Sunday, March 8, 2009

GMs vs. Media

There is often a great deal of similarity between voting for the Vezina Trophy, which is done by the league's general managers, and the voting for the year-end First and Second All-Star Teams done by the broadcast media. Last year was an example of a time when they did not agree, however, with Brodeur winning the Vezina and Nabokov getting named to the First All-Star Team. I thought it might be interesting to highlight some of the goalies who have had different results in Vezina and All-Star voting, not in terms of any particular season as in the case of Nabokov and Brodeur, but over the course of their careers to see if there was anyone that the GMs thought highly of but was overlooked by the writers.

Award voting is difficult to compare directly, since the number of voters changes from year to year. The best method is probably the one used at hockeygoalies.org, where the voting share is calculated for each goalie by dividing the number of voting points by the number of points a goalie would receive if he was a unanimous winner. That gives us a single scale that allows us to not only compare from year-to-year, but also between Vezina and All-Star voting. To get a career total, you just need to add up the total award "shares" for each goalie. For example, if a goalie gets 100 points out of a possible 200, that is a 0.50 share. If he has a 0.25 share the next season, and a 0.10 share the season after, his career total is 0.85.

For the well-known goalies there isn't too much difference between the two totals. Here are the results for the 4 most likely recent Hall of Famers, who all have All-Star voting totals between 8-15% higher than their Vezina results:

Belfour: 2.49 Vezina, 2.86 All-Star
Brodeur: 5.19 Vezina, 5.75 All-Star
Hasek: 4.76 Vezina, 5.22 All-Star
Roy: 4.64 Vezina, 5.00 All-Star

It is fairly standard for the well-known goalies that All-Star voting surpasses Vezina voting, since there tends to be more consensus. There are more a lot more writers voting on the awards than GMs, so one or two off-the-wall picks have less of an impact percentage-wise. The writers are also probably at least somewhat aware of each other's articles and published opinions, so that will influence some groupthink. It is also likely that some writers simply don't have a good idea of who the best goalie is, so they pick the most recognizable name.

On the other hand, some GMs don't get to see all the teams play so they may focus primarily on their own conference or division, or might even just throw some support behind their own goalie, which leads to more spread out voting results.

There were a number of goalies that accumulated substantially more Vezina shares than All-Star voting shares, meaning that they were valued higher by general managers than the media. If we assume that general managers have extra inside information from professional scouts as well as their hockey experience and expertise in evaluating talent, we would expect that the GMs should make better picks, and that the goalies that come up on this list should be goalies that are considered underrated.

There were 5 guys who met my threshhold of having Vezina shares that were at least 20% higher than All-Star shares:

Kelly Hrudey: 0.38 Vezina, 0.12 All-Star
Guy Hebert: 0.14 Vezina, 0.02 All-Star
Curtis Joseph: 1.02 Vezina, 0.42 All-Star
Daren Puppa: 0.82 Vezina, 0.60 All-Star
John Vanbiesbrouck: 1.52 Vezina, 1.07 All-Star

I have been less than flattering in some of the things I have written about Vezina voting and general managers, but here I'm going to be quite complimentary. That is a pretty good list.

Kelly Hrudey, Guy Hebert, and Daren Puppa are all guys that have pretty solid save percentage records, and often pop up in unexpected places when I look at various statistical comparisons. Curtis Joseph and John Vanbiesbrouck both have very good save percentage record compared to their peers in both the regular season and playoffs, but never had the team success that attracts recognition.

Curtis Joseph is the goalie who was most consistently rated higher by people inside the game of hockey than by casual observers. In the eyes of many fans, Curtis Joseph is a loser who always choked past the second round of the playoffs and never accomplished anything. Many people consider him to be a borderline Hall of Famer, despite being 4th all-time in the career wins list. Joseph never won a Vezina or was named a First or Second Team All-Star, and he never played in a Stanley Cup Final. However, if you look at insider perspectives on his play, including Vezina voting, he was repeatedly rated among the very best in the game.

