Round 2, Vote 1 (HOH Top Goaltenders)

[Hawkey Town 18]

I understand the math behind the "Brodeur's puckhandling stopped about 1 shot per game more than an average goalie" - you compare the number of shots he faces vs his backups. But that just seems too low to me. Are you adjusting for the quality of competition?

Just looking at Brodeur's 4 Vezina years - he only failed to appear in 25 games between all 4 seasons.

These are the teams where his backup played the whole season, their overall spot in the standings, and offense based on goals. (Where does one find team SOG rankings?)

2003: Lightning (12, 13), Predators (23, 28), Leafs (9, 8), Leafs (9, 8), Ducks (11, 22), Sabres (26, 25), Flames (22, 27), Penguins (29, 26) Sabres (26, 25)
2004: Leafs (5, 4) , Penguins (30, 22), Capitals (29, 27), Penguins (30, 22), Capitals (29, 27), Penguins (30, 22), Thrashers (21, 13)
2007: Predators (4, 4), Bruins (23, 23), Capitals (27 , 19), Islanders (17, 12)
2008: Panthers (22, 20), Islanders (27, 29), Penguins (4, 7), Capitals (15, 9), Bruins (13, 25)

• The average team his backups faced was 20th in the standings
• The average team his backups faced was 18th in offense
• 16 of 25 (64%) of the teams his backups faced were below average in the standings
• 16 of 25 (64%) of the teams his backups faced were below average offensively

In my opinion, your estimate underestimates the effects of Brodeur's puckhandling advantage over other goalies in two ways:

1. It doesn't take into account the fact that his backups faced teams that were below average both overall and offensively
   
2. It doesn't take into account the puck possession and shots for created by his puckhandling. Brodeur was a vital part of "The Trap." The trapping skaters would prevent the opposition from skating the puck into the zone, leaving a dump-in the only option. When Brodeur killed the opposing dump in, he not only killed the ability of the opposition to control the puck in his own zone, he created an opportunity for his team to control the puck in the opposition zone. The Devils of the era did not have a lot of skilled goal scorers, relying instead of volume shooting to score goals. To put it statistically, a goalie who is excellent at puckhandling should improve the Corsi ratings of all the skaters on his team, especially without the trapezoid.

How do we know Brodeur was facing average offenses over the course of the season? Couldn't some exceptionally bad teams in his division (whom he would play against a lot) or exceptionally good teams in the Western conference (whom he would rarely play against) skew things?

 

[Nalyd Psycho]

Not sure why this would have an effect. Devils disciplined play happened in both home and road away games.

Road teams are more likely to have questionable calls go against them.

Even if it's just 3 calls a season, that's probably 10 extra shots.

 

[tony d]

Brodeur's name is going to be interesting to see where he goes IMO. He has regressed a little over the past few seasons. It'll be interesting if he's in the middle of the list or at the bottom.

Oh and for the record the top 7 here is exactly how I thought it would work out.

 

[TheDevilMadeMe]

Sure, but you don't think a "positive" reputation among referees for disciplined play combined with the pressure of make calls in front of the home crowd could manifest itself in the count somewhere? I dunno, it seems at least plausible that the Devils could have been called for far less "marginal" penalties at home than on the road. Enough to make up enough difference to "meaningfully" impact SV% stats? That also, dunno, but seems possible.

If the Devils were called for fewer marginal penalties at home, you would expect the road team to have a lower shooting percentage in NJ. The exact opposite happens - Devils opponents have lower shot counts and higher shooting percentages in NJ than they do when the Devils play elsewhere.

Road teams are more likely to have questionable calls go against them.

Even if it's just 3 calls a season, that's probably 10 extra shots.

This explains why home teams have higher shooting percentages at home, on average.

It does nothing to explain why in certain arenas (like NJ and St Louis before the lockout), both the home and road teams have lower shot totals and higher shooting percentage than average. It does nothing to explain why in certain arenas (like Nashville both before and after the lockout), both the home and road teams have higher shot totals and lower shooting percentages than average.

 

[TheDevilMadeMe]

Brodeur's name is going to be interesting to see where he goes IMO. He has regressed a little over the past few seasons. It'll be interesting if he's in the middle of the list or at the bottom.

Oh and for the record the top 7 here is exactly how I thought it would work out.

Brodeur regressed a lot in the regular season since he was third in Vezina voting in 2009-10. But he certainly pulled it all together in the 2012 playoffs, very Terry Sawchuk in 1967 or Glenn Hall in 1968 where a goalie who looked past his prime had one last great hurrah in the playoffs.

 

[ContrarianGoaltender]

Road teams are more likely to have questionable calls go against them.

Even if it's just 3 calls a season, that's probably 10 extra shots.

Actually, 3 extra calls per season would be only 2 extra shots for an average team.

The average shots per 60 minute rates are about 29 at 5 on 5 and 49 at 5 on 5. That's an increase of 20 per 60, so a team with 6 extra minutes on the power play compared to 5 on 5 would take 2 extra shots. However, the extra shots are more difficult to stop for the goalie, so there would be a very small effect on expected save percentage.

New Jersey generally took more penalties on the road, but so did every other team (one of the biggest factors in home-ice advantage is getting more calls). I have the numbers from 1998 to 2010, and in that period the average team faced 9.3% more power plays on the road compared to at home. New Jersey faced 5.1% more on the road, the fourth-lowest difference in the league.

