Lemieux vs Gretzky - who had the Highest Offensive Peak? A thorough statistical analysis

[authentic]

(I thought I was done reading this tired thread, but somehow I got pulled back in....)

A few thoughts:

-- You seem a bit over-focused on the "early 80s" as if Gretzky's prime ended in 1984 or something, and not in 1991 as in reality (and only then with injury advancing it). Gretzky was the most dominant athlete in hockey, if not in the world, in 1985, 1986, and 1987. He was also the best player in 1988, and won scoring titles on 1990 and 1991 (and 1994, but by then he was way past his prime). We're now 10 years past the "early 80s", so I'm not sure why you're interring Gretzky in that era, which he clearly transcended.

-- "Gretzky's teammates are all Hall of Famers" is just a slight over-statement. We pause here to remember that Gretzky had the highest-scoring season in NHL history when he was still eligible for junior hockey and when he had exactly ZERO all-star teammates (1980-81), You may want to dismiss that by saying "early 80s", but Gretzky also scored 163 points (78 games) in 1991, a decade later, including 103 even strength points.

Mario had two Gretzky-like even strength points seasons, which were of course 1988-89 and 1992-93. If we project the ES points totals from after 1985 (as you dismiss the "early 80s", and so the two players' careers start at the same time) to 80-game seasons, we get this:
143 - Gretzky 1986
128 - Lemieux 1993 (didn't actually play full season)
126 - Gretzky 1987
114 - Gretzky 1988 (didn't actually play full season)
107 - Lemieux 1989
106 - Gretzky 1991
105 - Gretzky 1990
103 - Gretzky 1989

The only season here where Gretzky had the fully stacked line-up of productive Oilers' is 1986. Essentially, there were four other Oilers who consistently helped Gretzky score points -- Kurri, Messier, Coffey, Anderson. In 1987, Anderson was in decline and Coffey injured and having an off-season. In 1988 Coffey was long gone, Kurri and Anderson not as dominant. And you'll note positions 4, 5, 6, 7, and 8 on the above list are all Gretzky seasons in Los Angeles.

So, how does this (half of Gretzky's prime) compare with Lemieux's two (and only two) dominant ES points seasons? I personally doubt he hits 128 in 1993 in an 80-game season, but let's say 125 for an even number. Okay, so that's an utterly fabulous ES points season, at the all-time elite (Gretzky) level. Now, let's check which teammates were helping Lemieux score that season:
- Stevens, Tocchet, Francis, Jagr, Murphy (Mullen arguably)
The Pen's 1993 line-up of offensive players easily matches and probably beats any line-up the Oilers of the 1980s ever assembled (with the possible exception of spring 1987, a brief three-month period when the Oilers added Nilsson and Ruotsalainen... during which Gretzky's stats went down, not up). So, this idea that Lemieux did a Gretzky-like season with lesser teammates certainly doesn't hold up as far as 1992-93 goes (the focus of this thread, more or less).

From this perspective, I have always maintained that 1988-89 was Lemieux's best season and was his physical prime. His ES points production, though, is no better than Gretzky's in 1991 on the Kings, a club that scored (slightly) fewer goals than the 1988-89 Penguins. The 1989 Penguins could score, but Lemieux was the straw that stirred the drink, and they went as far as he took them only.

So, you keep going on about the "early 80s", but the record shows that Gretzky outscored Lemieux from 1986 to 1991, and massively outscored him at even strength even while in Los Angeles, with the notable exception of 1992-93, though we'll never know how that season would have ended if Mario had played 80 games. And that season, the Penguins were a great team with or without Lemieux.

If you read Staticans post above you'll see how this doesn't make any sense. It isn't a fair comparison to begin with because Lemieux and his team spent more total ice time on the powerplay and PK, thus playing far fewer ES minutes.

