Columbia Research Paper Looks at Fighting/PIM and Mortality Rates

[That Aud Smell]

The eye-grabbing headline is that NHL enforcers die, on average, 10 years earlier than their fellow NHL players.

The thing that really stood out to me was that the "control group" for this study -- NHL players who weren't "enforcers" -- had an average age of death of ~ 57 years old. (Compared with enforcers whose average age at death was 47.)

There are limitations and caveats with this research, to be sure. Several of them are noted in the paper itself. (I also note that they looked at player data going back to 1967, so, for many of the relevant periods of time, life expectancy wasn't what it is now.)

All that aside: Holy hot damn those are young average ages of death.

[PerreaultForever]

Well ya, and so do boxers and football players who take head shots and so on. 

High risk high reward. Nobody forces them to do it. 

[Taro T]

8 minutes ago, That Aud Smell said:

The eye-grabbing headline is that NHL enforcers die, on average, 10 years earlier than their fellow NHL players.

The thing that really stood out to me was that the "control group" for this study -- NHL players who weren't "enforcers" -- had an average age of death of ~ 57 years old. (Compared with enforcers whose average age at death was 47.)

There are limitations and caveats with this research, to be sure. Several of them are noted in the paper itself. (I also note that they looked at player data going back to 1967, so, for many of the relevant periods of time, life expectancy wasn't what it is now.)

All that aside: Holy hot damn those are young average ages of death.

Only looked very briefly (obviously) at the study but a few things stood out on a cursory review.

The items that stood out IMHO were that they had 331 fighters identified out of all the players that had played in the NHL over the past 55 years.  They then tried to match up characteristics of non-fighters as well as they could to the fighters they identified, so they only looked at 662 players (and they went similarly with their other study group - players with at least 3 PIM/G) of the full set of NHL players.  So, those "life expectancies" aren't actually life expectancies.  Those are the average ages at death of the 6% of enforcers looked at that have died and the 7% of non-enforcers that have died.  And after they handpicked their "control" group, that group actually has had a greater percentage die than of the enforcers group.

They also acknowdedge that they ended up with almost identical numbers of dead players from both smaller pools (21 dead enforcers and 24 dead non-enforcers) though the enforcers had 3 suicides, 2 ODs, 2 neurodegenerative disease induced deaths, and 4 vehicular deaths; the "control group" only had 1 death by any of those causes in their 24 deaths.  So, 52 percent of the enforcers died from non-natural causes and only 4 percent of the non-enforcers died from non-natural causes.  That right there can explain a large chunk of their results.  They are trying to say that the head trauma from fights is the primary cause of this disparity, but they don't really control to any degree whatsoever for other items which could account for these results such as the possibility that enforcers may have taken performance enhancing drugs at higher rates than their peers nor how significantly the control group of ALL non-enforcers would've compared to the enforcers.  There very well may be a bias injected into this study by the criteria used to reduce the control group from nearly 6,000 down to 331.  Intentionally only looking at 5% of the available data seems questionable at a minimum.  And, if you take out what they consider to be CTE related deaths then the enforcers are much more likely to still be alive (97% of the remaining 320 enforcers were still alive when the study was conducted compared with only 93% of the remaining 330 non-enforcers).

Also, the average life expectancy of the players seems to be artificially low for both groups as less than 7% of the players studied have actually died by the time the study was conducted and there have been far more NHLers that have been playing in the past 30 or so years than there were in the 1st 30 years covered by the study as the league has gone from 12 teams to 32.

Will try to take a deeper dive into this study, but have more questions right now about the methodologies than the results.

[That Aud Smell]

6 minutes ago, Taro T said:

So, those "life expectancies" aren't actually life expectancies.  Those are the average ages at death of the 6% of enforcers looked at that have died and the 7% of non-enforcers that have died.

This is helpful context.

Even so ... it seems odd that they'd have that many ostensibly healthy men dying at such ages. The lead researcher is quoted as saying that he was startled by those average ages of death, for those who did die. Then again, I suspect the lead researcher is looking to generate as much attention for his study as he can.

