What are your thoughts about Richard's 1944/45 season?

[Vilica]

I guess I hid my visual evidence too well in a previous post.

Tableau Public

That's all forwards who played at least 10 games in a season for the entire period 42-43 to 66-67, normalized to an 82 game season. You can see that by default, the 4 war years are unchecked (do not appear in the visual). If you include those seasons, you will see how the x-axis goes from 1.8 to 3.8 to 1.8 to 4.8, and the y-axis goes from 120 to 140. You can play around with teams and years to experiment yourself, and if you scroll over any dot, you can see player, team, season, actual games played/points, and normalized points.

 

[daver]

I think that a higher scoring environment allows for fewer flukes. Kind of like a basketball game vs a hockey game.

A lower scoring environment allows for "fluke" elite performances given the margins between players are that much thinner.

From 2010/11 to 2016/17, 80 points was good for a Top Ten Art Ross finish. One flukey 3 or 4 point game moved from barely Too 20 into the Top Ten or won you the Art Ross. That wouldn't happen in a higher scoring environment.

My point was the three players in NHL history with video game numbers played during a time where the league expanded more than four times over and scoring reached record levels. Do I think this means they weren't as good as the gaps between them and their peers would indicate? A bit but nothing that would move them from Big 4 status.

 

[daver]

The most ridiculous notion here is the using of journeymen like Cain and Liscombe who had absurd career years in 43-44 under the exact same conditions and then didn't build on them in 44-45 as evidence that Richard's 44-45 was legit, rather than the complete opposite.

What an absolutely ridiculous discussion.

Of course Richard's 44-45 season was hugely inflated because of WW 2. Every goalie he was facing was an AHLer and 2/3 of opposition players were AHLers, while most of Montreal's team and their entire superstar line were intact. It's the simplest of simple logic that an all-time dominant line will dominate more against far weaker competition.

As for statistical evidence, look what happens to Montreal's team GF/G during the 1940s. It practically *doubles* when the team is playing garbage opposition (players started leaving in 1941, the full exodus was 43-45, and then trickled back 45-47) before quickly returning to pre-WW2 levels almost as soon as the league returns to full strength.

40-41 2.52
41-42 2.79
42-43 3.62
43-44 4.68
44-45 4.56
45-46 3.44
46-47 3.15
47-48 2.45

The most ridiculous notion here is the using of journeymen like Cain and Liscombe who had absurd career years in 43-44 under the exact same conditions and then didn't build on them in 44-45 as evidence that Richard's 44-45 was legit, rather than the complete opposite.

This was a league in 1943-45 where a decent 2nd liner like Herb Cain could set insane scoring records that lasted for 25 years until post-1967 expansion. Literally everyone knows those Cain marks are totally bogus, but to be arguing that Richard in the exact same situations was totally legit and scored at exactly the same rate he would have against a full-strength NHL is just bonkers.

Who is disputing that scoring was up thus goal totals by the goal leaders were up? Noone that I can see.

The discussion is around how to value Richard's level of dominance that season due to some players being absent. I would say the idea that one or two players who did not play in 44/45 with perfect health and close to peak seasons could have gotten closer %-wise to Richard than Cain did is accepted by the majority. Anything more than that is far to speculative IMO.

Disputing Richard's dominance due to an imbalance in the league is an interesting take. Does this apply to his other great seasons or to great seasons by other players in the O6 era?

As for Cain and Liscombe, in 43/44 these two hung with the scoring champ from 1942/43 - Doug Bentley, whose goalscoring ability is above reproach correct? Why didn't Doug Bentley put up a significantly more dominant season in 43/44 under the "2/3rds AHL" conditions? What Lorne Carr or Cowley? They were tied for 5th place in 42/43 and could do no better than that over the '43 to '45 time period.

What about other Top Ten goalscorers from '43 to '45 like Blake, Kennedy and Mosieko still getting into the Top Ten after everyone was back?

All of this points to only a couple of players being closer to Richard from strictly a statistical perspective. If you want to add other elements to the discussion, that's cool.

 

[daver]

This was a league in 1943-45 where a decent 2nd liner like Herb Cain could set insane scoring records that lasted for 25 years until post-1967 expansion.

