Sidney Crosby can break the record for the most consecutive seasons OVER a PPG (2023 update: 18 consecutive PPG seasons)

[MadLuke]

Where are you getting 12 for Denneny? HR has him with this:

I didn't took the time to look player by player but simply took all the top 10 finish in excel and used the consolidate function, I manually correct for B.hull (Brett) and B. Hull (Bobby) but not for Corb Denneny and Cy Denneny that got combined, would imagine the same for Cook. It was a fast raw copy paste from that excel result. It is common enough that HR should put the complete first name imo.

 

[MadLuke]

Not sure what you mean.

It could a mathematic and situational phenomenon more than it require to be a better hockey player to achieve to be in the top 10 one.

An other phenomenon could be once you are establish it is easier, but even a big body and legendary junior player a la Beliveau do not get Lafreniere minutes on the PP and first length opportunity at 19 in the original 6 on a good team, he need to get to 22-23 to play, it could occur in a wider league (Hasek, Jagr did have to wait for first rate ice time) but it is much rarer and it mean second unit PP and second line versus not playing or much below the roster.

 

[daver]

It could a mathematic and situational phenomenon more than it require to be a better hockey player to achieve to be in the top 10 one.

An other phenomenon could be once you are establish it is easier, but even a big body and legendary junior player a la Beliveau do not get Lafreniere minutes on the PP and first length opportunity at 19 in the original 6 on a good team, he need to get to 22-23 to play, it could occur in a wider league (Hasek, Jagr did have to wait for first rate ice time) but it is much rarer and it mean second unit PP and second line versus not playing or much below the roster.

The Top Ten scorers in both eras are players getting prime icetime. It should be viewed as a direct comparison of their peers.

I think we can presume that Beliveau and Hull were getting prime icetime in their 8 best seasons.

 

[daver]

Most top 10 points finish (if he does it this year):

G. Howe*	21
W. Gretzky*	16
S. Crosby	12
J. Beliveau*	12
C. Denneny*	12
B. Hull* J. Jagr	11
A. Delvecchio*	11
M. Richard*	11
B. Cook*	11
J. Sakic*	10
M. Lemieux*	10
P. Esposito*	10

[TBODY] [/TBODY]

Most top 10 ppg (using hockey reference cut-off for games played), I do not think he is in the top 10 this season:

G. Howe*	20
W. Gretzky*	16
S. Crosby	13
J. Beliveau*	13
B. Hull*	12
M. Lemieux*	12
M. Richard*	12
C. Denneny*	12
E. Malkin	10
S. Mikita*	10
J. Sakic*	9
J. Jagr	9

[TBODY] [/TBODY]

He would enter that Beliveau/Hull/Jagr/Richard group of forward with offensive prime length only bested by howe/Gretzky does seem to be true, will see if he is able to put 1-2 Top 10 level season to cleanly separated himself for them (that can be fuzzy with Hull WHL has he was still a top 10 scorer when he did leave the NHL at 33).

For someone with injury during is Top 10 certain peak, he did end up with an impressive amount of actual top 10.

In Crosby's 12 Top 10 point finishes, he was, on average, 12% behind the leader.

In Crosby's 13 Top 10 PPG finishes, he was, on average, 8% behind the leader.

In Beliveau's 12 Top 10 point finishes, he was, on average, 22% behind the leader.

In Beliveau's 13 Top 10 PPG finishes, he was, on average, 14% behind the leader.


So if the scoring leader was 100, Crosby would be at 91 points, Beliveau would be at 85 points.

 

[MadLuke]

The Top Ten scorers in both eras are players getting prime icetime. It should be viewed as a direct comparison of their peers.

I think we can presume that Beliveau and Hull were getting prime icetime in their 8 best seasons.

yes fully agree for that 8 best season I was talking about the number of Top 10 finish, maybe Beliveau have 1 or 2 between 19 and 23 in today league (that would be one the reason why it would have harder to make the top 10 today if that is the case, today virtually no player playing at a possible TOP 15 finish rate do not get prime icetime with the rare Malkin sometime having lesser winger some year for a player of that level but still get first PP wave and lot of icetime overall with an offensive mission and deployment, while in the O6 some talent where buried when they were young on the best team like Montreal-Detroit.