The average team faced 175 power plays against at home compared to 190 against on the road. New Jersey averaged 144 against at home (82.2% of average) and 153 against on the road (80.6%) of average. 82.2% of 190 would be 156, so New Jersey would have faced three more road power plays if their relative home/road discipline was the same as average. This means that the team may have been relatively more disciplined on the road, but it's a small effect. It's possible that style of play still accounts for some of the percentage differences, but it seems unlikely that the number of penalties called is all that significant.

 

[Canadiens1958]

Understood

It's not just the home teams that had lower than normal shot totals in NJ and St. Louis - it was BOTH the home and road team.

Understood. The counter was conservative both ways in his SOG calls. Does not or did not change the game outcome. Just an additional curiosity added to the residual stats. Time to move on. No big deal. Accept the necessary adjustment if need be.

 

[ContrarianGoaltender]

Data Request: Does anyone know where I can find playoff Goals-Versus-Threshold for goalies? I'm mostly interesting in career playoff GVT, since I assume that individual season GVT won't look that much different from adjusted save percentages among goalies who made the finals.

You can download the historical GVT file directly from the Hockey Prospectus website. The link's right on the front page, just below the top article and the Unfiltered section.

 

[TheDevilMadeMe]

You can download the historical GVT file directly from the Hockey Prospectus website. The link's right on the front page, just below the top article and the Unfiltered section.

Thanks. I'll be sure to use it

 

[ContrarianGoaltender]

There hasn't been a lot of talk about international play so far in this thread.

Here's a hypothetical that might be worthy of discussion: If we completely disregard Hasek's entire North American professional career, where would he rank in the top 60 all-time? His international career seems to have been glossed over by some who didn't like the way his early NHL career unfolded, but I had five international goalies on my final top 60 list and feel it would be greatly inconsistent to not put a similar value on international and European accomplishments for NHL goalies compared to non-NHL goalies (of course with proper awareness of league strength and talent pool, opportunity and the limiting factors of small sample size tournaments).

Hasek has a long list of international accomplishments including being named the best goalie in the Czech league five times, three times the top player in the Extraliga, three league championships. Youngest professional player ever, and still able to compete at a high level in the KHL at the age of 45. Widely seen as the best goalie in Europe in the late '80s. Played in three Canada Cups, four Olympics and five world championships. Named best goalie at the world championships twice (and tournament All-Star three times) and was the MVP of the 1998 Olympics. Career total of 418 games played in European leagues, plus 69 senior international tournament matches.

To me his non-NHL career is pretty obviously better than anybody not named Tretiak or Holecek. I don't know if it's OK to bring up Hasek vs. Holecek at this point since only one of them is up for debate, but I think it might be an interesting question as to which one of them should be considered better based entirely on European and international games played. If the answer is Hasek, or even if the conclusion is that they are close, then it seems that would have pretty strong implications for Hasek's overall career value given that I think many would have rated Holecek in their top 20 or 25.

Considering international performance would also be a boost for Brodeur's record and a distinct negative for Ken Dryden.

With respect to Dryden, I wanted to also bring up blogger Pat McLean's terrific breakdown of the Summit Series, which has a detailed analysis and scoring chance breakdown and provides a strong argument that poor goaltending by Ken Dryden was a major reason (perhaps the single biggest reason) for that series being as close as it was. In the four games Dryden played Canada outchanced the Soviets in all four, by a combined total of 113-75, yet were outscored 19-15 in large part because of goaltending.

Dryden's overall series numbers were well off the pace compared to the other netminders in 1972 when you look at his "scoring chance save percentage":

Dryden: 19 goals on 75 chances, .747
Tretiak: 31 goals on 208 chances, .851
Esposito: 13 goals on 83 chances, .843

It's only an eight game sample in pretty unusual circumstances, but it's definitely not a feather in Dryden's cap.

 

[Dennis Bonvie]

How do we know Brodeur was facing average offenses over the course of the season? Couldn't some exceptionally bad teams in his division (whom he would play against a lot) or exceptionally good teams in the Western conference (whom he would rarely play against) skew things?

When someone plays a team during the season can also be a factor that is not considered. Playing Columbus when they have won 3 in row for the only time all season while your team has some injuries could be a tougher game than playing a banged up Rangers team on a losing streak.

 

[quoipourquoi]

Patrick Roy's Montreal Career (Part One)

I've contextualized the expected save percentage in each round of play to the team facing the Montreal Canadiens.


1985-86

League Average: .874
Roy’s Regular Season: .875 (19th; 12th among goalies with 40 decisions)
Roy’s Playoff: .923


1st Round
Expected Save Percentage: .874
Roy’s Save Percentage: .925
Quality Games: 3-0

Roy: 27/28; .964
Riggin: 23/26

Roy: 19/21; .905
Ranford: 21/24

Roy: 28/31; .903
Ranford: 16/20


2nd Round
Expected Save Percentage: .860
Roy’s Save Percentage: .918
Quality Games: 6-1 (9-1 Total)

Roy: 19/23; .826
Liut: 26/27

Roy: 24/25; .960
Liut: 27/30

Roy: 20/21; .952
Weeks: 12/14

Roy: 22/24; .917
Weeks: 18/19

(0-1 in OT)

Roy: 20/23; .870
Weeks: 25/30

Roy: 16/17; .941
Liut: 32/32

Roy: 24/25; .960
Liut: 30/32

(1-1 in OT)