[Taro T]

1 minute ago, That Aud Smell said:

This is helpful context.

Even so ... it seems odd that they'd have that many ostensibly healthy men dying at such ages. The lead researcher is quoted as saying that he was startled by those average ages of death, for those who did die. Then again, I suspect the lead researcher is looking to generate as much attention for his study as he can.

And at least one of those enforcers (Kordic) that killed himself was definitely one known to suffer from "'roid rage."  

And doubt that Brian Spencer was considered 1 of those enforcers but he likely wasn't in the control group either as his death was definitely of unnatural causes and due to being a drug dealer in later life.  

It seems like it's a very interesting study, but not sure just how much reliable info will be coming out of it based upon the way the study was set up.  (Again, this is only after a VERY cursory review of the study's methodology.  The methodology may be more rigorous than it appears at 1st blush (then again, it may be even less useful too).)

[shrader]

I'm not crazy about their matching process.  They reduced 5708 of the so called non-fighters down to the control group of 331.  They've thrown away a lot of data that they didn't necessarily have to.  Match 2 to 1, 3 to 1, ...  Hell, at the very least I'd like to see the characteristics of everyone since they do have them.

Can you tell that I'm not a fan of matching?

[Taro T]

1 minute ago, shrader said:

I'm not crazy about their matching process.  They reduced 5708 of the so called non-fighters down to the control group of 331.  They've thrown away a lot of data that they didn't necessarily have to.  Match 2 to 1, 3 to 1, ...  Hell, at the very least I'd like to see the characteristics of everyone since they do have them.

Can you tell that I'm not a fan of matching?

Yep.  And the fact that they did trim out so much data leads to the question of WHY?  Were they just lazy or were they needing to do that to get the results they wanted or did they actually have a truly legit reason for doing so?  REALLY doubt it's that 3rd option.

[thewookie1]

To be frank I can’t stand this Rick Westhead guy; he’s an overly sensitive media man who looks to crucify anything deemed not PC.

To be honest as interesting as this study was it doesn’t exactly give anything new. It doesn’t take a doctor to understand the more trauma you get the more likely you are to die younger than someone with less physical trauma. 
 

I understand the concern for player health but there are going to be downsides to making millions of dollars. People like Rick want fighting banned; then they’ll go after hitting in general since it too causes trauma and injuries. Next thing you know they look like women’s leagues and they let women into the NHL since they no longer have to fear being hip checked into the next county. It honestly scares me that we’ll have an unrecognizable sport where it’s purely based on skill and speed where players make 30mil a year but rarely get any drawback from it since all bodily harm is made illegal. *rant over*

[Zamboni]

Stu Grimson 57

Tony Twist is 55

Rob Ray is 54

Tie Domi is 53

Brad May is 51

Matt Barnaby is 50

 

Needless to say, Data is flawed a touch.

[Second Line Center]

Is it safe to say “enforcers” of the late 60s, 70s, and 80s didn’t take the kind of care of themselves as the non enforcers?  Why did they use 1967?   

Guaranteed a lot more alcohol   
 

Studying enforcers vs non enforcers is giving you a totally different human.

 

Edited May 10, 2023 by Second Line Center

[Sabel79]

Causation, correlation, and so on.  
 

I don’t think anyone thinks that getting punched/hit in the head a whole bunch is good for you.  Do we know more now than we did 30 years ago?  Absolutely.  Should that knowledge change our views on certain things?  Absolutely.  Has it?  Absolutely again.  
 

I genuinely believe that football (at least as we know it) will not exist in another 20 years.  It can’t.  The money/liability involved will become untenable once more and more science comes out.  This is, of course, assuming things in general do not take a turn we’ll have to go to the politics club to talk about.    
 

Hockey is already close to doing away with fighting, as it probably should.  It’ll have to evolve further to maintain the integrity/spirit of the game, but I think it stands a better chance.  Just nobody punching anybody.  

[shrader]

2 hours ago, Zamboni said:

Stu Grimson 57

Tony Twist is 55

Rob Ray is 54

Tie Domi is 53

Brad May is 51

Matt Barnaby is 50

 

Needless to say, Data is flawed a touch.

