I wasn't going down that route. I think Crosby has earned his place with Hull and Beliveau at this point, at least in terms of quality and quantity of prime. His PPG dominance in his prime is closer to Howe than it is to those two, something that shows the statistical anomaly of comparing Top X point/PPG finishes from dramatically different league sizes.
Comparing Top 3, 5 or 10 scoring and Hart finishes from different eras
- Thread starter daver
- Start date
I wasn't going down that route. I think Crosby has earned his place with Hull and Beliveau at this point, at least in terms of quality and quantity of prime. His PPG dominance in his prime is closer to Howe than it is to those two, something that shows the statistical anomaly of comparing Top X point/PPG finishes from dramatically different league sizes.
nothing is ever 100% provable but there are many players from outside the 5 traditional provinces since the 1970's.
It would be very difficult to argue that they are less elite players now than in the 06 era right?
I thought about this walking my dogs and it needs more exploration.
Comparing top 10 finishes in a 6 team league to a 30 team league(18 top line players to to 90 top line players) has a bit of a ratio problem and I'm not sure what it really tells us if anything over time.
Other than the extremely subjective argument that the general talent level in the league has not stayed generally the same as the league has expanded since 1967, there is clear statistical evidence that 10th place in the O6 was clearly finishing farther back from the scoring leader than 10th in the current era.
This should come as no surprise as the 10th place scorer today is finishing in the 94th percentile (out of 180 Top Six forwards) vs 10th place in the O6 being in the 72th percentile (out of 36).
This is strictly a numbers game, not an indication that talent or competition in the O6 was lower. It means that players fighting for a Top Ten finish in today's league are battling against a bigger field of other Top Tenish level players.
This is why top 10 finishes between different eras need to be taken with a grain of salt.
Of these Art Ross finishes:
Beliveau finished an average of 17% behind the leader in points and 10% behind the leader in PPG
Hull finished an average of 12% behind the leader in points and 9% behind the leader in PPG
Crosby finished an average of 9% behind in points and 5% behind the leader in PPG
If one choose to add Crosby's two seasons where he missed out on enough games to miss out on the Top Ten in scoring (07/08 - 31st and 10/11 - 32nd), his average behind the PPG leader drops to 2%.
In Howe's best 14 season stretch (49/50 to 62/63), he finished an average of 1% behind in points and 2% behind the leader in PPG.
I don't think I would expect there to be that many more Top Ten scorers when comparing the two but this wouldn't change the large difference between the percentile that each player is finishing in or the larger gap between 1st and 10th in each respective era.
Here is an interesting breakdown of the teams represented for all the players who had at least two Top Ten finishes:
O6 era (# of players, # of times represented in the Top Ten)
Montreal - 7 players, 31 times
Detroit - 7 players, 28 times
NYR - 3 players, 9 times
Boston - 3 players, 7 times
Chicago - 2 players, 6 times
Leafs - 2 players, 5 times
Current era (# of players, # of times represented in the Top Ten)
Pens - 3 players, 13 times
TBL - 3 players, 11 times
EDM - 3 players, 7 times
Van. - 2 players, 6 times
Dallas - 2 players, 5 times
Philly - 2 players, 5 times
SJ - 2 players, 4 times
Anah. - 2 players, 4 times
nine other teams had one player represented
This really hammers home the point the dynamic that a top O6 team playing significantly more games against poor defensive teams arguably boosted the point totals of its players vs. players on teams that played significantly more games against the best defensive teams.
Add this to the statistical realities of comparing a six team league to a 30 team league.
vadim sharifijanov
Registered User
- Oct 10, 2007
- 29,803
- 18,024
i guess the question is whether the bottom three defensive teams of the 1950s is in fact "poorer" than the bottom sixteen defensive teams in the 2010s. i don't know, i mean, this year the vancouver canucks were tied for 17th in goals allowed (i.e., tied for 14th most goals allowed). their defense was
alex edler - chris tanev
troy stecher - ben hutton
with the third pair being some ungodly combination of erik gudbranson, alex biega, derick pouliot, michael del zotto's corpse, and a surprisingly not completely decomposed luke schenn
They were "poorer" on a relative basis meaning the team with the worst GA in the O6 was farther away from the league average GA than the worse teams are today.
Subjective opinion on players does not change this statistical reality.
Sure you can say that facing the 1950s Habs or Wings would make any defense poorer just as easy as you can say the 1950s Habs or Wings had great offences due to facing these really poor defenses.
My point is to show the volatility that occurs in a six team vs. a 30 team league. One would think that if the O6 was at such a significantly higher level competition-wise than the current league, we would not see such discrepancy in the GFs and GAs.
Objectively, one should be able to point to the star players O6 playing 15% more games against the poorer defensive teams in the league than today's stars as something to consider and frankly a lot more tangible than subjective opinion on competition levels or the very questionable claim of how many "HOF" calibre d-men individual players had to face.
That the O6 tended to tilt towards two or three teams being based on league dynamics doesn't add anything to the aura that this era was the pinnacle of hockey. It actually stains it if anything.
Basically, I see no reason to not translate O6 results into into a fair comparison with results from a 30 team league. This mainly means not comparing Top X finishes without statistical context and playoff results without statistical context. For example, a team that reaches the 3rd round in the current league is the equivalent of reaching the SCF in the O6.