Joseph did not have a single season in his career where his All-Star voting share was higher than his Vezina voting share. The GMs always rated him higher than the writers did, and the most extreme example was during the 1998-99 season, when Joseph was the Vezina runner-up and actually received more first-place Vezina votes than eventual winner Dominik Hasek (!), yet finished just 4th in All-Star voting. Joseph was picked to play on Team Canada for three consecutive best-on-best tournaments, and began the tournament as the starter in two of them, despite playing in an era of strong goalie competition. Joseph was always a highly sought after free-agent and inked a series of lucrative deals, including signing for $8 million per year with Detroit in 2002 (the same amount the Red Wings were paying Hasek). Joseph was voted team MVP several times in both Edmonton and Toronto, and according to one source I found he was apparently a Lester B. Pearson finalist in 1998-99, which suggests that the other players around the league seemed to agree.

As I have argued many times before, NHL general managers are not always great at picking the best goalie in any given season, as they can be blinded by wins and shutouts like anyone else. However, the record shows that they are better than the writers at identifying overlooked goalies and rewarding goalies who play on weaker teams.

13 comments:

Bruce said...

Belfour: 2.49 Vezina, 2.86 All-Star
Brodeur: 5.19 Vezina, 5.75 All-Star
Hasek: 4.76 Vezina, 5.22 All-Star
Roy: 4.64 Vezina, 5.00 All-Star


Hey lookit! That Brodeur guy must be pretty good. Top of both lists. The only one still active, so bound to open a wider gap before hes done.

Statman said...

Interesting analysis.

Somewhere on HFBoards.com there is a listing of NHL trophy/allstar team voting, going back to the 50's or 40's.

The Contrarian Goaltender said...

Bruce: Nobody is saying Brodeur isn't pretty good. I don't put much weighting on awards voting, because I think the voters tend to focus too much on team success. The point of the post was to identify discrepancies between voter pools, rather than to use awards voting in goalie evaluation.

However, if you do place a lot of importance on award voting keep in mind that the #1 variable for awards voting is quality of competition. Just think of the classic example of centres from the 1980s - obviously their awards voting records are going to be worse than comparable players from other eras simply because they played the same position at the same time as two of the greatest ever.

Three of the goalies listed were born in 1965, and one of them was born in 1972. We would expect that it would be more difficult to amass awards votes when you are one of three Hall of Famers playing the same position and born in the same year. Recently Brodeur has seen his main rivals retire while also improving his own play, which is why the majority of his awards consideration has come over the last 5 seasons.

The Contrarian Goaltender said...

Statman: The problem is that Vezina voting started in 1982. Before that all we have is the All-Star Team voting, so a similar comparison isn't possible.

Scott said...

Thanks for this analysis CG. I think that this does put the GM's in a bit of a better light. Are there similar discrepancies on the other side (other than the four you mentioned)? It would be interesting to see if the writers value certain less heralded goaltenders more than the general managers did.

nu said...

Fun w/stats on a tangential note: one formulation for an HHOFer...which is, of course, based heavily on what other people think of you, starting with awards.

As far as goalies go...and a few more names, w/Cujo getting thrown in.

Something else to objectify them by!

Bruce said...

Bruce: Nobody is saying Brodeur isn't pretty good.

Hmmm, maybe the blog name should be changed to reflect that. "Brodeur is a Pretty Good Fraud", perhaps? :)

I don't put much weighting on awards voting, because I think the voters tend to focus too much on team success. The point of the post was to identify discrepancies between voter pools, rather than to use awards voting in goalie evaluation.

Well we all know where the latter will lead. The "experts" in both voter pools seem to agree that Brodeur is pretty Darn good. The question of discrepancy between the pools is a moot point if you don't value the process in the first place.

However, if you do place a lot of importance on award voting

I give it some value. It's another form of statistic, based on opinion rather than performance, but a collection of those opinions in the context of an entire career surely means something. I don't always agree with the voters either you may be sure, but by and large a group of informed people charged with the responsibility of assessing performance are mostly going to take their responsibility seriously, and their opinion as a group bears some weight. Bill James wrote a great essay on this (that surprised the shit out of me) which he applied especially to bygone eras, but from that foundation right across the board. Nowadays of course we all get a chance to see these guys for ourselves at least occasionally, and we have a bunch more stats to judge them by. Still, when I sit in my couch and opine that Martin Brodeur is the best goalie in the game that doesn't mean a whole lot, but when there are two pools of experts that agree with me year after year it does add some degree of confidence.

keep in mind that the #1 variable for awards voting is quality of competition.