Did you see the words “on average”? Each of these guys should be included based on what I briefly read. 

[PerreaultForever]

6 hours ago, Zamboni said:

Stu Grimson 57

Tony Twist is 55

Rob Ray is 54

Tie Domi is 53

Brad May is 51

Matt Barnaby is 50

 

Needless to say, Data is flawed a touch.

Not to mention all the even older guys who didn't wear helmets and got smacked around and smacked others a lot harder than they do today.

Just considering some Sabres who punched and got punched, Jim Schoenfeld is 70, Jerry Korab is 74, Rick Dudley is 74, we could go on..................

[shrader]

I can point you to several smokers who didn’t develop lung cancer or serious heart issues. I guess that means smoking’s not bad for you. 
 

This thing is not saying that the fighter will die at exactly 47.5 years old. Listing off a handful of names that lived well past that point, since that figure is an average, that means we’re going to be able to list a bunch who died well before it. That’s how averages work, there will be plenty above and below that mark. 
 

And I questioned the matching earlier, but that’s only an issue for their control, non-fighter group. The players they counted as fighters, every single one of them is included in that 47.5 figure. On its own, that’s a pretty startling number. You may want to question how they settled on that fighter definition (50+ fights) and that’s fair, but there are absolutely zero games being played in calculating that 47.5. It’s basic math at that point. With a little free time and google, you can verify it yourself. 

[JohnC]

Fighters do a lot of clutching and grabbing when they are engaged in a matchup. But they also throw punches to the head. You don't have to do a detailed statistical analysis to determine that it increases head injuries. And you don't need to have a medical background to recognize that over time the repeated hitting to the head could have serious health repercussions. 

More often than not fights between players from opposing teams agree in advance to go at it.  In reality, they are mostly pre-staged fights for a variety of reasons. It makes no sense to me that on one hand there is a lot of pontificating about player safety, and then on the other hand an activity is allowed that does jeopardize player's safety. 

Fighting is not allowed in the college game. Anyone who has watched college games recognizes that it is a physical game. No fighting is allowed in Olympic hockey. Yet, it is compelling to watch.

In one of Spellman's first game as a Sabre he got in a fight to demonstrate his toughness to his new teammates. The opposing fighter hit him in the head with a gruesome head shot. It was a quick dispatch. Spellman got close to being knocked out and went down. He's lucky that his season and his career were not ended in one of his first games with his new team. For what? To demonstrate that how tough he was?

Football is very physical. Fighting is not allowed. Basketball is a very physical sport. Fighting is not allowed. Fighting in the NHL has dramatically declined compared to the era before. It's time to stop with this archaic remnant of this game. You don't need fighting in this tough sport. 

[Zamboni]

Need vs. Want

 

Let’s not get the two confused

[Taro T]

9 hours ago, shrader said:

I can point you to several smokers who didn’t develop lung cancer or serious heart issues. I guess that means smoking’s not bad for you. 
 

This thing is not saying that the fighter will die at exactly 47.5 years old. Listing off a handful of names that lived well past that point, since that figure is an average, that means we’re going to be able to list a bunch who died well before it. That’s how averages work, there will be plenty above and below that mark. 
 

And I questioned the matching earlier, but that’s only an issue for their control, non-fighter group. The players they counted as fighters, every single one of them is included in that 47.5 figure. On its own, that’s a pretty startling number. You may want to question how they settled on that fighter definition (50+ fights) and that’s fair, but there are absolutely zero games being played in calculating that 47.5. It’s basic math at that point. With a little free time and google, you can verify it yourself. 

But that is ONLY the enforcers that have died to date.  There are 310 that are still alive, MANY of which are well above that 47.5 year "life expectancy."  It is true that of those that died, the average age of death was 47.5 (presuming that is true, haven't looked at the individual ages at the time of their death personally) but it is NOT true that the life expectancy of an NHL enforcer is 47.5 years.

