Monday, March 7, 2011

Longevity

Here are some career games played numbers, counting all games played in the regular season and playoffs in the NHL, all games played in a major European professional league and all games played in senior international competitions:

1. Dominik Hasek: 1,358
2. Martin Brodeur: 1,325
3. Patrick Roy: 1,282
4. Ed Belfour: 1,153
5. Curtis Joseph: 1,092

Brodeur and Roy have been rightly praised for their excellent longevity, but it's the guy at the top of the list that might surprise some NHL fans.

Hasek was the youngest professional hockey player ever as a 16 year old in the Czech League, and he led the KHL in shutouts this season three decades later at the age of 46. To put that into perspective, Gordie Howe's professional career spanned 34 years, just 4 more than Hasek's to date.

Hasek ranks 5th all-time among NHL players in career GVT, even though just 63% of his professional games were played in the NHL. Give him credit for those European games (the vast majority of which were played while Hasek was good enough to play in the NHL but prevented from doing so by communism), and there's probably a good argument to rank him no worse than 3rd behind only Howe and Wayne Gretzky for the highest total value career ever.

17 comments:

Robert Vollman said...

Tom added up career GVT for players across all leagues, including playoffs, from 1967-Present.

Hasek was 4th, behind Gretzky, Roy and Bourque, and ahead of Jagr, Lidstrom and Messier.

Unfortunately his study didn't include any of Hasek's time pre-Blackhawk except for a couple of international tournaments, so he could be higher on the list.

Anonymous said...

You didn't include QMJHL or AHL games. You also didn't factor in that Brodeur is still playing. Yes, Hasek had great longevity, but Brodeur has the most endurance of any goalie in the history of the NHL.

The Contrarian Goaltender said...

The QMJHL isn't a professional league. I could have included North American minor league pro games, but that would have only increased Hasek's lead (54 reg. season and playoff games in the IHL compared to 32 for Brodeur and 14 for Roy, who both played in the AHL).

Having said that, I would agree that Brodeur has more endurance and durability than Hasek did, and I also expect Brodeur will end up #1 on this list before he retires, since as you point out he is still playing.

For pure durability and endurance, you'd have to take Brodeur over Hasek for sure. But in terms of overall performance, Hasek's longevity combined with his elite peak makes it not even close.

Anonymous said...

If you're going to include all those games played in the KHL and other leagues, then you need to include Hasek's stats in those leagues.

I haven't done the stats myself other then this year myself, but for example, Hasek is the NHL all-time Save Percentage leader, however, this season his save percentage in the KHL is 0.864, not very good.

Some of your analyses are very good, but you can't only look at the peak years of a player and determine that they are the best goalie ever.

For example, would you rather have the number 1 or 2 goalie in the NHL for 7-10 years (Hasek), or would you rather have a top 5 (Brodeur) goalie for 15-20 years (keeping into account that the top 5 goalie probably had some seasons where he was in the top 1 or 2 or outside the top 5)?

I would love to see you do an analysis of SV% and GAA of Brodeur and Hasek over their entire professional careers, and I'm willing to bet the numbers are much closer then you make them seem.

If you only want to use NHL numbers, then you have to take into account that Hasek didn't play as long in the NHL.

Anonymous said...

By the way, the 0.864 save percentage is for the KHL playoffs, his regular season was .915.

But still the playoff games were included in your totals so they should count toward the totals for SV% as well.

The Contrarian Goaltender said...

Some of your analyses are very good, but you can't only look at the peak years of a player and determine that they are the best goalie ever.

Sure I can, and I'm not the only one that does it. Who's the better defenceman, Bobby Orr or Nicklas Lidstrom?

I would love to see you do an analysis of SV% and GAA of Brodeur and Hasek over their entire professional careers, and I'm willing to bet the numbers are much closer then you make them seem.

Regrettably, there aren't save percentage numbers available from the Czechoslovakian league in the 1980s, and GAA is pretty meaningless unless one is aware of the team and league context that the goalie plays in. Based on his multiple MVP and best goalie awards, though, I'm pretty sure Hasek was doing quite well for himself.

For example, would you rather have the number 1 or 2 goalie in the NHL for 7-10 years (Hasek), or would you rather have a top 5 (Brodeur) goalie for 15-20 years (keeping into account that the top 5 goalie probably had some seasons where he was in the top 1 or 2 or outside the top 5)?

The first one. Again, Hasek's total career NHL GVT is higher than Martin Brodeur's. Hasek played fewer games, but he nevertheless made a greater contribution towards helping his team win NHL hockey games.

That said, my main point is that rating Hasek based on his NHL career only is unfair to him. His legacy is even greater than what he achieved in North America. And if his North American career already beats out Brodeur's, then adding in his European exploits only puts the Dominator even farther ahead.