3rd Round
Expected Save Percentage: .890
Roy’s Save Percentage: .938
Quality Games: 5-0 (14-1 Total)

Roy: 26/27; .963
Vanbiesbrouck: 28/30

Roy: 19/21; .905
Vanbiesbrouck: 21/26

Roy: 44/47; .936
Vanbiesbrouck: 25/29

(2-1 in OT)

Roy: 27/29; .931
Vanbiesbrouck: 30/30

Roy: 19/20; .950
Vanbiesbrouck: 26/29


4th Round
Expected Save Percentage: .872
Roy’s Save Percentage: .897
Quality Games: 4-1 (18-2 Total)

Roy: 25/30; .833
Vernon: 22/24

Roy: 20/22; .909
Vernon: 32/35

(3-1 in OT)

Roy: 23/26; .885
Vernon: 12/16

Roy: 15/15; 1.000
Vernon: 23/24

Roy: 30/33; .909
Vernon: 29/33


1986-87

League Average: .880
Roy’s Regular Season: .892 (5th; 3rd among goalies with 40 decisions)
Roy’s Playoff: .873


1st Round
Expected Save Percentage: .875
Roy’s Save Percentage: .913
Quality Games: 3-1 (21-3 Career)

Roy: 31/33; .939
Keans: 22/28

Roy: 25/28; .893
Ranford: 21/25

(4-1 in OT)

Roy: 25/29; .862
Keans: 25/30

Roy: 35/37; .946
Ranford: 26/30


2nd Round
DID NOT PLAY FULL SERIES
Quality Games: 0-1 (3-2 Total; 21-4 Career)

Game #1
Roy: 19/26; .731
Gosselin: 27/32


3rd Round
DID NOT PLAY FULL SERIES
Quality Games: 0-1 (3-3 Total; 21-5 Career)

Game #4
Roy: 16/20; .800
Hextall: 25/28


1987-88

League Average: .880
Roy’s Regular Season: .900 (1st)
Roy’s Playoff: .890


1st Round
DID NOT PLAY FULL SERIES
Expected Save Percentage: .901
Roy’s Save Percentage: .878
Quality Games: 2-2 (23-7 Career)

Game #1
Roy: 21/24; .875
Brodeur: 9/11

Game #2
Roy: 28/31; .903
Brodeur: 21/28

Game #3
Roy: 30/33; .909
Liut: 23/27

Game #4
Roy: 36/43; .837
Liut: 26/31


2nd Round
DID NOT PLAY FULL SERIES
Expected Save Percentage: .884
Roy’s Save Percentage: .907
Quality Games: 2-1 (4-3 Total; 25-8 Career)

Game #3
Roy: 32/35; .914
Lemelin: 22/23

Game #4
Roy: 22/23; .957
Lemelin: 22/22

Game #5
Roy: 24/28; .857
Lemelin: 28/29


1988-89

League Average: .879
Roy’s Regular Season: .908 (1st)
Roy’s Playoff: .920


1st Round
DID NOT PLAY FULL SERIES
Expected Save Percentage: .878
Roy’s Save Percentage: .931
Quality Games: 3-0 (28-8 Career)

Roy: 25/27; .926
Whitmore: 27/33

Roy: 27/29; .931
Sidorkiewicz: 14/17

Roy: 43/46; .935
Whitmore: 36/40

(5-1 in OT)


2nd Round
DID NOT PLAY FULL SERIES
Expected Save Percentage: .885
Roy’s Save Percentage: .907
Quality Games: 3-1 (6-1 Total; 31-9 Career)

Roy: 26/28; .929
Moog: 12/15

Roy: 27/29; .931
Lemelin: 38/41

(6-1 in OT)

Roy: 22/26; .846
Lemelin: 23/28

Roy: 22/24; .917
Moog: 24/27


3rd Round
Expected Save Percentage: .871
Roy’s Save Percentage: .940
Quality Games: 6-0 (12-1 Total; 37-9 Career)

Roy: 23/26; .885
Wregget: 24/25

Roy: 24/24; 1.000
Wregget: 36/39

Roy: 25/26; .962
Wregget: 21/26

Roy: 17/17; 1.000
Hextall: 29/32

Roy: 21/23; .913
Hextall: 27/28

(6-2 in OT; six-game OT winning streak snapped)

Roy: 15/17; .882
Hextall: 22/26


4th Round
Expected Save Percentage: .872
Roy’s Save Percentage: .908
Quality Games: 5-1 (17-2 Total; 42-10 Career)

Roy: 32/35; .914
Vernon: 29/31

Roy: 30/32; .938
Vernon: 19/23

Roy: 34/37; .919
Vernon: 31/35

(7-2 in OT)

Roy: 31/34; .912
Vernon: 17/19

Roy: 25/28; .893
Vernon: 26/28

Roy: 15/18; .833
Vernon: 20/22


1989-90

League Average: .881
Roy’s Regular Season: .912 (1st)
Roy’s Playoff: .911


1st Round
Expected Save Percentage: .882
Roy’s Save Percentage: .923
Quality Games: 4-2 (46-12 Career)

Roy: 24/28; .857
Puppa: 34/35

Roy: 26/26; 1.000
Puppa: 33/36

Roy: 30/31; .968
Puppa: 31/33

(8-2 in OT)