How can you possibly compare the abilities of these d-men vs. the d-men from a team 60 years ago beyond comparing their contributions to their team's respective defensive performances?
Why isn't reasonable to state that these d-men were good enough to make their team average defensively while the 1954/55 Hawks who gave up 30% more goals than the league average, were a "poorer" team defensively. We have absolutely no way in to objectively measure how each defensive group would have done if they switched places.
The fact that the the league is made up of 1/2 non-Canadians should, at the very minimum, indicate an increase in the number of elite players, unless you believe that Canada is not producing as many elite players as it did in the O6 and the void is being filled by non-Canadians.
Another very reasonable belief is that Canada is not only producing as many elite players as it did in the O6 but is producing more as the population has generally increased and there are also many players from provinces that did not produce many NHLers in the O6.
If we want to go down the "well, we still literally cannot say for sure there are more elite players" road, which it sounds like some posters here want to, then I don't know what to say to that. At some point, ridiculously unreasonable opinions need to be ignored.
Last edited:
This would seem to be in line with their respective best 10 Art Ross finishes. How exactly does VsX works? Is it a player's point total compared the average point total for all 1st liners, or is it compared to the leading scorer?
Thanks
Cursed Lemon
Registered Bruiser
That's not what that meant.
You know that's not what that meant.
You're the one who was always constantly whining about cross-era comparisons.
You said there is no way to know whether there are more elite players or not currently. I would say that when you get beyond the Top 3 or so in any given year, statistically there is evidence to say there is.
Cursed Lemon
Registered Bruiser
The % gap from the #1 scorer over a large sample of seasons.
The #5 scorer from 1947 to 1967 had, on average, 18% of the leading scorers point total. The #5 scorer from 1997 to 2109 had, on average 13% of the leading scorers point total.
The #10 scorer from 1947 to 1967 had, on average, 50% of the leading scorers point total. The #10 scorer from 1997 to 2109 had, on average 30% of the leading scorers point total.
This is not surprising given there are, in theory, more 1st line caliber forwards in today's league. It makes complete mathematical sense that the #5 performer out of a larger group will be closer to the top %-wise than the #5 performer in the smaller group.
Dennis Bonvie
Registered User
10th place finishes in scoring from the last 8 seasons of the 06 era and the last 8 NHL seasons.
Proof of nothing, just thought it would be of interest.
1960 - Dean Prentice
1961 - Alex Delvecchio
1962 - Claude Provost
1963 - Bobby Hull and Murray Oliver (tie)
1964 - Dave Keon
1965 - Phil Esposito, John Bucyk and Ralph Backstrom
1966 - Murray Oliver
1967 - Alex Delvecchio
2012 - Erik Karlsson and Patrik Elias
2013 - Mike Ribeiro, Ryan Getzlaf and Pavel Datsyuk (tie, shortened season)
2014 - Nik Backstrom, Jaime Benn, Alex Ovechkin and Joe Pavelski (tie, 8-11)
2015 - Vladimir Tarasenko, Henrik Sedin, Claude Giroux and Nick Foligno (tie)
2016 - Evgeny Kuznetsov and Artemi Panarin (tie)
2017 - Vladimir Tarasenko
2018 - Sidney Crosby
2019 - Aleksander Barkov
Cursed Lemon
Registered Bruiser
Except you haven't dealt with the definition of "elite", which has always been the ability to separate one's self from the rest of the competition. That talent parity is greater in the league today does not mean that there are more "elite" players, and cross-era comparison still remains next to impossible.
That the modern players were significantly closer to the 1st place than the O6 players were should indicate something shouldn't it?
seventieslord
Student Of The Game
seventieslord
Student Of The Game
Again, this is a really misleading way to state it.
First, what you actually mean is:
The #5 scorer from 1947 to 1967 had, on average, 18% less than the leading scorers point total. The #5 scorer from 1997 to 2109 had, on average 13% less than the leading scorers point total.
The #10 scorer from 1947 to 1967 had, on average, 50% less than the leading scorers point total. The #10 scorer from 1997 to 2109 had, on average 30% less than the leading scorers point total.
Second, it shouldn't be stated that way anyway, when it says very little about the relative difference in the two figures. You wouldn't say that Columbus was twice as good as Montreal this year because Montreal was twice as far from 100 points as Columbus was, would you?
That just obscures the actual quantities that we're talking about. 98 is not much more than 96.
Similarly, you should be referring to the #5 scorers as 82% and 87% of the leading scorer, and the #10 scorers as 50% and 70% of the leading scorer.
87% is just 6% more than 82%. 70% is just 40% more than 50%. It's just misleading to make these differences appear to be 38% and 67% when they're not.
I agree on the last part as cross era comparissons are difficult due to bias and inability to account for different factors in any definitive formal manner.
The notion that there aren't more elite players simply does in any actual test with the emergence of elite players from beyond the 5 provinces providing 99+% of pre 1970s NHL.
I can't remember how many times it happened but wouldn't both Wayne and Mario be extreme outliers in any years they finished 1, 2 in league scoring.
And the Orr effect on Espositio makes the early 70s a bit of a blip in history as well.
seventieslord
Student Of The Game
Yes, there are times when there are more than one outlier, like the situations you mentioned.
I agree. It can be worded better.
There still is an obvious discrepancy between the #10 scorers, and an arguable one for #5.