They also don't look at whether there is a causation within the 11 enforcer deaths that they consider correlated to head trauma.  For example, look at the drug usage, for the less than a handful that had drug related deaths, did those players that had a propensity to become enforcers have a propensity towards developing an addiction to drugs and did that lead them to be more likely to become enforcers?  The ones that died in auto accidents, did the personalities that allowed them to become enforcers also have a tendency to lead those players to be "risk takers?"  If it was within the personality to be risk takers, they'd be expected to have more and more severe automobile accidents than the average person.

Also, am having a hard time understanding (and or finding what the meaning is) that of the players from the enforcer group and from the non-enforcer control group that the control group had SIGNIFICANTLY more deaths that couldn't be associated with head trauma (by the criteria established by the researchers to be proxies for head trauma) than the enforcer group had.  23 deaths out of 330 is double that of 10 out of 320.  The study doesn't even seem to acknowledge that as they are ONLY trying to see if they can correlate additional deaths to the head trauma.  It would seem that in their attempt to "match" the fighters to non-fighters that they have done a poor job of that as the deathrates from non-head trauma related causes are significantly different between the 2 groups.  And they are significantly higher for the non-enforcer control group.

And the authors would seem to be implying that the players that died from head trauma would've died anyway but about 10 years later without the head trauma.  (That does appear to be the result they are presenting.)  But from a cursory review of the study there doesn't seem to be any reason to believe those same players would have died anyway (just as there's no reason to assume they wouldn't have died).

On the surface, and without referring to this study, it would seem quite plausible that people that routinely suffer head trauma would have lowered life expectancies than those that don't routinely suffer head trauma.  And this study supports that presumption.  But, and again only took a cursory look at the study, it doesn't seem that the study as constructed actually lends much if any useful data to answering that question.

[Claude Balls]

Well seeing how "enforcers" have basically disappeared from the NHL, this doesn't seem very relevant nowadays. 

[matter2003]

Why do NHL players die so young? That's almost 20 years younger than the general population's life expectancy for males.

Seems strange although back then athletes didn't really take care of themselves the way they do today. Would expect that age to steadily increase as player starting in the late 90s began to really focus in on this.

Edited May 11, 2023 by matter2003

[MattPie]

21 minutes ago, matter2003 said:

Why do NHL players die so young? That's almost 20 years younger than the general population's life expectancy for males.

Seems strange although back then athletes didn't really take care of themselves the way they do today. Would expect that age to steadily increase as player starting in the late 90s began to really focus in on this.

I would expect that too, although I think it's probably more players *born* in the 90s. I don't feel like there was nearly the emphasis healthy living in the 90s/2000s as in the 2010s/2020s. But that might just be my perspective as someone who in their 40s started living far more healthy (and it f'n works; I feel way better than I did five years ago). A professional athlete (especially of the era that we're talking about) after  retirement should probably make drastic changes to their lifestyle since they (likely) don't have the same built-in activity level (practice, workouts, games). If you're exercising hard for a few hours six days a week, you can get away with some questionable diet choices.

As for the dying so young part, the study (as I understand it) is only looking at players that *have* died, so those average ages would be artificially low since many of the players in the age window are still alive. I would expect that if we come back to the same methodology in 20 or 30 years after the most of players have died for any reason, the average would be higher. but I bet you'd still see that same disparity where aggressive physical players die younger than players that aren't. As others above have said, this may be to personality traits or it may be from getting knocked in the head a lot too. My guess is the personality traits aren't as consistent as you'd think, and it's more down to getting knocked in the head.

This is part of the reason I take far less joy in the game than I used to, among other external reasons.

Edited May 11, 2023 by MattPie

[That Aud Smell]

12 hours ago, shrader said:

The players they counted as fighters, every single one of them is included in that 47.5 figure. On its own, that’s a pretty startling number. 

I agree.

And that's notwithstanding the fact that there are hundreds (?) in the fighters cohort that are still alive. The fact that a number of them have already died and died at that very young average age is, as you say, startling.

1 hour ago, Claude Balls said:

Well seeing how "enforcers" have basically disappeared from the NHL, this doesn't seem very relevant nowadays. 