Anonymous said...

Ever since I've been reading your blog, you've presented stats as the basis for your choosing of Hasek, you can't now say "judging by the amount of awards he won in the Czech league he was good there". If that's your argument then I say:

Brodeur was 1st or 2nd Team All-NHL (i.e. top 2 goalie) 7 times, Hasek was 1st or 2nd team 8 times. So I re-ask the question, would you rather have a guy who's in the top 1 or 2 for 8 years or a guy who's in the top 1 or 2 for 7 years and in the top 5 another 10 years after that?

By the same token, what if a player had 1 incredible peak year where he broke all the NHL single season scoring records, but never reproduced anything close to that ever again, certainly he's not the best player ever right? Now that's an extreme limit argument, but the point is a player's peak isn't the only factor in determining the greatest ever.

I added up Hasek's total goals allowed, versus games played in the Czech league, and in order for him to have had the save percentage he had in the NHL, 0.922 he would have had to be facing 34.45 shots per game, highly doubtful.

0.922 = (shots faced - goals allowed) / shots faced

Total goals allowed was 930.

Games played was 346.

Based on this his numbers in my opinion were not as good in the Czech league. Bringing me back to the point that yes Hasek's peak years 93-94 to 98-99 where higher then Brodeur's, but he didn't do it as long.

Brodeur's numbers are pretty consistent for 16 years now (and counting).

Also, only 3 times during Hasek's peak years did he play more then 60 games. What other reason would there be to take out the guy that everyone in the world considers one of the top 3 ever (first in your opinion obviously) for over 25% of the team's games, other then the fact that his body likely could not handle it. It's far fetched to say they took him out because of "team system". Almost every professional "team system" is play the best players for as long as they can play. If he could have played more there's no reason to think he wouldn't have. That's like the Chicago Bulls saying "we're not going to play Michael Jordan today because he needs some rest".

There will be no way of ever knowing if he would have put up the same numbers had he been playing an extra 15-20 games a season.

We do know that Brodeur's numbers remain relatively unchanged playing 70+ games a season.

We're obviously in disagreement about Hasek's European career benefiting his case, but I've tried to provide some numbers to suggest he was not as good when playing in Europe.

Agent Orange said...

"Who's the better defenceman, Bobby Orr or Nicklas Lidstrom?"

This would be a discussion that wouldn't end soon.

The one bit of nit-picking I would do is to say I wouldn't count Hasek's KHL games. The KHL is not the highest level available to Hasek. Are we going to count games in a beer league for guys?

I agree on using Hasek's czech league numbers as politics prevented him from playing at the highest level.

Just my 2 cents.

Anonymous said...

As long as I've been reading this blog, you've used statistics as your evidence to back up Hasek being "far better" then Brodeur. You can't now just say "judging by the amount of awards he won in the Czech league, he was really good then".

If that's your argument that I say this:

Hasek was voted 1st or 2nd team All-NHL 8 times, while Brodeur was voted 1st or 2nd team 7 times. So if that's the case, I re-ask the question: would your rather have a goalie who was in the top 1 or 2 in the league for 8 years or a goalie in the top 1 or 2 7 times, and in the top 5 for another 10 years after that?

Furthermore, here's an extreme example, what if a player had one really ridiculous season where he broke every single season scoring record there was but then never reproduced that form or anything close to it again, surely he wouldn't be considered the greatest ever right?

I ran some of Hasek's Czech league numbers, and he gave up 930 goals in 346 games. If he was going to have had the same save percentage he had in the NHL, 0.922, he would have had to have been facing 34.45 shots per game. Highly doubtful.

0.922 = (shots faced - goals allowed)/(shots faced)

Solve for shots faced as goals allowed is 930 and then divide that by 346. I think its safe to say Hasek's number in the Czech league were not as good.

Also, only 3 times during his peak 7 years 93-94 to 98-99 did he play more then 60 games. You can't blame "team system" for him not playing that long. Why would a team take out one of the 3 best goalies of all time for 25% of their games if his body could take playing more? That's like the Chicago Bulls saying "we're not going to play Michael Jordan tonight because he needs some rest, or we want to give the backup a chance to play a little". Every professional team's "system" is to play the best players for as much as they can handle. This is why I believe if Hasek could have played more, he would have. There's also no telling his number would have been as good if he was playing those extra 15-20 extra games a season.

We do know that Brodeur's numbers remain relatively unaffected by playing those extra games.

I'm not arguing that Hasek's peak years weren't better then Brodeur's, but it's hard to pick Brodeur's "peak years" because almost every single season he posts similar numbers, so you'd have to choose his 7 best individual seasons to compare with Hasek's and you could also argue that Brodeur never had a "peak" because he was consistently performing at generally the same level.

