HFBoards

Go Back   HFBoards > General Hockey Discussion > National Hockey League Talk
National Hockey League Talk Discuss NHL players, teams, games, and the Stanley Cup Playoffs.

Would Brodeur be a HOF Goalie Without the Devils?

Reply
 
Thread Tools
Old
01-14-2010, 04:12 PM
  #601
Chalupa Batman
Mod Supervisor
 
Chalupa Batman's Avatar
 
Join Date: Sep 2005
Posts: 23,164
vCash: 500
Quote:
Originally Posted by MadDevil View Post
As long as we're crunching numbers on Marty, clearly we need to take into account the teams he was playing against, the players on them, what system they were using, whether it was a back-to-back game, what phase the moon was in, and whether or not he had a bowel movement prior to the game. These could all skew his numbers one way or the other.
Whether or not you're willing to take this seriously, the first four things you mention are things which should be considered in this exercise.

Chalupa Batman is offline   Reply With Quote
Old
01-14-2010, 04:18 PM
  #602
haakon84
Registered User
 
Join Date: Dec 2003
Posts: 2,212
vCash: 500
Quote:
Originally Posted by seventieslord View Post
This is interesting stuff; nice work.

My position has been that there is no evidence to suggest facing more or less shots should be considered a factor in assessing sv%. Shot quality and PP time should be considered more. This hints that there can be a correllation, at least for these two goalies. I would sure like to see more of this.

Did you use only complete games? Games where they got pulled after allowing 4 goals on 15 shots would obviously skew the "under 20" category greatly.

Like any study, this seems to just bridge the gap slightly and never quite get Brodeur to that Roy/Hasek level. (i.e. gets him within 4-5 sv% points) Shot prevention studies do the same thing. As do studies about lack of shot recording.

A PP study could actually take things in the opposite direction, as you and I have discussed before. Shot quality studies, prior to 2007, also seemed to hint that Brodeur's shot quality was well below average.

But like I said, good work.
Thanks it gets much more interesting when you throw Roy's post 1993 numbers into the mix.

I used every appearance so the under 20 is definitely skewed but it's very similar save % for all three goaltenders.


haakon84 is offline   Reply With Quote
Old
01-14-2010, 04:21 PM
  #603
MadDevil
Registered User
 
Join Date: Feb 2007
Location: Bismarck, ND
Country: United States
Posts: 6,721
vCash: 500
Quote:
Originally Posted by Doctor No View Post
Whether or not you're willing to take this seriously, the first four things you mention are things which should be considered in this exercise.
Sure, and if somebody can go back and find those stats on every goaltender to play, we can have a discussion about it. The problem is, not everybody on the "greatest" list played in the same era, under the same rules, with the same equipment, etc. So it's basically a useless (although sometimes entertaining) exercise to begin with. Statistics are not the be all end all that some seem to think they are, especially with goalies.

The bottom line to me is that Brodeur has won more games (at least in the regular season) than any other goaltender in history, and the distance is getting greater. Not to mention the most shutouts in history. Something tells me that's not all the team in front of him. I'll give you this much, Brodeur is lucky that he had a GM running the team that recognized his abilities and built smartly around him. That being said, I truly believe Brodeur would have found success even if he did play for another team. The problem is, we'll never know if he would have put up the stats he has with another team or not. Just as we don't know if Roy or Hasek would have had more or less success if circumstances had been different.

MadDevil is offline   Reply With Quote
Old
01-14-2010, 04:26 PM
  #604
MathMan
Registered User
 
Join Date: Jan 2006
Posts: 16,940
vCash: 500
Quote:
Originally Posted by MadDevil View Post
The bottom line to me is that Brodeur has won more games (at least in the regular season) than any other goaltender in history, and the distance is getting greater. Not to mention the most shutouts in history.
Except wins and shutouts are both team stats. Yes, the goalie counts for a significant chunk of the team, but so does the #1 center and I don't see anyone bringing up Gretzky's or Lemieux's win percentages or shutouts when trying to decide which was better. It makes exactly the same amount of sense. Crediting goalies with a win-loss record was one of the most boneheaded decisions in hockey.

Any argument that rests on evaluating goaltenders based on wins or shutouts is so inherently flawed as to be useless. It's highly unfortunate that so much evaluation of goalies occurs on that basis considering how invalid it is.