This is a prime example of why I keep mentioning that some people are so focused on the raw totals from that season, they are incapable of an objective discussion surrounding Richard's dominance.

It doesn't matter that Richard may not have scored 50 goals if there is no war, or that he didn't reach 50 again, this should not bear any relevance to placing context on his level of dominance that year.

 

[blogofmike]

Who is disputing that scoring was up thus goal totals by the goal leaders were up? Noone that I can see.

The discussion is around how to value Richard's level of dominance that season due to some players being absent. I would say the idea that one or two players who did not play in 44/45 with perfect health and close to peak seasons could have gotten closer %-wise to Richard than Cain did is accepted by the majority. Anything more than that is far to speculative IMO.

Disputing Richard's dominance due to an imbalance in the league is an interesting take. Does this apply to his other great seasons or to great seasons by other players in the O6 era?

As for Cain and Liscombe, in 43/44 these two hung with the scoring champ from 1942/43 - Doug Bentley, whose goalscoring ability is above reproach correct? Why didn't Doug Bentley put up a significantly more dominant season in 43/44 under the "2/3rds AHL" conditions? What Lorne Carr or Cowley? They were tied for 5th place in 42/43 and could do no better than that over the '43 to '45 time period.

What about other Top Ten goalscorers from '43 to '45 like Blake, Kennedy and Mosieko still getting into the Top Ten after everyone was back?

All of this points to only a couple of players being closer to Richard from strictly a statistical perspective. If you want to add other elements to the discussion, that's cool.

I'm one who says balance does apply to other years, but can't think of anywhere it impacts as much as the WW2 years.

Lorne Carr stepping up was war-related, wasn't it? And Bill Cowley did great in 1944. He had 71 points in 36 games. If not for injury, the points record would have been Cowley's not Cain's. Someone earlier (@overpass ?) noted that Cain's scoring pace fell off as soon as Cowley was out. Purely as a goal scorer, Cowley being in the Top 10 is kind of a red flag in and of itself. He was a playmaker and wasn't really a candidate for the Retro Rocket. It would be like WW3 breaking out in the 90s and Adam Oates making the Top 5 in goals.

Doug Bentley scored 37 goals in 40 games in 1944. And had 1 goal in 10 games against a juggernaut Habs team that went over .830. Because the Habs had a relatively stacked roster that created an imbalanced league.

As for the Top Ten guys, Blake was on a line with Lach and Richard (HHOF quality linemates often help), a teenaged Ted Kennedy was 3rd in goals and never reached 3rd again, and Bill Mosienko was 5th, but actually declined in goals from the previous year when Bentley was there to help (HHOF quality linemates help in the Central Time zone too.)

 

[tarheelhockey]

I am interested in your idea that one event is more likely to lead to unlikely results than two events.

It’s not really my idea, it’s an objective mathematical principle.

I’m sure you’re familiar with the coin-flipping idea that you’re FAR more likely to flip 3 heads in a row than 10 heads in a row, and that 100 heads in a row is for all intents and purposes impossible. Even though each flip is independent, the likelihood of the total series approaching its weighted odds goes up dramatically with each flip.

Now, imagine each NHL player were given a coin with “goal” on one side and “no goal” on the other. Each player gets to flip his coin every time he (officially) takes a shot on goal. Clearly, this coin isn’t weighted evenly... an average coin is only 10% likely to land on heads, but certain elite players might have a heavily-weighted 15-20%.

In an average season, Connor McDavid would flip that coin about 250 times. His coin is weighted 15.07% likely to land heads, so on average he will land heads about 37 times per season. Ovechkin’s coin is weighted only 12.7%, but he flips it nearly 400 times a season. You get the drift.

Now, let’s give the players a second coin, labeled “point” and “no point”. They get to flip this coin every time they or one of the players on the ice with them takes a shot on net. This time, McDavid’s coin is weighted more heavily AND he flips it more often (1100 times a season vs 1000).

Don’t get me wrong... we can’t prove anything about the quality of the individual players just by flipping their coins. Hockey is far too complex for that. Perhaps certain coins are a heavier because of the contributions of teammates, or lighter due to lingering injury, or whatever. Perhaps the coins are measuring the wrong things entirely. Whether the coins accurately represent “true” ability is a rabbithole that only enthusiasts like ourselves would willingly explore.