Roy: 23/27; .852
Puppa: 31/33

Roy: 30/32; .938
Puppa: 22/25

Roy: 22/24; .917
Puppa: 24/28


2nd Round
Expected Save Percentage: .889
Roy’s Save Percentage: .894
Quality Games: 3-2 (7-4 Total; 49-14 Career)

Roy: 27/28; .964
Moog: 20/20

Roy: 19/24; .792
Moog: 30/34

Roy: 8/12; .667
Moog: 20/23

Roy: 28/29; .966
Moog: 32/35

Roy: 28/30; .933
Moog: 23/24


1990-91

League Average: .886
Roy’s Regular Season: .906 (2nd; 2nd among goalies with 40 decisions)
Roy’s Playoff: .898


1st Round
Expected Save Percentage: .879
Roy’s Save Percentage: .874
Quality Games: 2-4 (51-18 Career)

Roy: 24/29; .828
Puppa: 10/15

Roy: 23/27; .852
Puppa: 26/31

Roy: 28/33; .848
Malarchuk: 18/22

Roy: 26/30; .867 (Racicot: 11/13; Decision)
Malarchuk: 33/37

Roy: 29/32
Malarchuk: 24/28

(9-2 in OT)

Roy: 23/24; .958
Malarchuk: 24/29


2nd Round
Expected Save Percentage: .881
Roy’s Save Percentage: .918
Quality Games: 6-1 (8-5 Total; 59-23 Career)

Roy: 23/25; .920
Moog: 18/19

Roy: 41/44; .932
Moog: 33/37

(10-2 in OT)

Roy: 23/26; .885
Moog: 39/41

Roy: 23/25; .920
Moog: 22/27

Roy: 28/32; .875
Moog: 18/19

Roy: 32/24; .941
Moog: 38/41

(11-2 in OT)

Roy: 31/33; .939
Moog: 35/36


1991-92

League Average: .888
Roy’s Regular Season: .914 (1st)
Roy’s Playoff: .904


1st Round
Expected Save Percentage: .891
Roy’s Save Percentage: .918
Quality Games: 4-3 (63-26 Career)

Roy: 32/32; 1.000
Pietrangelo: 28/30

Roy: 15/17; .882
Pietrangelo: 23/27

Roy: 28/33; .848
Pietrangelo: 32/34

Roy: 33/35; .943
Pietrangelo: 24/25

Roy: 19/23; .826
Pietrangelo: 23/29; .793

Roy: 24/26; .923
Pietrangelo: 42/43

(11-3 in OT; five-game OT winning streak snapped)

Roy: 39/41; .951
Pietrangelo: 53/56

(12-3 in OT)


2nd Round
Expected Save Percentage: .899
Roy’s Save Percentage: .876
Quality Games: 1-3 (5-6 Total; 64-29 Career)

Roy: 21/27; .778
Moog: 24/28

Roy: 21/24; .875
Moog: 20/22

(12-4 in OT)

Roy: 21/24; .875
Moog: 24/26

Roy: 29/30; .967
Moog: 26/26


1992-93

League Average: .885
Roy’s Regular Season: .894 (8th; 6th among goalies with 42 decisions)
Roy’s Playoff: .929


1st Round
Expected Save Percentage: .861
Roy’s Save Percentage: .936
Quality Games: 6-0 (70-29 Career)

Roy: 34/37; .919
Hextall: 36/38

(12-5 in OT)

Roy: 34/37; .919
Hextall: 32/33

Roy: 34/35; .971
Hextall: 48/50

(13-5 in OT)

Roy: 25/27; .926
Hextall: 34/37

Roy: 35/37; .946 (Racicot: 7/9)
Hextall: 25/30

(14-5 in OT)

Roy: 28/30; .933
Hextall: 18/23


2nd Round
Expected Save Percentage: .875
Roy’s Save Percentage: .915
Quality Games: 4-0 (10-0 Total; 74-29 Career)

Roy: 32/35; .914
Fuhr: 18/22

Roy: 28/31; .903
Fuhr: 21/25

(15-5 in OT)

Roy: 33/36; .917
Fuhr: 23/27

(16-5 in OT)

Roy: 37/40; .925
Fuhr: 26/30

(17-5 in OT)


3rd Round
Expected Save Percentage: .871
Roy’s Save Percentage: .932
Quality Games: 5-0 (15-0; 79-29 Career)

Roy: 20/21; .952
Healy: 18/21

Roy: 39/42; .929
Healy: 37/41

(18-5 in OT)

Roy: 31/32; .969
Healy: 21/23

(19-5 in OT)

Roy: 21/24; .875
Healy: 23/24

Roy: 26/28; .929
Healy: 28/33


4th Round
Expected Save Percentage: .882
Roy’s Save Percentage: .929
Quality Games: 5-0 (20-0 Total; 84-29 Career)

Roy: 34/37; .919
Hrudey: 31/32

Roy: 22/24; .917
Hrudey: 38/41

(20-5 in OT)

Roy: 30/33; .909
Hrudey: 32/36

(21-5 in OT)

Roy: 40/42; .952
Hrudey: 36/39

(22-5 in OT)

Roy: 18/19; .947
Hrudey: 25/29


1993-94

League Average: .895
Roy’s Regular Season: .918 (3rd; 3rd among goalies with 42 decisions)
Roy’s Playoff: .930


1st Round
DID NOT PLAY FULL SERIES
Expected Save Percentage: .903
Roy’s Save Percentage: .930
Quality Games: 3-3 (87-32 Career)