In a large league with 180+ forward getting first wave of PP time, you compete really with all the talents in the league getting prime opportunity and all the would be top 20-25 players usually but having a special season (lucky-teammate just a short peak) has well that you would not in the past, maybe it change little for the Art Ross has people competing for those would have raised to the top, but for top 10 possibly.

 

[daver]

Top 10 scoring finishes, and AST finishes can't be compared "raw" across eras. I'm surprised this still needs to be justified on this forum, should be pretty self-explanatory.

I think some people are defensive about players from the past being dismissed on mass when this should be viewed from strictly a statistic perspective with no subjective interpretation needed. If, on average, Player X finished closer to the top of the scoring race percentage-wise than Player Y did, that is an advantage for Player X. Finishing in a higher percentile of your peer group should be recognized. If these are not viewed as an advantage then we should also disregard other metrics such as scoring titles and scoring finishes.

 

[daver]

More data:

From 1949/50 to 1968/89, there were 64 different players who finished in the Top 10.

38 (59%) had more multiple Top Ten finishes
28 (44%) had three or more
11 (17%) had five or more
3 ( 5%) had 10 or more

From 1999/00 to 2019/20, there were 85 different players (33% more) who finished in the Top 10.

49 (57%) had more multiple Top Ten finishes
30 (35%) had three or more
9 (11%) had five or more
1 ( 1%) had 10 or more

More clear evidence that finishing in the Top 5/10 in the current era is significantly more impressive than in the O6.

This should be viewed as Crosby separating himself from Hull and Beliveau in terms of longevity of prime.

 

[vadim sharifijanov]

daver, you throw out numbers and then extrapolate conclusions from them that you don't explain.

no offence, but even if they are right, your conclusions is certainly not self-evident. you have to show your reasoning.

 

[daver]

daver, you throw out numbers and then extrapolate conclusions from them that you don't explain.

no offence, but even if they are right, your conclusions is certainly not self-evident. you have to show your reasoning.

It is a statistical position. If you want to put those numbers into context, please do so. Just like any other stat can be put into context.

Dominance vs. peers is the cornerstone of player comparison. I just showed that Crosby's Top 10 finishes are more dominant vs. his peers than Hull's and Beliveau's were vs. their peers. It is something that, like Beliveau's playoff resume or Hull's goalscoring, give him an edge over the other two.

It, IMO, places him right there with those two for #5 all-time.

 

[vadim sharifijanov]

It is a statistical position. If you want to put those numbers into context, please do so. Just like any other stat can be put into context.

Dominance vs. peers is the cornerstone of player comparison. I just showed that Crosby's Top 10 finishes are more dominant vs. his peers than Hull's and Beliveau's were vs. their peers. It is something that, like Beliveau's playoff resume or Hull's goalscoring, give him an edge over the other two.

It, IMO, places him right there with those two for #5 all-time.

what i mean is, when you say something like

More clear evidence that finishing in the Top 5/10 in the current era is significantly more impressive than in the O6.

This should be viewed as Crosby separating himself from Hull and Beliveau in terms of longevity of prime.

it is not clear at all without an argument that explains how you came to that conclusion

 

[MadLuke]

More data:

From 1949/50 to 1968/89, there were 64 different players who finished in the Top 10.

38 (59%) had more multiple Top Ten finishes
28 (44%) had three or more
11 (17%) had five or more
3 ( 5%) had 10 or more

From 1999/00 to 2019/20, there were 85 different players (33% more) who finished in the Top 10.

49 (57%) had more multiple Top Ten finishes
30 (35%) had three or more
9 (11%) had five or more
1 ( 1%) had 10 or more

More clear evidence that finishing in the Top 5/10 in the current era is significantly more impressive than in the O6.