If you want to use statistics to make any argument about a goalie's ability, use some sort of goaltending statistic.

MathMan is offline   Reply With Quote
Old
01-14-2010, 04:31 PM
  #605
Chalupa Batman
Mod Supervisor
 
Chalupa Batman's Avatar
 
Join Date: Sep 2005
Posts: 23,164
vCash: 500
Quote:
Originally Posted by MadDevil View Post
Sure, and if somebody can go back and find those stats on every goaltender to play, we can have a discussion about it. The problem is, not everybody on the "greatest" list played in the same era, under the same rules, with the same equipment, etc. So it's basically a useless (although sometimes entertaining) exercise to begin with. Statistics are not the be all end all that some seem to think they are, especially with goalies.
Agreed with the bolded part. There are people out there (not you) who seem to think that all I do is look at Excel spreadsheets all day, and that I've never watched a hockey game in my life. And that's not the case.

The same people also consider it a slight when I say that Brodeur is a top-five goaltender of all time - they act as if I'd just called him Gregg Naumenko or something. Rest assured that I have a lot of respect for what Martin Brodeur has accomplished in his career thus far, and am looking forward to seeing what he will do with the remainder.

(I also have a great deal of respect for Gregg Naumenko's work - anyone who's been in the show has more on-ice talent than I could ever dream of having. But that's a different thread)

Statistics are a tool - they have an advantage of being memoryless (all of us have flawed memories of what we've seen, and none of us have seen every game played by every goaltender in history - or even the last N years). A lot of people have thrown out (usually derogatory) metaphors for statistics in this thread, so I'll give mine as well: statistics are like lightpoles. You can use them for support (as a drunk would), or you can use them for illumination (I prefer the latter).

Chalupa Batman is offline   Reply With Quote
Old
01-14-2010, 05:08 PM
  #606
seventieslord
Moderator
 
seventieslord's Avatar
 
Join Date: Mar 2006
Location: Regina, SK
Country: Canada
Posts: 23,846
vCash: 500
Quote:
Originally Posted by MoonDragn View Post
Look at the stats, they are mostly not there, they aren't displayed correctly.
You will have to better explain to me what is not showing up correctly. I checked 2002 and 2007 and everything looks fine. 1997-98's numbers are wacky but other than that these seem fine.

Quote:
I still wouldn't use the PPOA stat, because some teams face more shots on the PP than others on average. A good example is that same year in 2007-2008, Buffalo had more penalty minutes but faced less shots on net than NJ during the PP.
So they were better at preventing shots against while killing penalties. This doesn't change the fact that the more penalties they take, the higher percentage of their goalie's shots will be from a PP.

Quote:
As I mentioned before, I think it would be more meaningful to calculate career SV% for normal 5-5, pp and sh. But it still doesn't take into account all the other factors like deflected shots, breakaways etc.
You may as well include SH with ES because they seem to have about the same sv% anyway, and SH shots are infrequent and random. Then you're dealing with PP and non-PP which makes it so much easier.

It absolutely would be more meaningful to calculate PP and non-PP sv% (not on career, year by year, otherwise it is too subject to fluctuation) but when you're done doing that, then you can start to play with the numbers and see what each goalie's overall sv% would look like if their diet included a "standardized" dosage of PP shots.

And we can easily do this for all goalies going back to 1999 because the data exists. it is before that where estimations have to come in. They would not be perfect but the assumptions that have to be made for such a system to work are more than reasonable.

Quote:
I was calculating the percentage of shots on pp vs shots on 5-5 not against the total shots faced.
OK. That was incorrect. I gave you the correct percentages.

Quote:
Again, PPOA doesn't show how many shots faced during each PP, or total shots faced. If you only adjust based on PP/G, you would be off in your numbers vs PP shots/G.
If you have a good reason to believe there would be a such a variance in this between two goalies on the same team, please share it.

Second, I assume you're talking about pre-1999? We don't need to speculate on post-1999 situational shots. We have the data.

For pre-1999 seasons I just demonstrated that in a very extreme circumstance you could find yourself as much as 9% off on shots that comprise as much as 25% of a goalie's total shots. So about 2.25% off in total... in an extreme case like this. Like I said, it's not perfect, but it is more than reasonable. No one is ever going to claim a bastardized stat like that PP-adjusted sv% is authoritative, but it is closer to the truth, not further.