Roy: 25/28; .893
Casey: 24/26

Roy: 40/42; .952
Casey: 21/24

Misses Game #3 (Appendicitis)

Roy: 39/41; .951
Riendeau: 10/15

Roy: 60/61; .984
Casey: 34/36

(23-5 in OT; Montreal career ends on an 11-game OT winning streak)

Roy: 22-25; .880
Casey: 20/22

Roy: 26/31; .839
Casey: 26/29


1994-95

League Average: .901
Roy’s Regular Season: .906 (14th; 11th among goalies with 24 decisions)



Career

Save Percentage
(1, 1, 1, 1, 2, 3, 5, 8*, 14, 19*)
*Conn Smythe Season

Save Percentage (among goalies with 50% of their team’s decisions)
(1, 1, 1, 1, 2, 3, 3, 6*, 11, 12*)
*Conn Smythe Season


Overtime: 23 Wins, 5 Losses

Quality Games: 87-32

Quality Series (Positive): 18
Quality Series (Even): 2
Quality Series (Negative): 4

Series (Quality Games)
6-0 (1989 PHI, 1993 QUE)
5-0 (1986 NYR, 1993 NYI, 1993 LAK)
6-1 (1986 HFD, 1991 BOS)
4-0 (1993 BUF)
5-1 (1989 CGY)
3-0 (1986 BOS, 1989 HFD)
4-1 (1986 CGY)
3-1 (1987 BOS, 1989 BOS)
4-2 (1990 BUF)
2-1 (1988 BOS)
3-2 (1990 BOS)
4-3 (1992 HFD)
3-3 (1994 BOS)
2-2 (1988 HFD)
0-1 (1987 QUE)
0-1 (1987 PHI)
2-4 (1991 BUF)
1-3 (1992 BOS)


From 1985-86 to 1993-94, Patrick Roy was a de facto top-three save percentage starter in seven seasons. On top of those seven seasons, each of the two remaining seasons ended with Patrick Roy winning the Conn Smythe Trophy. I would argue that this nine-year stretch alone is more than enough to earn him a place in the HOF. I will be examining his Colorado play and making an assessment as to whether it was mere compiling on to an already HOF career, or a distinct HOF career in its own right.

 

[Nalyd Psycho]

This explains why home teams have higher shooting percentages at home, on average.

It does nothing to explain why in certain arenas (like NJ and St Louis before the lockout), both the home and road teams have lower shot totals and higher shooting percentage than average. It does nothing to explain why in certain arenas (like Nashville both before and after the lockout), both the home and road teams have higher shot totals and lower shooting percentages than average.

Uhm, were exactly did I argue against different arenas having different shot counting methods. Please don't jump to radical conclusions about what people are saying, it only hurts discourse.

 

[Canadiens1958]

Let's Extrapolate

There hasn't been a lot of talk about international play so far in this thread.

Here's a hypothetical that might be worthy of discussion: If we completely disregard Hasek's entire North American professional career, where would he rank in the top 60 all-time? His international career seems to have been glossed over by some who didn't like the way his early NHL career unfolded, but I had five international goalies on my final top 60 list and feel it would be greatly inconsistent to not put a similar value on international and European accomplishments for NHL goalies compared to non-NHL goalies (of course with proper awareness of league strength and talent pool, opportunity and the limiting factors of small sample size tournaments).

Hasek has a long list of international accomplishments including being named the best goalie in the Czech league five times, three times the top player in the Extraliga, three league championships. Youngest professional player ever, and still able to compete at a high level in the KHL at the age of 45. Widely seen as the best goalie in Europe in the late '80s. Played in three Canada Cups, four Olympics and five world championships. Named best goalie at the world championships twice (and tournament All-Star three times) and was the MVP of the 1998 Olympics. Career total of 418 games played in European leagues, plus 69 senior international tournament matches.

To me his non-NHL career is pretty obviously better than anybody not named Tretiak or Holecek. I don't know if it's OK to bring up Hasek vs. Holecek at this point since only one of them is up for debate, but I think it might be an interesting question as to which one of them should be considered better based entirely on European and international games played. If the answer is Hasek, or even if the conclusion is that they are close, then it seems that would have pretty strong implications for Hasek's overall career value given that I think many would have rated Holecek in their top 20 or 25.

Considering international performance would also be a boost for Brodeur's record and a distinct negative for Ken Dryden.

With respect to Dryden, I wanted to also bring up blogger Pat McLean's terrific breakdown of the Summit Series, which has a detailed analysis and scoring chance breakdown and provides a strong argument that poor goaltending by Ken Dryden was a major reason (perhaps the single biggest reason) for that series being as close as it was. In the four games Dryden played Canada outchanced the Soviets in all four, by a combined total of 113-75, yet were outscored 19-15 in large part because of goaltending.

Dryden's overall series numbers were well off the pace compared to the other netminders in 1972 when you look at his "scoring chance save percentage":

Dryden: 19 goals on 75 chances, .747
Tretiak: 31 goals on 208 chances, .851
Esposito: 13 goals on 83 chances, .843

It's only an eight game sample in pretty unusual circumstances, but it's definitely not a feather in Dryden's cap.

Your appreciation of European Goalies is interesting.