Still, we can say with complete confidence that the point coin will reliably give us a “truer” result than the goal coin. Why? Because one coin is being flipped 4 times as many as the other. Irrespective of all other factors, it’s a statistical certainty that the point coins will reliably produce a more accurate representation of the underlying odds than the goal coins.

 

[tarheelhockey]

You seem to be saying, "let's completely ignore what Richard actually did in 44/45 and start anew with everyone else in the field". If that was the case then your 50/50 claim makes more sense given Richard lost goalscoring titles after 44/45 to players that were missing.

You sound like you truly believe that another player could have reached Richard's level of performance in 44/45 based only on the fact said player did not have the chance to compete that year.

I think that is completely unreasonable given that there were players capable of competing for the goal title in your hypothetical "full roster" season that actually played in 43/44 and 44/45 and were over 50% lower than Richard.

Richard was the only goalscoring leader from the '40s capable of anything close to that level of domination.

It seems to be consensus that Richard the player most capable of dominating the league to the level of 50%.


The key questions are:

1) Would Richard dominate ordinary NHL’ers at the same rate that he dominated minor leaguers?
- The common-sense answer is no. It’s obviously easier to score points against weaker competition. This should not need to be discussed in depth... it’s self evident.

2) Would players who missed time in ‘45 have exceeded the bar set by Cain/Kennedy?
- This is less common-sense than #1, but still seems to be a demonstrable “yes”. Virtually all of the non-MTL stars were out of the league at this time — including every active goal scoring champion other than Bill Thoms (1936). One of the few talents remaining, Schriner, was on pace for 40+ when he broke his leg. It seems overwhelmingly likely that at least one of the missing stars would have surpassed the production of support scorer Cain and teenaged Kennedy.

Taking 1 and 2 together, it appears overwhelmingly likely that Richard’s 50% lead would have been considerably smaller if the league had been at full strength. Arguing otherwise is either arguing against common sense (1) or arguing against all past and future evidence of player productivity (2).

The only way I can see to get Richard to a 50% win in a war-less 1945, is to argue that he actually didn’t score as much as he could have against the weakened league... that 50% is not an artificially high but rather an artificially low margin of separation between himself and the pack. I do think there’s some hint of evidence for that theory, but confirming it would mean we’d have to stop spinning our wheels in statistical arguments and get down to the harder work of finding out how the Habs deployed Richard that season, game-by-game.

 

[daver]

I'm one who says balance does apply to other years, but can't think of anywhere it impacts as much as the WW2 years.

Lorne Carr stepping up was war-related, wasn't it? And Bill Cowley did great in 1944. He had 71 points in 36 games. If not for injury, the points record would have been Cowley's not Cain's. Someone earlier (@overpass ?) noted that Cain's scoring pace fell off as soon as Cowley was out. Purely as a goal scorer, Cowley being in the Top 10 is kind of a red flag in and of itself. He was a playmaker and wasn't really a candidate for the Retro Rocket. It would be like WW3 breaking out in the 90s and Adam Oates making the Top 5 in goals.

Doug Bentley scored 37 goals in 40 games in 1944. And had 1 goal in 10 games against a juggernaut Habs team that went over .830. Because the Habs had a relatively stacked roster that created an imbalanced league.

As for the Top Ten guys, Blake was on a line with Lach and Richard (HHOF quality linemates often help), a teenaged Ted Kennedy was 3rd in goals and never reached 3rd again, and Bill Mosienko was 5th, but actually declined in goals from the previous year when Bentley was there to help (HHOF quality linemates help in the Central Time zone too.)

Was 42/43 a war year? That's when Lorne Carr stepped up and matched players like 2-time goalscoring leader Hextall, and Cowley who you are promoting as elite.

Cowley's pace in 1944 would have reasonably put him the 35 to 40 goal range and the 5th to 1st range (many others had similar GPGs after 36 games like Cain rather ironically) which puts him close to Cain and Liscombe. Nothing indicates that as being anywhere close to Richard's season in 44/45 when scoring actually was down from 43/44.