I believe the study's authors cite to their conclusions and recommend removal of fighting from the game. E.g., by making fighting a game misconduct with supplemental discipline, etc.

[PerreaultForever]

I just heard a thing about this study on the CBC and in the deaths they even include cases of high speed traffic fatalities as if these were caused by their fighting. That's a heck of a reach imo. They attribute every death to brain trauma issues not accounting at all for the fact that a lot of these guys were not bright lights to begin with and their "risky behavior" is just from being idiots. 

Bob Probert didn't take massive amounts of cocaine because Tie Domi hit him in the head a few times. He took it because he was party animal with too much money and too few IQ points to begin with. 

[Claude Balls]

44 minutes ago, That Aud Smell said:

I believe the study's authors cite to their conclusions and recommend removal of fighting from the game. E.g., by making fighting a game misconduct with supplemental discipline, etc.

Yeah, I don't think (or hope) that happens. This years playoffs have had plenty of fighting and hard hitting. I love it. Although that slash by Pieterangelo last night was not part of the game. That was absolutely brutal. I hope he gets more than one game for that.

[shrader]

3 hours ago, Taro T said:

But that is ONLY the enforcers that have died to date.  There are 310 that are still alive, MANY of which are well above that 47.5 year "life expectancy."  It is true that of those that died, the average age of death was 47.5 (presuming that is true, haven't looked at the individual ages at the time of their death personally) but it is NOT true that the life expectancy of an NHL enforcer is 47.5 years.

They also don't look at whether there is a causation within the 11 enforcer deaths that they consider correlated to head trauma.  For example, look at the drug usage, for the less than a handful that had drug related deaths, did those players that had a propensity to become enforcers have a propensity towards developing an addiction to drugs and did that lead them to be more likely to become enforcers?  The ones that died in auto accidents, did the personalities that allowed them to become enforcers also have a tendency to lead those players to be "risk takers?"  If it was within the personality to be risk takers, they'd be expected to have more and more severe automobile accidents than the average person.

Also, am having a hard time understanding (and or finding what the meaning is) that of the players from the enforcer group and from the non-enforcer control group that the control group had SIGNIFICANTLY more deaths that couldn't be associated with head trauma (by the criteria established by the researchers to be proxies for head trauma) than the enforcer group had.  23 deaths out of 330 is double that of 10 out of 320.  The study doesn't even seem to acknowledge that as they are ONLY trying to see if they can correlate additional deaths to the head trauma.  It would seem that in their attempt to "match" the fighters to non-fighters that they have done a poor job of that as the deathrates from non-head trauma related causes are significantly different between the 2 groups.  And they are significantly higher for the non-enforcer control group.

And the authors would seem to be implying that the players that died from head trauma would've died anyway but about 10 years later without the head trauma.  (That does appear to be the result they are presenting.)  But from a cursory review of the study there doesn't seem to be any reason to believe those same players would have died anyway (just as there's no reason to assume they wouldn't have died).

On the surface, and without referring to this study, it would seem quite plausible that people that routinely suffer head trauma would have lowered life expectancies than those that don't routinely suffer head trauma.  And this study supports that presumption.  But, and again only took a cursory look at the study, it doesn't seem that the study as constructed actually lends much if any useful data to answering that question.

I grazed it so quickly, I didn't notice that it was limited to only those who were already dead.  The total numbers always felt low to me, and now I know why.  All of this just makes me wish even more that they would have skipped out on the whole matching part.  They're very limited in any conclusions they can make by limiting to that population.  They limit even further thanks to the matching.  The whole thing is pilot work hoping to drum up further research, nothing more.

It is kind of funny though seeing this rightfully scrutinized.  I've seen this exact type of approach applauded in other places simply because it fits the narrative they want to see.  I'm not throwing you under that bus (did they mention if any of the players were hit by a bus?), but I'm guessing you know exactly what I'm talking about.

[kas23]

Correlation does not equal causation. Guys whose only job is to beat the snot out of opposing players likely have a much different mentality or view on life than finesse players and I highly doubt this is limited to the hockey rink.