Quote:
I don't remember how many games each division played each other and against the rest of the league. Perhaps that accounts for some of the differences in SV%? When you face the same teams more than others, you tend to play better/worse against those teams.
Up until 1994, I think every team played every other team at least twice and I don't recall divisional matchups being more than 6 times a season. In any case, all you have to do to see how minor this is, is pick a year where one division was higher scoring than others, look at the schedule of one of the teams in that division, calculate the average GPG of their opponents throughout the year (if they play their own division more than others it will come out a bit higher than the others) and do the same with a goalie from an average scoring division. Note how small the differences are. The reason: they all still play most of the same teams.

Quote:
While the Oilers dynasty were gone, their players were still around in the western divisions. Also during the 1990s on, there were a few expansion teams added, diluting the talent, maybe that had an effect on the scoring.
Coffey and Simpson went to Pittsburgh, a lot of guys went to the Rangers... I know Kurri and Gretzky went to the Kings... Tikkanen went everywhere. I'm sure if you track the GP of oilers dynasty players from 1990-1995, you'll find that they played more in the east than the west. And even if they didn't, it's statistically insignificant.

Quote:
I think it would mean more. If you played your division more, and your division has tougher opponents, then your numbers are skewed against the average. Brodeur plays against Gaborik, Malkin and Crosby right now, and against Ovechkin in his conference and other very talented forwards like Kovalchuk, Semin, Backstrom etc.

Either way, my point is that they don't play on the same teams with the same opponents. You cannot really adjust SV% against environment norms because each team plays differently against different opponents. Just simply averaging everything doesn't give you the complete picture. only the rough picture.
It means more nowadays than it did before (because inter-conference play was all but eliminated there for a few years) but it still means very little in the long run. Any "difficult opponent" in your division still accounts for 10% of your games.

Quote:
The rough picture says that Hasek saves .008 more than Brodeur, thats .8 percent, less than 1% of the shots faced. The few times you've seen them going head to head, the contests are usually very close, including that one which Hasek won with 70 saves in the 4th OT in 1994 with Brodeur saving 49. Watching that game you couldn't really say one was better than the other. Both were spectacular.

Same thing could be said about the 51 save vs 45 save between Brodeur and Lundqvist recently.

Considering all the intangibles between the goalies, 1% really isn't that much of a difference.
It's thinking that it's just 1% that is the major error there.

A sv% of .9141 means you allowed 8.59% of your shots. A sv% of .9223 means you allowed 7.77%. 8.59 is 10.55% higher than 7.77. Simply stated, it means that the goalie that was .008 behind in sv% was 10% more likely to allow any given shot than the other goalie.

A difference of .008 where both goalies were 20 sv% higher than that, or 20% sv% points lower, would yield different results too. That's why I said use error rates. I'm surprised you never found variance in the 80s when you did that study as I was writing the reply advising you of this. Your variance would be even lower if you did.

Quote:
Originally Posted by MoonDragn View Post
Look at the competition. Who would you say was Roy's competition during the 80s? There were nobody as good as him, maybe due to his butterfly style.

Once Hasek came into the scene, there was also Potvin, Belfours and Joseph, to name a few. Brodeur simply played in a time with more competition. How would he have compared if he played in the same time period as Roy without Roy there?

Plus another point is that Brodeur's strength comes from puck handling and consistency. Something that other of his peers didn't possess. They might be able to beat him one year or another, but not every year.

Another thing to consider, both Roy and Hasek were veterans in their craft when Marty came in as a rookie. He has only gotten better over time, but comparing a young Brodeur vs two experienced goaltenders doesn't seem fair at all.

We could compare each goaltender at the same age in their careers, but then you'd want to adjust for league averages, which isn't fair because again, the competition wasn't the same.

Simply looking at SV% isn't enough, now you start to look at how valuable a save is for low scoring teams, how valuable is a save for bad defense teams etc. Goal prevention and ultimately how much a goalie contributes to the team win is what we really should look at in judging which goalie is better. I just don't think that can be done accurately.
You're reading into competition too much here. All these goalies had the best goalies in the world as their competition, and all faced shots from the best shhoters in the world.