 

[tabness]

Here is the table of "important points" (points on goals that tied/broke the tie), as while the way I defined blowout points (leading by 4 or more) probably makes them less important than rally points (trailing by 2 or more), you can't just look at blowout points alone given the different fortunes of Gretzky's and Lemieux's teams throughout different stages of each player's career:

I'll mostly refrain from commenting on the data lol, but I will note that just eyeballing these stats vs other players, both Gretzky and Lemieux are definitely among the lowest in percentage of "important points"

---------------GRETZKY-----------------LEMIEUX------------

year	important	total	%	important	total	%
1980​	52​	137​	37.96%​
1981​	66​	157​	42.04%​
1982​	86​	210​	40.95%​
1983​	76​	196​	38.78%​
1984​	62​	205​	30.24%​
1985​	90​	205​	43.90%​	50​	98​	51.02%​
1986​	80​	211​	37.91%​	69​	135​	51.11%​
1987​	64​	177​	36.16%​	55​	106​	51.89%​
1988​	56​	148​	37.84%​	77​	162​	47.53%​
1989​	83​	164​	50.61%​	73​	182​	40.11%​
1990​	61​	138​	44.20%​	50​	122​	40.98%​
1991​	61​	160​	38.13%​	19​	45​	42.22%​
1992​	53​	117​	45.30%​	50​	131​	38.17%​
1993​	30​	63​	47.62%​	66​	155​	42.58%​
1994​	59​	128​	46.09%​	19​	37​	51.35%​
1995​	25​	48​	52.08%​
1996​	43​	102​	42.16%​	63​	155​	40.65%​
1997​	45​	95​	47.37%​	49​	121​	40.50%​
1998​	50​	89​	56.18%​
1999​	27​	62​	43.55%​
2000​
2001​				43​	75​	57.33%​
2002​				14​	31​	45.16%​
2003​				43​	85​	50.59%​
2004​				5​	9​	55.56%​
2006​				10​	22​	45.45%​

 

[ContrarianGoaltender]

By the logic of the OP, Mario Lemieux isn't the greatest power play scorer of all-time. He has now been surpassed by Connor McDavid:

PP Performance, Adjusted by League Average PPG Per Game:

Season	Player	LgAvg PPG	GP	PPG	PPA	PPP	Adj PPP / 82 GP
2023​	McDavid	0.67​	62​	18​	37​	55​	97.7​
1996​	Lemieux	0.90​	70​	31​	48​	79​	92.5​
2021​	McDavid	0.57​	56​	9​	28​	37​	85.5​
2020​	McDavid	0.60​	64​	11​	32​	43​	82.6​
1989​	Lemieux	1.06​	76​	31​	48​	79​	72.4​
2001​	Lemieux	0.76​	43​	16​	16​	32​	72.3​
1988​	Lemieux	1.11​	77​	22​	58​	80​	69.1​
2003​	Lemieux	0.73​	67​	14​	31​	45​	67.9​
2022​	McDavid	0.60​	80​	10​	34​	44​	67.7​
2002​	Lemieux	0.65​	24​	2​	12​	14​	66.2​
1993​	Lemieux	1.03​	60​	16​	39​	55​	65.7​
1990​	Lemieux	0.95​	59​	14​	35​	49​	64.5​
1986​	Lemieux	1.02​	79​	17​	49​	66​	60.4​
1994​	Lemieux	0.90​	22​	7​	8​	15​	55.9​
1992​	Lemieux	0.97​	64​	12​	35​	47​	55.9​
2019​	McDavid	0.58​	78​	9​	24​	33​	53.8​
1997​	Lemieux	0.67​	76​	15​	22​	37​	53.6​
1987​	Lemieux	0.90​	63​	19​	19​	38​	49.5​
2017​	McDavid	0.57​	82​	3​	24​	27​	42.6​
2004​	Lemieux	0.70​	10​	0​	4​	4​	42.2​
1991​	Lemieux	0.89​	26​	6​	7​	13​	41.5​
2016​	McDavid	0.58​	45​	3​	11​	14​	39.6​
1985​	Lemieux	0.89​	73​	11​	22​	33​	37.5​
2018​	McDavid	0.61​	82​	5​	15​	20​	29.5​

(Note: Adjusted PPP is normalized to a 0.90 PPG environment)

McDavid is currently having a better power play season relative to the league average power play goal rate than Lemieux ever did, and McDavid also currently has 3 of the top 4 seasons between the two of them. Seems like a pretty open-and-shut case, especially if we're focusing on peak (the way the OP focused on peak ES scoring for Lemieux in 1993 vs. Gretzky's longer extended prime, for example).

Now, you might be saying, that doesn't seem to make sense, it doesn't really match the eye test, McDavid is historically great on the PP but it seems unlikely he is surpassing the greatness of Mario. And I would completely agree with you.