As for Bentley's "37 goals in 40 games besides the Habs", did Cain and Liscombe not perform under the exact same conditions and got within 2 goals of him? That seems to be a meaningless narrative.

As for Toe Blake, have you applied the same "stacked line" premise in evaluating all the missing elite goalscorers that could have potentially done better than Cain in 44/45? Perhaps they aren't as elite once the same evaluation is applied.

Ted Kennedy was 5th in goals in 46/47, two goals away from 3rd ; hardly overwhelming evidence that his 3rd place in 44/45 was unusual. He was Top 5ish in goals for the period of 46/47 to 50/51; further proof of his eliteness. His age is meaningless. Many 19 year olds have had prime/close to prime seasons in NHL history. The dynamic of teenagers not getting icetime in the O6 seems to apply here. Kennedy did get icetime and performed.

Mosienko was 3 goals away from being 5th again in 46/47; again hardly overwhelming evidence that his 5th place in 44/45 was unusual. Re: your remark about his numbers going down in 44/45 having some significance, you are too focused on raw numbers. Scoring was noticeably up in 43/44 when NINE players hit the 30 goal mark vs. only TWO hitting that mark in 44/45.

To reiterate my position, I have no issue placing one or two players closer to Richard in 44/45 in a hypothetical "full roster" 44/45 season which would have a marginal affect on his relative dominance that year. Whether he scores less goals than 50 goals in a hypothetical scenario is irrelevant.

If you want to ADD in the balance issue to that, please go ahead.

 

[Michael Farkas]

Was 42/43 a war year?

Yup. Still is.

 

[daver]

Yup. Still is.

Ok, so we should also question the scoring leaders in 42/43 in a similar manner as 43/44 and 44/45 are being questioned?

There doesn't seem to be anything unusual about 42/43 in terms of turnover of names in the Top Ten from the previous season (it was five, the average for the decade) or elite goalscorers like Hextall or Drillon having unusually dominant seasons like Richard's 44/45.

 

[Michael Farkas]

'43 probably not quite on the same level as '44 and '45, but close...

 

[MXD]

Ok, so we should also question the scoring leaders in 42/43 in a similar manner as 43/44 and 44/45 are being questioned?

There doesn't seem to be anything unusual about 42/43 in terms of turnover of names in the Top Ten from the previous season (it was five, the average for the decade) or elite goalscorers like Hextall or Drillon having unusually dominant seasons like Richard's 44/45.

Hextall and Drillon increased their goals totals in 42-43, but finished further down the order. In fact, 42-43 would be, numerically at least, their best season.

Hextall and Drillon would also never have a season as good as 42-43 afterwards either (well, only Hextall, as Drillon left the NHL).

If anything, what happens to these two players (increased totals + lower finish + possibly the moment where they hit the proverbial age wall) is good reason enough to asterisk that season as well. And we know that, because the players who passed them were not generationnal goalscorers or anything of like.

 

[tarheelhockey]

Was 42/43 a war year?

Per the list on page 1 of this thread, 65 players with immediate past/future NHL experience were in active service by 1943. This is equivalent to about half of the players in the NHL at the time. I believe the absences were unevenly distributed due to mid-season enlistments, meaning minor leaguers were being called up to replace legit stars all season. Among the players missing: the entire Kraut Line, Ken Reardon, Neil Colville, Wally Stanowski, Eddie Wiseman, Leo Reise, Chuck Rayner.

As for Bentley's "37 goals in 40 games besides the Habs", did Cain and Liscombe not perform under the exact same conditions and got within 2 goals of him? That seems to be a meaningless narrative.

Here are the goal scoring spreads.

Bentley
Boston - 7
Detroit - 12
Montreal - 1
Rangers - 13
Toronto - 5

Cain
Chicago - 7
Detroit - 8
Montreal - 4 (missed 1 game)
Rangers - 13
Toronto - 4

Liscombe
Boston - 11
Chicago - 6
Montreal - 3
Rangers - 14
Toronto - 2

One thing is very clear: if you wanted to contend for a goal scoring title, it helped a lot not to have to play against Montreal.

