Wednesday, October 29, 2008

Adjusted Shutouts

In an earlier post, I calculated how well goalies did against league-adjusted save percentage. Some of the feedback from that post was that save percentage doesn't show everything. Fair enough, I will be looking at a couple of other traditional goalie statistics in a series of upcoming posts. With Martin Brodeur chasing down the career shutout record, I decided to focus first on shutout statistics. Anyone can look up the raw stats and see the all-time and active leaders, but as we have seen many times, dealing with unadjusted goalie stats is almost always a bad idea. As such, we need to add some context to the stats to be able to isolate who was actually outstanding and who was merely average or worse. I came up with two different ways of adjusting shutout totals. The first is adjusting shutouts for era, which is described in this post. The second, adjusting shutouts for the number of shots faced, will be described in a future post.

If you look at the career shutout ranking, you notice right away that shutouts are heavily impacted by era effects. The best periods for shutouts were the pre-World War II era, the 1950s, and the so-called "Dead Puck Era" (1997-2004). Goalies who played during these periods dominate the career list. The most difficult times to record shutouts were during World War II and the entire 1980s decade.

To take this into account, I have calculated an era-adjusted shutouts statistic, done by adjusting each goalie's number of shutouts based on the league average frequency of shutouts on a season-by-season basis. If shutouts were common around the league that season then each one received less of a weighting, but if shutouts were scarce goalies received bonus points for each one they recorded.

Here are the top 25 goalies of all-time in adjusted shutouts, from 1917-18 through 2007-08. I have included total shutouts, adjusted shutouts, as well as adjusted shutouts per 1000 games played to compare goalies with different career lengths.


RankGoalieSOAdj SOAdjSO/1000 GP
1.Clint Benedict57107.7297.5
2.Martin Brodeur9694.797.8
3.Tony Esposito7694.0106.1
4.Ed Belfour7689.793.1
5.Terry Sawchuk10387.890.4
6.Patrick Roy6687.484.9
7.Glenn Hall8484.993.7
8.Dominik Hasek8183.8114.0
9.Jacques Plante8280.996.7
10.Turk Broda6270.3111.8
11.Tiny Thompson8168.3123.5
12.Mike Liut2564.697.3
13.George Hainsworth9464.5138.7
14.Harry Lumley7163.679.2
15.Rogie Vachon5163.279.5
16.Bernie Parent5462.3102.5
17.Tom Barrasso3861.979.7
18.Alec Connell8161.3147.0
19.John Vanbiesbrouck4059.167.0
20.Ed Giacomin5458.596.1
21.Ken Dryden4655.9140.8
22.Lorne Chabot7155.1133.7
23.Curtis Joseph5150.955.2
24.Andy Moog2849.669.6
25.Frank Brimsek4049.596.3


Martin Brodeur does well by this metric, hanging onto the same #2 spot he holds in actual career shutouts. Tony Esposito, Ed Belfour and Patrick Roy move up the adjusted rankings while Terry Sawchuk and George Hainsworth drop down. Clint Benedict is the new leader, but his adjusted shutouts per game rate makes it obvious that his numbers are somewhat inflated. For most of his career he had the advantage of playing on the best team in a 4 team league, which makes it much easier to beat the average. Benedict has a good shutout record, but I'm not sure it is quite that good.

Probably the most interesting result on the list is Mike Liut, who had only 25 career shutouts but ends up 12th all-time in adjusted total. Over the course of Liut's career, his adjusted shutout per games mark has been virtually equivalent to Martin Brodeur's. Other goalies who suffered from playing in the wide-open 1980s but still posted very strong shutout totals include Tom Barrasso, John Vanbiesbrouck, and Andy Moog.

9 comments:

Anonymous said...

this is somewhat bogues too. looking at overlapping numbers from 93-94 through 2002-03 here are the results

hasek 60 shutouts
brodeur 63 shutouts
roy 39 shutouts

now unlike the contrarian goaltender i will give you all the information here, not just the stats that help make my case. it is clear from a study done by bruce earlier that the difference between roys avalanche and brodeurs devils was not significant. thus 63 shutouts from brodeur obliterates roys 39. hasek on the other hand comes in a close second with 60 shutouts, but it is only fair to mention hasek did not play 2002-03. so in this case again, if there is a point to this study, it is misleading because to see patrick roy moved up the list the way he is, isnt correct. i'm sick of all this credit you give roy for having played in the mid-late 80's. you use this as the foundation for infating his statistics, yet i dont see you giving brodeur any credit for playing in the currect era 05-09.