If it is not too complicated and possible with your database/code:

a) If you look at Top 3 and 5 instead of top 10 of the respective era is sensibly the same (33% or more pool of players doing it from 1999 to 2020)
b) Is there a way to roughly estimate what a top 10 of the past would be today (say top 13) or vice versa, a top 10 today being what a top 8 in the 06 ?
Like maybe by looking at how many achieved three or more top 9 finish and look when it look like at the 35% of the modern era.

 

[daver]

it is not clear at all without an argument that explains how you came to that conclusion

A better resume of scoring and PPG finishes once statistical context is applied to raw placements.

 

[Black Gold Extractor]

In Crosby's 12 Top 10 point finishes, he was, on average, 12% behind the leader.

In Crosby's 13 Top 10 PPG finishes, he was, on average, 8% behind the leader.

In Beliveau's 12 Top 10 point finishes, he was, on average, 22% behind the leader.

In Beliveau's 13 Top 10 PPG finishes, he was, on average, 14% behind the leader.


So if the scoring leader was 100, Crosby would be at 91 points, Beliveau would be at 85 points.

Consider the points-per-game leader of 2013-14 and 2014-15. One year the leader has 1.300 points/game, and the other he has 1.091. The leader is, of course, Sidney Crosby. By using Crosby as the benchmark for comparison (since he's the points/game leader), it seems that the entire league improved by 19% going from one season to the next... and then suddenly collapsed again when Kane took the points/game throne in 2015-16 (1.293 points/game).

Using the leader as the benchmark for any comparison is not particularly useful, especially when the leader is typically an outlier among outliers. Scoring is not a test where errors are noted and removed from "100". Rather, scoring is entirely additive. (Goaltending, on the other hand...)

That being said, I do recognize that a top-10 finish in different eras are not made equally. Perhaps a comparison with 3rd on the leaderboard would be more useful. That should be high enough to not be significantly different across various eras while low enough to not be an outlier in most years.

 

[Baby Punisher]

That's a very impressive stat. More so for Lemuiex. It is very impressive considering the health issues he faced and retiring several times in his career only to come back and still be at the top of his game.

 

[seventieslord]

Consider the points-per-game leader of 2013-14 and 2014-15. One year the leader has 1.300 points/game, and the other he has 1.091. The leader is, of course, Sidney Crosby. By using Crosby as the benchmark for comparison (since he's the points/game leader), it seems that the entire league improved by 19% going from one season to the next... and then suddenly collapsed again when Kane took the points/game throne in 2015-16 (1.293 points/game).

Using the leader as the benchmark for any comparison is not particularly useful, especially when the leader is typically an outlier among outliers. Scoring is not a test where errors are noted and removed from "100". Rather, scoring is entirely additive. (Goaltending, on the other hand...)

That being said, I do recognize that a top-10 finish in different eras are not made equally. Perhaps a comparison with 3rd on the leaderboard would be more useful. That should be high enough to not be significantly different across various eras while low enough to not be an outlier in most years.

that's pretty much VsX.

 

[daver]

Consider the points-per-game leader of 2013-14 and 2014-15. One year the leader has 1.300 points/game, and the other he has 1.091. The leader is, of course, Sidney Crosby. By using Crosby as the benchmark for comparison (since he's the points/game leader), it seems that the entire league improved by 19% going from one season to the next... and then suddenly collapsed again when Kane took the points/game throne in 2015-16 (1.293 points/game).

Using the leader as the benchmark for any comparison is not particularly useful, especially when the leader is typically an outlier among outliers. Scoring is not a test where errors are noted and removed from "100". Rather, scoring is entirely additive. (Goaltending, on the other hand...)

That being said, I do recognize that a top-10 finish in different eras are not made equally. Perhaps a comparison with 3rd on the leaderboard would be more useful. That should be high enough to not be significantly different across various eras while low enough to not be an outlier in most years.

I agree that using one stat, be it the points leader or # 2, or #3 etc.. should always be scrutinized. I prefer using a stat that reflects the average of the peer group e,g. Top 10, 20, #1 line forwards etc..