Ted Kennedy was 5th in goals in 46/47, two goals away from 3rd ; hardly overwhelming evidence that his 3rd place in 44/45 was unusual. He was Top 5ish in goals for the period of 46/47 to 50/51; further proof of his eliteness. His age is meaningless. Many 19 year olds have had prime/close to prime seasons in NHL history. The dynamic of teenagers not getting icetime in the O6 seems to apply here. Kennedy did get icetime and performed.

FWIW, as of 1942 no teenager had ever had a top-10 goal finish in the history of the NHL. Now, you do make a good point about opportunity — a teenager breaking into the league is highly unlikely to be given the chance to score often, so there’s a chicken-egg dynamic in play. But still. Kennedy’s season was unprecedented and clearly the result of being given a unique opportunity under very unique circumstances. The alternate interpretation, that Kennedy was the most NHL-ready player in history to that point, strains credulity.

 

[JackSlater]

I counted 43 players who had played at least 50% of the 1942 NHL season and missed the 1943 season due to WW2. I also counted 5 such players who left before the 1942 season due to WW2. This is somewhat mitigated by Brooklyn shutting down and some of its depth players dispersing throughout the rest of the league, but you're still looking at over 1/3 of the NHL being gone in 1943 due to WW2. 8 HHOFers, 5 additional players who had made or would make the post season all star team, and three of the league's starting goaltenders.

30 more such players would leave before 1944 and 7 more would leave by 1945, though a few (fewer than 10 I believe) would come back by 1944 or 1945. 1943 was a clear outlier compared to seasons outside of the WW2 years and to me is a clear war impacted season. 1944 an 1945 clearly saw a greater impact. This also didn't count players who had the beginning of their NHL career delayed like Meeker or Leswick.

 

[daver]

Hextall and Drillon increased their goals totals in 42-43, but finished further down the order. In fact, 42-43 would be, numerically at least, their best season.

Hextall and Drillon would also never have a season as good as 42-43 afterwards either (well, only Hextall, as Drillon left the NHL).

If anything, what happens to these two players (increased totals + lower finish + possibly the moment where they hit the proverbial age wall) is good reason enough to asterisk that season as well. And we know that, because the players who passed them were not generationnal goalscorers or anything of like.

The league increased in scoring. Why raw points are continually brought up with no context is baffling to me.

The point is their respective placing among the elite goalscorers did not change much in such a "war weakened season". That it actually went down goes against the common sense that we are supposed to believe.

If you want to argue that Hextall was regressing, that is reasonable but then there were also plenty of players at his age still in their primes throughout the '40s.

What about Lynn Patrick? Why didn't he improve on his strong goalscoring title in 41/42 in finish in the "war weakened" 42/43.

What about Red Hamill? Same thing.

What about Max Bentley? Same thing.

Where are the numerous examples of other elite goalscorers reaching anything close to Richard's level of domination in the 42/43 to 44/45 war era?

You have to admit that there isn't an overwhelming amount of evidence showing that players like Conacher (or who else is there?) would have dominated the war years to a degree that would have approached Richard.

 

[daver]

Here are the goal scoring spreads.

Bentley
Boston - 7
Detroit - 12
Montreal - 1
Rangers - 13
Toronto - 5

Cain
Chicago - 7
Detroit - 8
Montreal - 4 (missed 1 game)
Rangers - 13
Toronto - 4

Liscombe
Boston - 11
Chicago - 6
Montreal - 3
Rangers - 14
Toronto - 2

One thing is very clear: if you wanted to contend for a goal scoring title, it helped a lot not to have to play against Montreal.

Sure, that Bentley was better than his 38 goal total would indicate certainly isn't thus the point that Cain and Liscombe performed at his level (Top 5 - 10 in any season in the '40s) remains.

 

[MXD]

The league increased in scoring. Why raw points are continually brought up with no context is baffling to me.

Reformulate your post.

 

[daver]

Reformulate your post.

I have been extremely consistent in referring to goalscoring placements and dominance vs. peers as the only relevant metric in this discussion.

 

[MXD]

I have been extremely consistent in referring to goalscoring placements and dominance vs. peers as the only relevant metric in this discussion.