That's a fair point. Brodeur certainly was held back from awards early in his career but you're right that he's made hay in the last few years with the retirements of the Class of '65, while the younger guys like Luongo, Giguere and Lundqvist have themselves struggled to make inroads on the awards front cuz the Class of '72 is pretty good.

The Contrarian Goaltender said...

Hmmm, maybe the blog name should be changed to reflect that. "Brodeur is a Pretty Good Fraud", perhaps? :)

Still hung up on the blog name, eh? I might get around to changing it at some point, but I think you are well aware of how I rate Brodeur overall.

The question of discrepancy between the pools is a moot point if you don't value the process in the first place.

I don't think so. It's not that I think the voters know nothing at all about goaltending, it is just that there are a couple of factors that are weighted too heavily in my opinion - wins and games played. Voters probably do a fairly decent job comparing goalies in similar team situations with similar workloads, but they go crazy when they see the big win totals. I'd say that is true for both voting pools, which means that a head-to-head comparison removes most of the team play bias affecting the results. The guys I identified above never really posted big win totals. For example, none of them ever led the league in wins, except for Vanbiesbrouck in 1985-86 when he, surprise, surprise, won the Vezina Trophy. They must have received all those general manager votes for other reasons, and that makes the results interesting to me.

Bill James wrote a great essay on this which he applied especially to bygone eras, but from that foundation right across the board.

Sure, I agree with the general point that awards voting has value. I place a lot more importance on awards voting for forwards and defencemen. The problem is that hockey goaltender is one of the most team-dependent positions in all of professional sports, and is simply more difficult to evaluate. I don't think that is a particularly controversial point - as I've said before, get people to name the top 5 players ever, and then get them to name the top 5 goalies ever. The first list is going to be a lot more consistent than the second one.

The Brodeur debate is another illustrative case - I'm struggling to think of a skater that is ranked as the greatest ever by one group, while ranked as the third best of his own era and in the lower half of the top 10 by another group. I'm not even sure there is a good parallel to that in any other professional sport. Goaltending depends on the team, and there are a lot of people out there who don't recognize that.

Bruce said...

I'm struggling to think of a skater that is ranked as the greatest ever by one group, while ranked as the third best of his own era and in the lower half of the top 10 by another group.

CG: Brodeur's "own era" is now. Who are the other two guys from this era that are better?

Brodeur is fully 7 years behind your Class of '65, which you were very quick to point out how that worked to his advantage in awards votes. But he won those awards over the past five seasons because he's been the Best of that era, after being one of the best of the preceding era, which happened to include several of the greatest in NHL history. He belongs in that group. The Greatest? Let him finish making his case.

The Contrarian Goaltender said...

Bruce: I tend to think of eras of goaltending based on league scoring levels rather than goalie birthdays, because that makes the most sense in terms of directly comparing numbers. I'd split it up into pre-WWII, Original Six, expansion era, the high-scoring '80s, the Dead Puck Era, and the new NHL. Brodeur's career basically started at the same time the Dead Puck Era did, so I think it is natural to include him there.

It's a fair point, though, that Brodeur's age difference means he shouldn't really be compared directly to Hasek and Roy, and Marty has certainly been at or near the top post-lockout.

My point about the discrepancy in the way he is evaluated still stands, but as you say he is still making his case. In my view he is doing a lot better of job of that now than he did between 1998 and 2004, and if he keeps it up he will also naturally improve his all-time status.

Bruce said...

CG: Fair enough.

Anonymous said...

I laugh when you say Curtis Joseph choked in the 2nd round of the playoffs. His save percentage was 2nd overall in the playoffs both in 2000 and 2001, the only problem was he had to face Brodeur and the A Line.

The Contrarian Goaltender said...

I laugh when you say Curtis Joseph choked in the 2nd round of the playoffs.

I never said that, I said that many fans think that he choked. I have been a big advocate of Curtis Joseph and his playoff record. Click on his name in the underrated goalies section on the right to see an article about his playoff performances.