toronto goalies penalty killers been bad!) Still, given Cory has only faced 300+ shots look for his PK Save % to regress strongly downward when he becomes a starter. To once again hammer home the importance of sample size take a look at:

The current top 3 all-time leaders in combined Save %

1. Cory Schnieder .928

2. Tuukka Rask .926

3. Dominik Hasek .922

Yes! Rask and Schneider are really good but better than,'The Dominator' - I don' think so! Also, it is no coincidence that the top two young goalies play for the top two teams from last year.

Okay, now we know were our two goalies rank in the league let's compare them one on one.

Save %(Even) Last 4 yrs. Last 2 yrs. This year Diff.(+/-) this yr.
Luongo .930 .930 .929
Schneider .926 .931 .932 +2 goals / 29 games

Virtually identical performances. Of course, Cory 's numbers are in only ~57 career starts while Luongo has played in 138 over the same time period. Should this matter? Most Cory advocates assume he would only get better as a starter. However, there is a logical contrary opinion to this view. Playing goalie is clearly one of the most difficult positions in pro sports, placing a high level of both physical and mental demands on the individual. It makes sense that the more they play the more demanding the position. Yes 'getting in a groove' is important but for high level athletes this ought not to take more than a handful of shots.This report by Philadelphia blog, "Broad street Hockey" agrees.

"There is no evidence whatsoever that a lot of consecutive games will help a goalie get in a rhythm"

So there is a good chance that a large increase in starts for Cory will mean at least a slightly reduced performance. So, here is the only notable difference between Vancouver's two goalies:

Save.%(PK) Last 4 yrs. Last 2 yrs. This year ~ Diff.(+/-) this yr.
Luongo .871 .886 .874
Schneider .920 .936 .957 ~ +12 goals in 28 games

But,we know that this difference is not likely to continue in the long term. I have also thrown in the Quality Starts & Really Bad game stats ("Blow-ups") for the two goalies as they were asked for in the comment section of my previous post: "Playoff Goaltending"

Quality Starts%s Last 4 yrs. Last 2 yrs. This year Diff(+/-) This year
Luongo 62% 65% 62%
Schneider 69% 74% 75% +2.5 wins this year

{Quality Starts and Blow-ups created by HockeyProspectus.com}

Yes, Schneider has been better, worth close to 3 wins this year over what we could expect from Luongo but we know now that this has been driven primarily by Cory's PK save %.

Blow ups % Last 4 yrs. Last 2 yrs. This year
Luongo 13% 14% 12%
Schneider 17% 13% 10%

And, not much of a difference here.

To summarize: Since we only have a small sample size and since goalie stats regress very strongly we cannot say that Cory Schneider is better than Roberto Luongo. We expect Cory''s Pk % will come down. And, there is a good chance his Even Strength Save % might dip as well as he faces the grind of being an every game starter. And, strictly by the numbers, we can say right now that Cory has an ~80% chance of being an above average goalie long term.

So back to our original question, Who do you start in the playoffs? Subjectively, it appears the Canucks play better in front of Schneider so I checked and I have one more interesting stat (from timeonice.com)

Team Pos. F (close games) League Rank
Luongo 50.5% 11th
Schneider 53% 4th

When it counts (during close games) The Canucks play noticeably better in front of Cory Schneider

We are not imagining it! Vancouver plays better in front of Schneider. The difference was also present last year although not as dramatic. What are the possible reasons for this?

  1. Cory Schneider handles the puck better (Not!)
  2. He controls rebounds better (possible small effect)
  3. The team has more confidence in Cory so they are able to control the puck more. (possible)
  4. Cory has played an easier schedule (not true, Boston, Chicago, Nashville and road games)
  5. This is just random variation (perhaps? but- Pos. F is known to be reliable over 60 games).

Conclusion:

One thing is clear that Luongo is not responsible for how other players play., especially since his performance in Vancouver has been top notch.That is unfair. Still, in the playoffs the margin for error is acutely small. Boston with the greatest goalie ever - ("gotta pump Timmy's tires"), went to game seven three times including one overtime game last year. So, it's clear, every edge counts. As big a fan of Roberto as I am, I feel the time has come! I say start Cory.

  1. He has proven to be as good as Luongo in short term
  2. He has no history. So, if things go wrong which will happen (all champion ship teams face adversity) the team and city won't have Lui as their familiar scapegoat to carry 40+ years of Canuck failure.
  3. The Team seems to play better in front of Cory (for whatever reason).
  4. We can expect Cory to be as good as Lui in the short-term
  5. If Cory fails, Lui can come in with nothing to lose, play more focused, trying to prove himself and potentially save the day and be the hero.

(This happened to injured Tim Thomas who lost his job to Tuukka Rask in the 2009-10 playoffs, and we all know what happened subsequently!)

Will the Canucks do this? -- Not likely.

Of course, if the Canucks were planning to do this I would have expected them to play Cory even more this year. They easily could have given his strong performance and the huge lead the Canucks have in the division.

The short leash?

Okay, so if Luongo starts then we just pull him quicker. (But remember, in one of the strangest coaching moves ever, Alain Vigneault left Luongo in for 8 goals!! in Game three of the Finals). Who knows why? So, how quick? A bad game? one bad goal? two bad goals? It's time for your input.

Tell us what you think!

X
Log In Sign Up

forgot?
Log In Sign Up

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior users will need to choose a permanent username, along with a new password.

Your username will be used to login to SB Nation going forward.

I already have a Vox Media account!

Verify Vox Media account

Please login to your Vox Media account. This account will be linked to your previously existing Eater account.

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior MT authors will need to choose a new username and password.

Your username will be used to login to SB Nation going forward.

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,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Nucks Misconduct

You must be a member of Nucks Misconduct to participate.

We have our own Community Guidelines at Nucks Misconduct. You should read them.

Join Nucks Misconduct

You must be a member of Nucks Misconduct to participate.

We have our own Community Guidelines at Nucks Misconduct. You should read them.

Spinner.vc97ec6e

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