You also have been consistently shown up, so a certain measure of respect to someone who just wanted to post something regarding two players you SPECIFICALLY named, without painting the whole picture because I had no time to do so, would've been appropriate.

 

[daver]

FWIW, as of 1942 no teenager had ever had a top-10 goal finish in the history of the NHL. Now, you do make a good point about opportunity — a teenager breaking into the league is highly unlikely to be given the chance to score often, so there’s a chicken-egg dynamic in play. But still. Kennedy’s season was unprecedented and clearly the result of being given a unique opportunity under very unique circumstances. The alternate interpretation, that Kennedy was the most NHL-ready player in history to that point, strains credulity.

I agree that Kennedy's 19 year old season doesn't happen in our "what if" scenario. I just don't see it as one of the cornerstones to the argument that numerous less than elite goalscorers, i.e. had no business being in the Top Ten, excelled artificially in 44/45?

 

[tarheelhockey]

Sure, that Bentley was better than his 38 goal total would indicate certainly isn't thus the point that Cain and Liscombe performed at his level (Top 5) remains.

They did score goals at a similar rate... Hart voting (Bentley was a finalist in 43 and 44, Cain and Liscombe combined for 0 votes) suggests that Bentley was doing considerably more of it on his own. That interpretation is supported by Bentley scoring a much larger portion of his team’s goals (21%) than Cain (16%) or Liscombe (17%). His somehow also having more assists than Liscombe is almost inexplicable unless one accepts that Bentley was simply doing a lot more to drive his team offensively.

 

[daver]

You also have been consistently shown up.

I counted 43 players who had played at least 50% of the 1942 NHL season and missed the 1943 season due to WW2. I also counted 5 such players who left before the 1942 season due to WW2. This is somewhat mitigated by Brooklyn shutting down and some of its depth players dispersing throughout the rest of the league, but you're still looking at over 1/3 of the NHL being gone in 1943 due to WW2. 8 HHOFers, 5 additional players who had made or would make the post season all star team, and three of the league's starting goaltenders.

30 more such players would leave before 1944 and 7 more would leave by 1945, though a few (fewer than 10 I believe) would come back by 1944 or 1945. 1943 was a clear outlier compared to seasons outside of the WW2 years and to me is a clear war impacted season. 1944 an 1945 clearly saw a greater impact. This also didn't count players who had the beginning of their NHL career delayed like Meeker or Leswick.

Ok, so the point that no pre-war elite scorers were able to dominate a "war weakened" season is even stronger isn't it?

Hextall
Drillon
Patrick
Hamill
Max Bentley

None were able to exploit the 42/43 season, in terms of relative production, like Richard did in 44/45.

 

[MXD]

His somehow also having more assists than Liscombe is almost inexplicable unless one accepts that Bentley was simply doing a lot more to drive his team offensively.

If you're talking about Herb Cain vs. Liscombe, well...

Wings : 216G, 274A
Bruins : 223G, 323A

 

[MXD]

Ok, so the point that no pre-war elite scorers were able to dominate a "war weakened" season is even stronger isn't it?

Hextall
Drillon
Patrick
Hamill
Max Bentley

None were able to exploit the 42/43 season, in terms of relative production, like Richard did in 44/45.

I see two reasons for that :

- The NHL wasn't as depleted in 42-43.
- Maurice Richard was a signficantly better player than any of Hextall, Drillon, Patrick, Hamill and was a significantly better goalscorer than Max Bentley.

...These feel obvious? Good. Because they are.

 

[daver]

They did score goals at a similar rate... Hart voting (Bentley was a finalist in 43 and 44, Cain and Liscombe combined for 0 votes) suggests that Bentley was doing considerably more of it on his own. That interpretation is supported by Bentley scoring a much larger portion of his team’s goals (21%) than Cain (16%) or Liscombe (17%). His somehow also having more assists than Liscombe is almost inexplicable unless one accepts that Bentley was simply doing a lot more to drive his team offensively.

Ok, he was more valuable to his team. What does this have to with goal totals?

And whatever it does mean, have you applied the same evaluation to the seasons top goalscorers who were missing to see if some of their numbers need context too?