69 Senior International Tournament games still ranks Hasek well behind Viktor Konovalenko - 117, Nikolay Puchkov 90, Vladimir Myshkin 87, without looking at other Swedish, Finnish, Czech, American. The internal European leagues awarded few honours so the honours and awards columns in their cases are sparse. They do have a nice collection of Olympic and WHC medals mainly gold. Same could be said for Seth Martin - Canadian, well viewed international, but no Canadian Senior recognition since the leagues never bothered with such expenses and sponsors could not be found.

Glad to consider a Dryden/Hasek comparison but then it reduces to a 6th/7th place discussion.

 

[seventieslord]

That all reminds me, there was a thread not too long ago about players who had 2 hhof careers. Bourque, orr and roy were the three favourites IIRC. Someone even argued roy might have 3 hhof careers if cut off strategically, I think.

 

[Rob Scuderi]

Your appreciation of European Goalies is interesting.

69 Senior International Tournament games still ranks Hasek well behind Viktor Konovalenko - 117, Nikolay Puchkov 90, Vladimir Myshkin 87, without looking at other Swedish, Finnish, Czech, American. The internal European leagues awarded few honours so the honours and awards columns in their cases are sparse. They do have a nice collection of Olympic and WHC medals mainly gold. Same could be said for Seth Martin - Canadian, well viewed international, but no Canadian Senior recognition since the leagues never bothered with such expenses and sponsors could not be found.

Glad to consider a Dryden/Hasek comparison but then it reduces to a 6th/7th place discussion.

Shouldn't we care more about how they fared while representing their country than the amount of times they did?

I understand the awards are sparse (even though we get two different sets of people providing all-star awards) but it doesn't seem helpful to count appearances and gloss over the performances (e.g. Konovalenko ranks ahead of Hasek).

I agree though, it's hard to meaningfully discuss international play with these candidates. What can we say about the pre-expansion guys? Plante played great for one game coming out of retirement on a team of juniors against the Soviet national team. Sawchuk and Hall never got their shot like many of their contemporaries so that leaves them in the dark.

------------------
To take this a step further, Dryden clearly fared poorly against the Soviet national team playing behind a team of NHL all-stars. Can we make anything of Plante's performance against the Soviets on a much lesser team? Or is this dismissible based on different ways they played in the net/not particularly meaningful?

Of course there's the major difference that Plante played only one game unlike Dryden.

 

[vadim sharifijanov]

That all reminds me, there was a thread not too long ago about players who had 2 hhof careers. Bourque, orr and roy were the three favourites IIRC. Someone even argued roy might have 3 hhof careers if cut off strategically, I think.

hmm, maybe.


'86-'89: conn smythe/cup, second finals, vezina/1st all-star team, 2nd all-star team, three jennings. short career, but hard to leave him out. 132 wins in 240 games. 35 wins in 53 playoff games. still good for top 30 all time in playoff wins.


'90-'95: conn smythe/cup, vezina/1st all-star team, 2nd all-star team, jennings. 175 wins in 342 games. 35 wins in 61 playoff games. again, very short. but both stretches aren't far off from tim thomas' regular season career, and both top his playoff resume. consider also in this second stretch, roy's save percentage wins over the next highest guy in the league were enormous. not as enormous as hasek's best years would be, but still incredibly dominant. also, of course, the '93 cup run was the stuff of legend.


'96-'03: conn smythe, two cups, 1st team all-star, jennings. 274 wins in 500 games. 81 wins in playoff 133 games. other than the smythe, the individual hardware isn't there the way it is in the first two stretches. but he would be 36th in all-time regular season wins, and a staggering 6th in playoff wins (exactly one cup run more than hasek's entire career).

 

[pluppe]

Montreal Canadiens defensive dicipline 86-92

It is well known that Brodeur has had the opportunity to play behind a diciplined defence for large parts of his career. But I was surprised to see that Roy had similar support during the start of his career.

Between 85/86 and 91/92 Roy finished 14th, 5th, 1st, 1st, 1st, 2nd, 1st in the save % race. During these seasons Montreal finished the season with the least amount of power play opportunities against 5 times and second least 2 times. Often with much lower numbers than the second least. They also ranked consistently low in power play opportunities which might also results in tougher shots against in forms of breakaways. These are the years of Roys best save % placements. It is also interesting to note that in 92/93 when Montreal finished with the 8th least PPOA Roy fell to 5th in save % again after 5 consecutive years in the top-2.

Buffalo ranked 22nd, 10th, 4th, 3rd, 3rd, 4th in most power play opportunities against during Haseks 6 consecutive save % titles. They also consistently had more power play opportunities than Roys Canadians.

This indicates that Roys save % numbers from his peak could be inflated. I know that late in his career he did rank high in even strengh save % but does anybody have his even strengh save % from these years compared to the rest of the league.

It also indicates that Haseks titles could be even more impressive than the numbers show.

 

[overpass]

So why are you responding to me? I have time and time again said that we should be careful about making to sure deductions because random chance plays a role.



And then you go on quoting a article I never linked too. And saying that I "attribute all variation in past performance to random chance". And try to claim that it´s not a response to my words but to a philosophy. Like I said, please reread all my posts and then tell me that I am not saying that it is a combination of thing where chance plays a part.



Thanks for leaving out the following line when quoting "I´m not saying it is true in this case but it´s worth mentioning." I think it is worth mentioning and you can see why in my answer to qpq below. And again: Where is this my philosophy?