It turns out if you use a sensible adjustment instead of league average goals per game, such as, say, the scoring rate of the #10 scorer in the league in PPP, then it turns out that Lemieux still has the best peak season, and McDavid's current season is just slightly above Lemieux in 1988 and 1989:

PP Performance, Adjusted by #10 in PP Scoring:

Season	Player	#10 PPP	GP	#10 PPP/GP	GP	PPG	PPA	PPP	Adj PPP
1996​	Lemieux	41​	82​	0.50​	70​	31​	48​	79​	88.8​
2023​	McDavid	28​	60​	0.47​	62​	18​	37​	55​	74.8​
1989​	Lemieux	44​	80​	0.55​	76​	31​	48​	79​	74.4​
1988​	Lemieux	45​	80​	0.56​	77​	22​	58​	80​	72.7​
2020​	McDavid	27​	71​	0.38​	64​	11​	32​	43​	69.5​
2021​	McDavid	22​	56​	0.39​	56​	9​	28​	37​	66.2​
2002​	Lemieux	29​	82​	0.35​	24​	2​	12​	14​	64.9​
2001​	Lemieux	37​	82​	0.45​	43​	16​	16​	32​	64.9​
2003​	Lemieux	35​	82​	0.43​	67​	14​	31​	45​	61.9​
1986​	Lemieux	43​	80​	0.54​	79​	17​	49​	66​	61.2​
1990​	Lemieux	43​	80​	0.54​	59​	14​	35​	49​	60.8​
1993​	Lemieux	51​	84​	0.61​	60​	16​	39​	55​	59.4​
1994​	Lemieux	41​	84​	0.49​	22​	7​	8​	15​	55.0​
1987​	Lemieux	35​	80​	0.44​	63​	19​	19​	38​	54.3​
2022​	McDavid	35​	82​	0.43​	80​	10​	34​	44​	50.7​
1997​	Lemieux	31​	82​	0.38​	76​	15​	22​	37​	50.7​
1992​	Lemieux	46​	80​	0.58​	64​	12​	35​	47​	50.3​
2004​	Lemieux	31​	82​	0.38​	10​	0​	4​	4​	41.6​
2019​	McDavid	33​	82​	0.40​	78​	9​	24​	33​	41.4​
1985​	Lemieux	36​	80​	0.45​	73​	11​	22​	33​	39.5​
1991​	Lemieux	40​	80​	0.50​	26​	6​	7​	13​	39.4​
2017​	McDavid	27​	82​	0.33​	82​	3​	24​	27​	39.4​
2016​	McDavid	27​	82​	0.33​	45​	3​	11​	14​	37.2​
2018​	McDavid	33​	82​	0.40​	82​	5​	15​	20​	23.9​

I do think McDavid is in the middle of rocketing up the list in the conversation of the greatest power play performers ever. But I don't think he has yet surpassed Lemieux, and I think the above ranking is much more realistic than the previous one.

I continue to maintain that there is no good reason to ever use league average goals per game as the basis for adjusting stats. There seems to be a strong contingent of posters here who love those adjustments because they make guys who played in the '90s look better than guys who played in the '80s, and that seems to justify their subjective opinions about the relative strengths of those eras. I'm just saying that even though that might feed your confirmation bias in those comparisons, it can become a double-edged sword if you start moving to other eras, like how you now have to bow to McDavid as the top PP scorer ever or get called out for being inconsistent.

And yes, I know that using VsX or Vs10 or similar estimates of scoring levels for top players can be quite variable over the course of one season, they can greatly depend on a couple of key injuries or how many super teams or generational talents might be active in the league, and a result they can sometimes be a bit fuzzy from year to year. But they're still way better than using league average rates, which give a false precision and simply don't accurately tell us what we really want to know, which is how easy it is to score points for a top-line NHL forward.