What I am showing is the # players in that peer group needs to be adjusted depending on league size. I can guarantee you that Art Ross winners from the O6 are going to show more impressive showings in comparison to the other Top 10 scorers than the Art Ross winners from the past 20 year are going to show.

Specific to Crosby, Hull and Beliveau though, we have up to 15 seasons of data to use so it should flatten any statistical anomalies like the 14/15 season which saw a weak Art Ross winner and the 2018/19 season that saw a very strong Art Ross winner.

Question for you

If you accept the numbers I provided that show, on average, the #5 scorer from the O6 and the #10 scorer from the current era seem to be equal, would you think it is reasonable to establish the peer group for Hull and Beliveau as being the Top 5 scorers and establish the peer group for Crosby as the Top 10 scorers?

 

[daver]

that's pretty much VsX.

The issue is that the average #3 scorer was farther away from the leader in the O6 than in the current era (14% vs. 11%). Using the #5 scorer in the current era would be a fairer measurement (14% to 15%).

Or use the #5 scorer from the O6 vs. the #10 scorer from the current era (21% vs. 21%).

 

[plusandminus]

I think these "benchmarks", and "vsx" things, sometimes compares apples to oranges... And why ever compare to the scoring leader? Or even the guy finishing 2nd or 3rd? Sometimes completely different "benchmarks" and circumstances.
Why not just look at plain scoring finishes among Canadians/North Americans instead? In Crosby's case points per game scoring finishes might be interesting.

 

[Black Gold Extractor]

I agree that using one stat, be it the points leader or # 2, or #3 etc.. should always be scrutinized. I prefer using a stat that reflects the average of the peer group e,g. Top 10, 20, #1 line forwards etc..

What I am showing is the # players in that peer group needs to be adjusted depending on league size. I can guarantee you that Art Ross winners from the O6 are going to show more impressive showings in comparison to the other Top 10 scorers than the Art Ross winners from the past 20 year are going to show.

Specific to Crosby, Hull and Beliveau though, we have up to 15 seasons of data to use so it should flatten any statistical anomalies like the 14/15 season which saw a weak Art Ross winner and the 2018/19 season that saw a very strong Art Ross winner.

Question for you

If you accept the numbers I provided that show, on average, the #5 scorer from the O6 and the #10 scorer from the current era seem to be equal, would you think it is reasonable to establish the peer group for Hull and Beliveau as being the Top 5 scorers and establish the peer group for Crosby as the Top 10 scorers?

That's reasonable, although I'd consider the talent pool rather than simply league size. The league was way bigger in 1972-73 than 1966-67, but I doubt that there's significant difference in the talent pool (rather, it might even be weaker in 1972-73 due to the WHA). For whatever reason, it seems to correlate relatively well with the percentage of Canadians in the NHL (at least if we disregard outside events, like WWII or the WHA). Historically, that would mean maybe #7 for the 1980's, #8 for the 1990's (75% of the NHL was Canadian at the end of the 80's going into the 90's), #9 for the 2000's, #10 for the 2010's (roughly half of the NHL was Canadian in 2015-16), and possibly #11 for the 2020's (only 43% of the NHL is now Canadian).

I've also noted a strengthening of the league using a couple of different methods. The quality of a "typical 2nd-place scorer" has improved at a rate of roughly 5% per century. As such, if you're looking at a a VsX score from 50 years ago, it's probably ~2.5% higher than it should be compared with a modern-day VsX score. Whether that makes any real difference is up to the reader to decide.

 

[wetcoast]

I would have a hard time downplaying Crosby's ''record'' when Lemieux's streak had the following seasons.

26 games out of 80
22 games out of 82
43 games out of 82
24 games out of 82

Fully agree here and finding it funny that people focused on games missed by one guy and not the other one

Also, not this thread is about this, but it was mentioned above, the longer That Crosby stays at his current level the more we might see a legit argue meant for a Big 5; that is if people look past video game stats and rank actual full impacts of players involved.