All goalies have an entire carer of data to go by. No one's comparing young Brodeur to veteran Roy and Hasek.... (edit) And Mike Veron and Ron Hextall.

Roy, as I said, was not really a superstar until the 1989 season. From 1988 to 1993, his competition included Andy Moog (who was much better than Potvin, who you named), Mike Richter, Ed Belfour, Curtis Joseph, and John Vaniesbrouck.

You said yourself: "Goal prevention and ultimately how much a goalie contributes to the team win is what we really should look at in judging which goalie is better." and that is exactly why adjustments for league averages need to be done. When scoring is lower, the value of a goal and what it means to a win is higher. The reverse when scoring is higher.


Last edited by seventieslord: 01-14-2010 at 05:39 PM.
seventieslord is offline   Reply With Quote
Old
01-14-2010, 05:13 PM
  #607
hockeyf4n
Registered User
 
Join Date: Apr 2007
Posts: 115
vCash: 500
Quote:
Originally Posted by MathMan View Post
Except wins and shutouts are both team stats. Yes, the goalie counts for a significant chunk of the team, but so does the #1 center and I don't see anyone bringing up Gretzky's or Lemieux's win percentages or shutouts when trying to decide which was better. It makes exactly the same amount of sense. Crediting goalies with a win-loss record was one of the most boneheaded decisions in hockey.

Any argument that rests on evaluating goaltenders based on wins or shutouts is so inherently flawed as to be useless. It's highly unfortunate that so much evaluation of goalies occurs on that basis considering how invalid it is.

If you want to use statistics to make any argument about a goalie's ability, use some sort of goaltending statistic.

Therein lies the problem. there is no goalie statistic devoid of a team's influence. If you value save % as a goalie statistic then you can't state that a shutout is also meaningless because a shutout is equivalent to a perfect save %.

hockeyf4n is offline   Reply With Quote
Old
01-14-2010, 06:12 PM
  #608
RabbinsDuck
Registered User
 
RabbinsDuck's Avatar
 
Join Date: Feb 2008
Location: Brighton, MI
Country: United States
Posts: 4,728
vCash: 500
Quote:
Originally Posted by hockeyf4n View Post
Therein lies the problem. there is no goalie statistic devoid of a team's influence. If you value save % as a goalie statistic then you can't state that a shutout is also meaningless because a shutout is equivalent to a perfect save %.
And there is no player statistic either that is devoid of team influence... except perhaps faceoff percentage.

You work with what you have, and obviously some are better than others.

RabbinsDuck is offline   Reply With Quote
Old
01-15-2010, 02:38 PM
  #609
MathMan
Registered User
 
Join Date: Jan 2006
Posts: 16,940
vCash: 500
Quote:
Originally Posted by hockeyf4n View Post
Therein lies the problem. there is no goalie statistic devoid of a team's influence. If you value save % as a goalie statistic then you can't state that a shutout is also meaningless because a shutout is equivalent to a perfect save %.
Sure, save percentage isn't perfect, but it's a damn sight better than wins, which are heavily affected by things the goaltender has practically no control over (especially, goals for!). Save percentage is affected by shot quality and the defense, of course, but at least it's a stat that belongs to the goalie and not the team.

Scoring stats are personal stats for skaters, yes they are affected by linemates, yes they are imperfect, but nobody claims that makes them useless when evaluating skaters. Like I said, if wins are a valid statistic for goalies, then it's a valid stat for top-line centers, and defensemen, too.

BTW, a shutout implies equivalent to a perfect save % over one game, which is a significant difference with a season-long save percentage which has a much bigger sample size and thus is much less vulnerable to momentary variations. More importantly, getting a shutout is also a factor of how many shots against are allowed during a game, which is not a component of save percentage. The two stats are not near equivalent.

MathMan is offline   Reply With Quote
Reply

Forum Jump


Bookmarks

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off



All times are GMT -5. The time now is 09:58 AM.

monitoring_string = "e4251c93e2ba248d29da988d93bf5144"
Contact Us - HFBoards - Archive - Privacy Statement - Terms of Use - Advertise - Top - AdChoices

vBulletin Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
HFBoards.com is a property of CraveOnline Media, LLC, an Evolve Media, LLC company. 2014 All Rights Reserved.