 

[authentic]

By the logic of the OP, Mario Lemieux isn't the greatest power play scorer of all-time. He has now been surpassed by Connor McDavid:

PP Performance, Adjusted by League Average PPG Per Game:

Season	Player	LgAvg PPG	GP	PPG	PPA	PPP	Adj PPP / 82 GP
2023​	McDavid	0.67​	62​	18​	37​	55​	97.7​
1996​	Lemieux	0.90​	70​	31​	48​	79​	92.5​
2021​	McDavid	0.57​	56​	9​	28​	37​	85.5​
2020​	McDavid	0.60​	64​	11​	32​	43​	82.6​
1989​	Lemieux	1.06​	76​	31​	48​	79​	72.4​
2001​	Lemieux	0.76​	43​	16​	16​	32​	72.3​
1988​	Lemieux	1.11​	77​	22​	58​	80​	69.1​
2003​	Lemieux	0.73​	67​	14​	31​	45​	67.9​
2022​	McDavid	0.60​	80​	10​	34​	44​	67.7​
2002​	Lemieux	0.65​	24​	2​	12​	14​	66.2​
1993​	Lemieux	1.03​	60​	16​	39​	55​	65.7​
1990​	Lemieux	0.95​	59​	14​	35​	49​	64.5​
1986​	Lemieux	1.02​	79​	17​	49​	66​	60.4​
1994​	Lemieux	0.90​	22​	7​	8​	15​	55.9​
1992​	Lemieux	0.97​	64​	12​	35​	47​	55.9​
2019​	McDavid	0.58​	78​	9​	24​	33​	53.8​
1997​	Lemieux	0.67​	76​	15​	22​	37​	53.6​
1987​	Lemieux	0.90​	63​	19​	19​	38​	49.5​
2017​	McDavid	0.57​	82​	3​	24​	27​	42.6​
2004​	Lemieux	0.70​	10​	0​	4​	4​	42.2​
1991​	Lemieux	0.89​	26​	6​	7​	13​	41.5​
2016​	McDavid	0.58​	45​	3​	11​	14​	39.6​
1985​	Lemieux	0.89​	73​	11​	22​	33​	37.5​
2018​	McDavid	0.61​	82​	5​	15​	20​	29.5​

(Note: Adjusted PPP is normalized to a 0.90 PPG environment)

McDavid is currently having a better power play season relative to the league average power play goal rate than Lemieux ever did, and McDavid also currently has 3 of the top 4 seasons between the two of them. Seems like a pretty open-and-shut case, especially if we're focusing on peak (the way the OP focused on peak ES scoring for Lemieux in 1993 vs. Gretzky's longer extended prime, for example).

Now, you might be saying, that doesn't seem to make sense, it doesn't really match the eye test, McDavid is historically great on the PP but it seems unlikely he is surpassing the greatness of Mario. And I would completely agree with you.

It turns out if you use a sensible adjustment instead of league average goals per game, such as, say, the scoring rate of the #10 scorer in the league in PPP, then it turns out that Lemieux still has the best peak season, and McDavid's current season is just slightly above Lemieux in 1988 and 1989:

PP Performance, Adjusted by #10 in PP Scoring:

Season	Player	#10 PPP	GP	#10 PPP/GP	GP	PPG	PPA	PPP	Adj PPP
1996​	Lemieux	41​	82​	0.50​	70​	31​	48​	79​	88.8​
2023​	McDavid	28​	60​	0.47​	62​	18​	37​	55​	74.8​
1989​	Lemieux	44​	80​	0.55​	76​	31​	48​	79​	74.4​
1988​	Lemieux	45​	80​	0.56​	77​	22​	58​	80​	72.7​
2020​	McDavid	27​	71​	0.38​	64​	11​	32​	43​	69.5​
2021​	McDavid	22​	56​	0.39​	56​	9​	28​	37​	66.2​
2002​	Lemieux	29​	82​	0.35​	24​	2​	12​	14​	64.9​
2001​	Lemieux	37​	82​	0.45​	43​	16​	16​	32​	64.9​
2003​	Lemieux	35​	82​	0.43​	67​	14​	31​	45​	61.9​
1986​	Lemieux	43​	80​	0.54​	79​	17​	49​	66​	61.2​
1990​	Lemieux	43​	80​	0.54​	59​	14​	35​	49​	60.8​
1993​	Lemieux	51​	84​	0.61​	60​	16​	39​	55​	59.4​
1994​	Lemieux	41​	84​	0.49​	22​	7​	8​	15​	55.0​
1987​	Lemieux	35​	80​	0.44​	63​	19​	19​	38​	54.3​
2022​	McDavid	35​	82​	0.43​	80​	10​	34​	44​	50.7​
1997​	Lemieux	31​	82​	0.38​	76​	15​	22​	37​	50.7​
1992​	Lemieux	46​	80​	0.58​	64​	12​	35​	47​	50.3​
2004​	Lemieux	31​	82​	0.38​	10​	0​	4​	4​	41.6​
2019​	McDavid	33​	82​	0.40​	78​	9​	24​	33​	41.4​
1985​	Lemieux	36​	80​	0.45​	73​	11​	22​	33​	39.5​
1991​	Lemieux	40​	80​	0.50​	26​	6​	7​	13​	39.4​
2017​	McDavid	27​	82​	0.33​	82​	3​	24​	27​	39.4​
2016​	McDavid	27​	82​	0.33​	45​	3​	11​	14​	37.2​
2018​	McDavid	33​	82​	0.40​	82​	5​	15​	20​	23.9​