 

[wetcoast]

I think these "benchmarks", and "vsx" things, sometimes compares apples to oranges... And why ever compare to the scoring leader? Or even the guy finishing 2nd or 3rd? Sometimes completely different "benchmarks" and circumstances.
Why not just look at plain scoring finishes among Canadians/North Americans instead? In Crosby's case points per game scoring finishes might be interesting.

Very interesting way to look at it.

Of course comparing players from different eras to a common factor would make for a much stronger argument than simply counting top 10 finishes right?

 

[daver]

I've also noted a strengthening of the league using a couple of different methods. The quality of a "typical 2nd-place scorer" has improved at a rate of roughly 5% per century. As such, if you're looking at a a VsX score from 50 years ago, it's probably ~2.5% higher than it should be compared with a modern-day VsX score. Whether that makes any real difference is up to the reader to decide.

From 1949/50 to 1974/75, the 2nd place scorer was, on average, 12.5% behind 1st.

From 1989/90 to 2019/20, the 2nd place scorer was, on, average, 8.5% behind 1st.

So pretty close to what you have come up with. IMO, it's enough of a difference to give a marginal edge over another, in a strictly a statistical discussion.


Using VsX (#2 scorer) for Belliveau, Crosby and Hull, the numbers do not change much vs. a VsX (#1 scorer).

In their 12 best point finishes

Crosby was, on average, 6% behind 2nd place (best was 20% ahead, and worst was 21% behind)

Hull was, on average, 8.5% behind 12nd place (best was 24% ahead, and worst was 31% behind)

Beliveau, was, on average, 10% behind 2nd place (best was 11% ahead, and worst was 36% behind)


In their 11 best PPG finishes*

Crosby was, on average, 5.5% ahead of 2nd place (best was 27% ahead, and worst was 3% behind)**

Hull was, on average, 4.5% behind 12nd place (best was 30% ahead, and worst was 23% behind)

Beliveau, was, on average, 2.5% behind 2nd place (best was 11% ahead, and worst was 21% behind)


* Hull's 12th best was 45%, a statistical anomaly, IMO.

** assumed a tie in PPG with Malkin in 11/12, too many missed games

 

[daver]

Making it their best 14 best PPG finishes which takes Hull out of the picture:

Crosby was, on average, 0.5% ahead of 2nd place (best was 27% ahead, and worst was 20% behind)

Beliveau, was, on average, 7.5% behind 2nd place (best was 11% ahead, and worst was 30% behind)


For comparison:

In his 11 best PPG finishes, Howe was, on average, 8% ahead of 2nd place (best was 35% ahead, and worst was 7% behind)

In his 14 best PPG finishes, Howe was, on average, 4.5% ahead of 2nd place (best was 35% ahead, and worst was 10% behind)

 

[daver]

More comparison:

In his 11 best PPG finishes, Jagr was, on average, 2% behind 2nd place (best was 13% ahead, and worst was 19% behind)

In his 11 best point finishes, Jagr was, on average, 3% behind of 2nd place (best was 19% ahead, and worst was 19% behind)

In his 14 best PPG finishes, Jagr was, on average, 11% behind 2nd place (best was 13% ahead, and worst was 48% behind)

In his 14 best point finishes, Jagr was, on average, 12% behind of 2nd place (best was 19% ahead, and worst was 49% behind)

 

[daver]

One more piece of data re: comparing Top Ten from different eras:

From 1949/50 to 1969/70 there were:

68 different players who had a Top Ten finish in scoring

38 (56%) of those players had multiple Top Ten finishes

28 (41%) of those players had three or more Top Ten finishes

11 (16%) of those players had five or more Top Ten finishes

2 (3%) of those players had ten or more Top Ten finishes


From 1998/99 to 2019/20 there were:

85 different players who had a Top Ten finish in scoring

49 (58%) of those players had multiple Top Ten finishes

30 (35%) of those players had three or more Top Ten finishes

9 (11%) of those players had five or more Top Ten finishes

1 (1%) of those players had ten or more Top Ten finishes


In the current era, it is more of an accomplishment to finish in the Top Ten and an even bigger relative accomplishment to have five or more Art Ross finishes.