I think you and MasterOfDistricts are fine posters and worth responding to. I don't mean to denigrate you by responding to you. I simply had something to say and wanted to get it across clearly in one post rather than breaking it up into point-by-point responses to different posts.

I'm quoting an article that you never linked to because someone else linked to it, and because it's a good example of the "statistical philosophy" that all performance only has value as a sample of ability, rather than having value as an end in itself. You have not stated this explicitly but it is implied in all your posts when you talk about small samples and random variation.

You've asked why I have responded to you specifically. Here's one of your previous posts that explains why.

Bolded: No it doesn´t . Numbers will always vary in these cases because of random variance. The question is if there is additional variance because of differences in ability. Otherwise the fact that 2 goalies have different quality starts % for 100 games says nothing about what is likely for the next 100.

(For context you were responding to a post that said "Er doesn't the fact that different goalies have different "quality starts" numbers kind of prove that some are more likely to have extreme performances than others?" Which was itself a response to a post saying "There's no evidence that some goalies are more likely to have extreme performances than others, relative to skill level." So you did not establish the "likely" framework, but you are posting within it. I disagree with the whole "likely" framework as it applies to this project. IMO it's about evaluating past performance, not projecting into the future.)

Anyway, I saw this post as fitting with the "statistical philosophy" I was attacking. In order to be more specific about how I disagree with it, I have reworded your paragraph to reflect my views. Hope this helps.

Numbers will always vary in these cases because of random variance specific factors unaccounted for in the model such as home/road, support from the team in front of the player, and opposing shooters making their shots. The question is if there is additional variance because of differences in ability performance. Otherwise the fact that 2 goalies have different quality starts % for 100 games says nothing about what is likely for the next 100 their actual performance was for those 100 games.

Now I realize that you have primarily been saying that random variation may play a role in performance distribution, as opposed to MasterOfDistricts saying it does play a role. But since I disagree that random variation is a factor when assessing historical performance, I disagree with your view as well. And in fact, in pointing to sample sizes in this thread, you are implicitly saying that random variation does play a role in performance distribution.

Of course they won´t. But that does not make it not true. Goalies like all people are affected by attributional biases. So? Or are you claiming that there is no such thing? Is the project not about discussing these things with as much context as possible. If we can´t mention that chance plays a role (which I remind you is all I have done) then I think you are missing out on important information.

(I also think that if we did evaluate them based on the goaltenders own perception, Roy would be the runaway winner)

at your Roy comment. How would these seven goaltenders rank by this evaluation method?

But seriously, I'm not only talking about the views of specific goaltenders here, or even goaltenders in general. I'm saying that our universal human experience* is that we perform better at some times that others (at our jobs, in our relationships, in life), and that we are responsible for our good performance and our bad performance. As human beings, goaltenders experience their career in this way. I think it is extremely wrong to call a universal human experience an attribution error because someone has run a crude statistical model that hasn't found a large difference between actual distribution and random distribution. Attribution errors may exist but I think there is a high burden of proof for statistical methods when attempting to go against human experience.

*Or if it's not universal it's the dominant way that people have experienced the world, and much better than blaming any variation in performance on the whims of the gods, or randomness, or whatever.

 

[Canadiens1958]

My Points

Shouldn't we care more about how they fared while representing their country than the amount of times they did?

I understand the awards are sparse (even though we get two different sets of people providing all-star awards) but it doesn't seem helpful to count appearances and gloss over the performances (e.g. Konovalenko ranks ahead of Hasek).

I agree though, it's hard to meaningfully discuss international play with these candidates. What can we say about the pre-expansion guys? Plante played great for one game coming out of retirement on a team of juniors against the Soviet national team. Sawchuk and Hall never got their shot like many of their contemporaries so that leaves them in the dark.

------------------
To take this a step further, Dryden clearly fared poorly against the Soviet national team playing behind a team of NHL all-stars. Can we make anything of Plante's performance against the Soviets on a much lesser team? Or is this dismissible based on different ways they played in the net/not particularly meaningful?

Of course there's the major difference that Plante played only one game unlike Dryden.

My points exactly:

http://www.hockey-reference.com/players/h/hasekdo01.html

Dominik Hasek, other than 1998 Olympics there is little of note internationally, Canada Cup is 3W- 10L-2T. Not even a Vladimir Dzurilla like cameo 1976 great performance. Holecek vs the Soviets. Jack McCartan 1960, Jim Craig 1980, Seth Martin 1961 are all legit comparisons to Hasek 1998.

Plante 1965 with the Junior Canadiens reinforced by 5 minor pros against the Soviet Nationals.Was at the game. What matters is that out of retirement, within ten minutes - complete line rotations for the Soviets, Plante had adjusted his spots and timing to their game. Tretiak did the same in Gane 1 of the 1972 Summit Series.Down 2-0 wuthin six minutes he made little positional adjustments. Dryden never did. Whether a goalie is playing with stars or scrubs the recognition and adjustments either come quickly or they do not. Hasek took about 4 years to adjust to the NHL. Goalie recognizes the key elements very quickly or he does not.

Saw Sawchuk and Hall enough and how quickly they adjusted to new teams and circumstances - quickly. They did not have Plante's puckhandling skills which threw the Soviets a bit.

Both Hasek and Dryden brought plenty of great attributes to the rink but they also had notable shortcomings.

 

[overpass]

Others have posted on Ken Dryden's lack of success in international play. And it turns out that, as with every other topic, Ken Dryden has written about this. Here are a couple of excerpt from Dryden's pen, published in Sports Illustrated. Bolded the parts about Dryden's style specifically.