Think of Brodeur's career as a plateu, he quickly climbed to a high level and just stayed there, whereas Hasek's career has a definite spike (a level higher then Brodeur's) and a definite dropoff (a level lower).

The Contrarian Goaltender said...

You can't now just say "judging by the amount of awards he won in the Czech league, he was really good then".

You are correct that awards do not conclusively prove anything, but we can't do a full statistical analysis because the numbers aren't available. That leaves us to make a probability assessment. In my estimation, it is likely that somebody who won a lot of awards, was recognized at the time as one of the top goalies in the world, and subsequently dominated the NHL was quite likely performing at a high level in his domestic league. You may disagree, but I think that's the most likely explanation given the facts.

I re-ask the question: would your rather have a goalie who was in the top 1 or 2 in the league for 8 years or a goalie in the top 1 or 2 7 times, and in the top 5 for another 10 years after that?

And I re-answer the question: I'd rather have Hasek. Hasek's top 2 seasons were significantly better than Brodeur's top 2 seasons, and you're giving Hasek no credit for his late career while overrating Brodeur's consistency (much of it driven by his durability and his teammates).

My position is that Hasek prevented more NHL goals against than Martin Brodeur did. That's not more goals per game, or more goals comparing prime to prime. It's more total goals, everything included, over his entire career, even despite playing in almost 400 fewer games than Brodeur.

Therefore, even if you prefer to look at total career value without even considering peak value, I think you'd still have to take Hasek because his total career value is higher. And that's North American numbers only. If you add in a reasonable estimate of what he achieved in Europe then the gap gets even wider.

That's the conclusion I've reached from a variety of different methods such as looking at special teams adjusted numbers, comparing to backup goalies, save percentages compared to league average, win thresholds, etc. And I'm not the only analyst to suggest that, as I pointed out earlier the GVT numbers make the same conclusion.

I think its safe to say Hasek's number in the Czech league were not as good.

Save percentages in the 1980s were lower than they were in the 1990s. No kidding. Are you also willing to argue that Patrick Roy was a worse goalie in Montreal because from 1985 to 1991 his save percentage was a mere .900?

Raw numbers don't tell you anything, you need to measure relative dominance, but you can't do that unless you know something about league strength and scoring environment. Unfortunately, we don't know much at all about the Czech league in the 1980s, which again means we have to rely on the contextual evidence that suggests Hasek was a difference-maker over there.

Also, only 3 times during his peak 7 years 93-94 to 98-99 did he play more then 60 games. You can't blame "team system" for him not playing that long. Why would a team take out one of the 3 best goalies of all time for 25% of their games if his body could take playing more?

You need to review the context if you think there was anything lacking with Dominik Hasek's workload during his prime.

Games played, 1993-94 to 1998-99:
1. Martin Brodeur, 371
2. Dominik Hasek, 361
3. Patrick Roy, 360
4. Curtis Joseph, 351
5. Mike Richter, 345

Shots against, 1993-94 to 1998-99:
1. Dominik Hasek: 10,987
2. Patrick Roy: 10,469
3. Curtis Joseph: 10,205
4. Felix Potvin: 10,072
5. Guy Hebert: 10,051

Think of Brodeur's career as a plateu, he quickly climbed to a high level and just stayed there, whereas Hasek's career has a definite spike (a level higher then Brodeur's) and a definite dropoff (a level lower).

This chart disagrees with you.

Anonymous said...

I didn't not give credit to Hasek's late career, but as your chart even shows, he had a significant dropoff in his play later on in his career.

Also, that chart is misleading about Brodeur's SV% numbers because yes it looks like an upside down bell, but the entire spread of difference is only 0.01 throughout the whole career, whereas in Hasek and Roy's case, you can see a significant dropoff. This basically proves my point that his numbers remained relatively unchanged (only 0.01% difference). So I don't see that chart disagreeing with me.

Also, choosing to compare just those 7 years in games played is also unfair because a) Brodeur had just broken into the league and wasn't a full time starter in 93-94 and b) there was a lockout shortened season. Instead look at the 7 years Brodeur played before the lockout, 97-98 to 03-04, and let's call that his "peak".

Games played: 505. Now let's assume that the one season wasn't shortened and Hasek played more games then usual that season, that still would leave him 110 games short of what Brodeur played in his "peak". The point was not to compare the same 7 seasons between the two, it was that during his "peak" Brodeur played over 65 games every season, whereas Hasek did it just 3 times during his.

Sadly there is no way of knowing just how good Hasek was in the Czech league, but I was just merely pointing out that there's very little likelihood that his SV% was as good over there (playing in a worse league). Most players career's follow a normal bell curve where their best play is in the middle part of their career, and it's likely that Hasek was the same, which along with the worse SV% number is how I based my argument for him being worse in the Czech league then he was in the NHL.