The Contrarian Goaltender said...

Anonymous: This is not about giving anyone credit, it is simply adjusting for league average. All these numbers are based on identical mathematical formulas, so unless you are disputing the league averages then I don't understand your disagreement.

In the last three seasons, 5.8% of games have ended as shutouts. From 1986 to 1993, 2.9% of games ended as shutouts. Why would Brodeur get extra "credit" for those past three seasons, when shutouts were twice as frequent as they were during the early part of Roy's career?

You "give all the information" by cherry-picking a few seasons after Roy's prime and using that to justify how he should be ranked? I'll let the reader decide who is providing all the information and who is merely providing the stats that help their case.

Anonymous said...

Your are a complete moron, and keep re-analyzing stats til you find the answer you want. Take your head out of your ass and pay respect to the best goaltender of all time!

Anonymous said...

Anonymous - your giving us Anonymous' a bad name!

I realize math & spelling are not your strong suits, but what criteria do you use to label Brodeur as the best of all time?

Anonymous said...

bottom line is brodeur has put up numbers that no one else in the history of the game has ever put up. you can make all the excuses or theoretical arguments you want but bottom line is no one else has done it. this whole sight is about "if this, or what if that" bullshit. regardless of how you manipulate numbers, you can not takes away the records brodeur will have. any idiot can write a blog. if this guy was so good at what he did he would have an opinion that mattered or at least was supported by mainstream media, not an opinion that could only be found on the depths of a google search engine.

Anonymous said...

That's right, Brodeur has put up those numbers. But what exactly does that mean? That he's the greatest of all time? Or the goalie who will soon have the most wins etc? Or both? If both, why?

Gretzky has about 1,000 more pts than Howe... so Gretzky is [[2,800/1,800] -1] x 100 = 56% better at scoring points than Howe? Should we factor in things such as average goals per game during Gretz's era vs. Howe's era?

Thats the kind of even basic math that most "mainstream media" doesn't touch... because most mainstream sports media are poorly educated & basically bumbled into their careers.

"supported by mainstream media"

haha, good one!! If you're going to wait for mainstream sports media to come up with reasoned, advanced analysis you'll wait for a long, long time.

If these analyses anger you so much, you should go back to reading the basic Grade 8-quality sports pages.

The Contrarian Goaltender said...

Brodeur and his teammates have put up numbers that no one else in the history of the game has ever put up. Look at numbers that are more of a measure of Brodeur's individual play, like save percentage or some of my team-adjusted stats, and there are many guys who equal or surpass him. And that is the point of this site, and why I don't even need hypotheticals to make the case - just look at Brodeur's stats and his teammates and what you will see is certainly not the greatest goalie ever.

You are right, though, I can't take away any of the records Brodeur will have. Just like Brodeur can't take away the save percentage records from Dominik Hasek.

Anonymous said...

you seriously sound like a bitter 14 year old. are you from new york?
you seem like the typical new yorker, whether you root for the sabres or rangers, i dont know but comments like "brodeur cant take away the save percentage records from hasek" are juvenile. not to mention if thats how you wanna argue then... hasek cant take away the wins or shutout records brodeur will have. hasek cant get more cups than brodeur, and couldnt even get one without jumping to a front runner.

The Contrarian Goaltender said...

No, I'm not from New York or a fan of any New York-based teams. My arguments on this site come strictly from my interpretation of the numbers and my evaluation of goalie play, rather than as a result of any team affiliation.

Save percentage is by far the most important goalie stat, whereas wins, shutouts and Cups are team stats that have little value in evaluating goalies. That is why I brought it up and why Hasek's lead in that area is so significant. It is my standard counterargument these days because the popular perception seems to be that Brodeur is about to set every goaltending record ever.