March 26, 1973 - The Rinks Were Running Red

SEPTEMBER 24: Ever feel so nervous, so on edge that you almost can't stand up? That's the way I feel right now—and it's only a quarter to nine in the morning, still 11 hours away from game time. I haven't felt this shaky in years. The fear of failure is wicked. I did not have any such feelings even the night before my first Stanley Cup game against Boston in 1971. After all, I had played six NHL games at the end of the regular season and we had won all six. But my record against the Russians....

I wonder how the other players feel about Harry's decision to play me again. I'm sure they think Tony should be in goal tonight, and I can't really blame them. I haven't given them many reasons to believe that all of a sudden I'll play well against the Russians. But now that I think about it, there are a few reasons why I might:

1) I was rusty in Canada but now I have two games and a month of practice under my belt.

2) I have changed my style to accommodate the strong points of the Soviet game. Mentally, at least, this seems to have caused a big improvement in my performances. Instead of coming out of the net as I did in Canada—only to get hit by the Russians' short-passing game—I'll stay closer in. Although I still haven't tried the new style in a game, I have developed a quiet confidence in it. At first it was new and I didn't know if I could play that way or indeed if it was the right way. Now I'm certain; so certain, in fact, that I know that under pressure I will not revert back to the old Dryden style.

3) The team is playing much better.

October 8, 1979 - A Game In Search Of Some Contests

While most fans have chosen to remember the Soviet victory over the NHL All-Stars in last February's Challenge Cup as more decisive than it was, it is at least arguable that it represented for us the first step down the other side of the competitive mountain. The second and third games of that series exposed the dangerously one-dimensional nature of our game. Until that time, Soviet-NHL games had assumed a predictable pattern. The NHL would generate a wide territorial edge in play, continuously applying pressure in the Soviet zone; the Soviets would counter with Vladislav Tretiak's strong goaltending and a quick-breaking, opportunistic offense. But in the Challenge Cup the Soviets widened the scope of their offense—adding pursuit to their possession style—and improved their defensive game by adding quickness, toughness and patience. And they took away our game.

Simply doing what we've been doing—only better—may not be good enough. Before the Challenge Cup I thought back on the mixed results that I had had in games against the Soviet team, and I wondered if I was simply not made for the quickness of the international game. I asked myself, "Am I a dinosaur? Am I simply too big to adapt?" After the most recent series, perhaps the more appropriate question was: "Are we all dinosaurs? Are we all traveling on our bellies when others are using wings?"

Also, a comment on TCG's posting the scoring chance numbers from 1972. I think that series breakdown is very cool, but I'm hesitant to take the scoring chance numbers as gospel. They have been recorded by observers who define scoring chances based on NHL hockey of the last few years, which is far more stylistically homogenous than the 1972 Summit Series. Specifically I wonder if they give enough credit to the Soviets creating chances with lateral movement and passing, since they define chances based on shots taken from a specific area of the ice.

While this may affect a comparison of Canadian goaltending to Soviet goaltending as a whole, it says something that Esposito's numbers were so much better than Dryden's while playing behind the same team.

 

[quoipourquoi]

Does anyone have the list of international goaltenders Hasek was competing against for his All-Star World Championships and the list of international goaltenders Salo was competing against for his?

 

[overpass]

It is well known that Brodeur has had the opportunity to play behind a diciplined defence for large parts of his career. But I was surprised to see that Roy had similar support during the start of his career.

Between 85/86 and 91/92 Roy finished 14th, 5th, 1st, 1st, 1st, 2nd, 1st in the save % race. During these seasons Montreal finished the season with the least amount of power play opportunities against 5 times and second least 2 times. Often with much lower numbers than the second least. They also ranked consistently low in power play opportunities which might also results in tougher shots against in forms of breakaways. These are the years of Roys best save % placements. It is also interesting to note that in 92/93 when Montreal finished with the 8th least PPOA Roy fell to 5th in save % again after 5 consecutive years in the top-2.

Buffalo ranked 22nd, 10th, 4th, 3rd, 3rd, 4th in most power play opportunities against during Haseks 6 consecutive save % titles. They also consistently had more power play opportunities than Roys Canadians.

This indicates that Roys save % numbers from his peak could be inflated. I know that late in his career he did rank high in even strengh save % but does anybody have his even strengh save % from these years compared to the rest of the league.

It also indicates that Haseks titles could be even more impressive than the numbers show.

Good point. Those Montreal teams were the forerunners of the Devils when it came to team defence and discipline.

 

[Rob Scuderi]

Does anyone have the list of international goaltenders Hasek was competing against for his All-Star World Championships and the list of international goaltenders Salo was competing against for his?

Eliteprospects is a good site to use when looking for international competition info. You can see all the goalies Hasek competed against, but it's missing the stats for some.

 

[Rob Scuderi]

My points exactly:

http://www.hockey-reference.com/players/h/hasekdo01.html

Dominik Hasek, other than 1998 Olympics there is little of note internationally, Canada Cup is 3W- 10L-2T. Not even a Vladimir Dzurilla like cameo 1976 great performance. Holecek vs the Soviets. Jack McCartan 1960, Jim Craig 1980, Seth Martin 1961 are all legit comparisons to Hasek 1998.

Why exactly are we dismissing Hasek's accomplishments at the World Championships?