Anonymous said...

Furthermore, you say much of Brodeur's consistency was driven by durability and teammates.

Durability yes, teammates, again he posts pretty consistent (only 0.01% difference based on your chart) numbers no matter who has been in front of him his entire career and with whatever system the team was playing (they didn't always play the neutral zone trap).

The Contrarian Goaltender said...

If you think Hasek's decline was steep, you should see what the graph would look like if I extended it to the present date. There would be a veritable Alpine ski hill on the right side to account for Brodeur's .902 thus far in 2010-11. Age-related decline happens to everybody, and trying to pass that off as a lack of consistency is not a good argument.

The point about games played is that Hasek played as much as the other top goalies in the league. Martin Brodeur is a games played outlier, there is no denying that, but that doesn't mean every other team is crazy because they didn't start their goalie 75 times per year.

As I demonstrated, compare Hasek to everyone other than Brodeur and his GP numbers are perfectly fine. Patrick Roy only started over 65 games twice. Ed Belfour did it three times, the same as Hasek. There were some pretty good long-time starting goalies who never played over 65 games (e.g. Barrasso, Vanbiesbrouck, Giguere). Yet all of them still had excellent deep playoff runs after seasons where they started most of the games, proving that they could indeed thrive despite a heavy workload.

There aren't very many reasons to start your goalie 77 times in a season. To me it would only make sense if you are a playoff bubble team playing an easy travel schedule with a great starter and a terrible backup with no future prospects within the organization. New Jersey decided to repeatedly use that strategy, to Brodeur's benefit of course, but I simply do not see that alone as making him better than goalies on other teams with different strategic goals.

it's likely that Hasek was the same, which along with the worse SV% number is how I based my argument for him being worse in the Czech league then he was in the NHL.

You're almost certainly right that Hasek was worse in the Czech league than in the NHL. It seems only logical that his play would improve when facing the better shooters playing in North America.

That said, so what? Nobody is arguing that Hasek was as good in the Czech Republic, merely that he was easily good enough to play in the NHL, and therefore would have added to his career totals if he had been given that opportunity. I don't need to make the case that Hasek had an all-time great peak over there to be justified in claiming that communism cost Hasek potential NHL career value.

Even if he only played as well as a league average goalie, he would have still added quite a bit to his legacy here if those 300+ pre-NHL games came in North America rather than in Europe. And based on what came afterwards, I think there is no doubt at all that Hasek would have been at least an average goalie in the NHL in the late '80s.

Agent Orange said...

Anon said

"0.922 = (shots faced - goals allowed)/(shots faced)"

Wouldn't the reasonable thing to do be to set this to league average over that time period? I'm not sure what that number is but using Roy's number of 0.900 over those years you get about 26.7 shots per game (I didn't have his minutes to do the per 60 calculation). This isn't an unreasonable number. Based on this it wouldn't be unreasonable to assume his sv% was north of 90%.

Thoughts?

Anonymous said...

I think you hinted above at one reason why Brodeur was able to rack up 75-game seasons... he was facing fewer shots than his contemporaries. From 93-94 to 01-02, Brodeur faced 24.8 shots/60.

Roy - 28.4
Belfour - 25.6
CuJo - 29.2
Hasek - 29.5
Osgood - 26.3 (26.1 if you only include DET)
Potvin - 29.8
Random - 28.75

"Random" is as random as I could contrive: I sorted all goalies by their number of assists that year, took the highest guy on the list who A) wasn't already on the list and B) played at least 3000 minutes (or 1500 for 1995), and tallied them as if it were one guy. (The only guy on the list twice is Sean Burke, once with the Whalers and another with the Panthers.)

None of these goalies played as many minutes overall as Brodeur, but five of them still faced more shots total. (Belfour and Osgood were well behind.) Hasek was first on the list, with our composite fellow just nine shots behind (but in 809 more minutes). Heck, Brodeur had a full year's worth of minutes on Hasek and Potvin and still faced fewer shots.

That really makes it easier to keep on going out there.

Host Pay Per Head said...

well sometimes I think that some players that are near to retirement play better than other young players and should not retire just yet

Anonymous said...

I would take Hasek over anyone 7 days a week. Best goaltender of all time by a large margin. I respect Broduer and Roy a lot but would place them at 4th and 11th G.O.A.T. respectively. There are a lot of things to factor in and I could debate for days on the subject but for all intensive purposes Hasek wins this race by a half lap.
His entire professional career should be taken into account and if he only started 8 years earlier he would own pretty much every record.

I respect everyone and their opinions but this isn't even up for argument people. Gretzky is the best player ever and Hasek is the best goalie ever.....period.