I do think McDavid is in the middle of rocketing up the list in the conversation of the greatest power play performers ever. But I don't think he has yet surpassed Lemieux, and I think the above ranking is much more realistic than the previous one.

I continue to maintain that there is no good reason to ever use league average goals per game as the basis for adjusting stats. There seems to be a strong contingent of posters here who love those adjustments because they make guys who played in the '90s look better than guys who played in the '80s, and that seems to justify their subjective opinions about the relative strengths of those eras. I'm just saying that even though that might feed your confirmation bias in those comparisons, it can become a double-edged sword if you start moving to other eras, like how you now have to bow to McDavid as the top PP scorer ever or get called out for being inconsistent.

And yes, I know that using VsX or Vs10 or similar estimates of scoring levels for top players can be quite variable over the course of one season, they can greatly depend on a couple of key injuries or how many super teams or generational talents might be active in the league, and a result they can sometimes be a bit fuzzy from year to year. But they're still way better than using league average rates, which give a false precision and simply don't accurately tell us what we really want to know, which is how easy it is to score points for a top-line NHL forward.

The Oilers have the best powerplay % since 1977, largely because of McDavid. I don't think this is such an open and shut case as you present it here. It's seems pretty plausible to me that this could be the best powerplay performance ever. He basically has as many powerplay points as anyone else has ES points.

 

[tabness]

By the logic of the OP, Mario Lemieux isn't the greatest power play scorer of all-time. He has now been surpassed by Connor McDavid:

It's a bit simplistic and uncharitable to simply hone in on the one factor of adjusted points that @TheStatican was making no? The scoring levels/help from teammates stuff is old news anyway, for me at least, the interesting part which hasn't gotten much honest engagement was the stat padding stuff.

But I guess that's to be expected, very entrenched paradigm here on ranking and all that, Lemieux just don't got the Hart Trophies or the VsX lol

 

[TheStatican]

Here is the table of "important points" (points on goals that tied/broke the tie), as while the way I defined blowout points (leading by 4 or more) probably makes them less important than rally points (trailing by 2 or more), you can't just look at blowout points alone given the different fortunes of Gretzky's and Lemieux's teams throughout different stages of each player's career:


---------------GRETZKY-----------------LEMIEUX------------

year	important	total	%	important	total	%
1982​	86​	210​	40.95%​
1983​	76​	196​	38.78%​
1984​	62​	205​	30.24%​
1985​	90​	205​	43.90%​	50​	98​	51.02%​
1986​	80​	211​	37.91%​	69​	135​	51.11%​
1987​	64​	177​	36.16%​	55​	106​	51.89%​
1988​	56​	148​	37.84%​	77​	162​	47.53%​
1989​	83​	164​	50.61%​	73​	182​	40.11%​
1993​	30​	63​	47.62%​	66​	155​	42.58%​
1996​	43​	102​	42.16%​	63​	155​	40.65%​

Blowout points taken from @tabness chart;

Gretzky
1982 blowout: 34 16.19%
1983 blowout: 32 16.33%
1984 blowout: 46 22.44%
1985 blowout: 29 14.15%
1986 blowout: 36 17.06%
1987 blowout: 22 12.43%

Lemieux
1989 blowout: 22 12.09%
1993 blowout: 22 14.19%
1996 blowout: 25 16.13%

This corroborates the data I was pointing out in my OP. The one year where Gretzky exceeds his already extremely high plateau is the year in which he scores by far both the highest percentage of blowout points and the lowest percentage of important points of the top 10 greatest offensive seasons.

Honestly I'm starting to believe 84-85 was Gretzky's best season and that I should be comparing Lemieux best with that year.

 

[TheStatican]

By the logic of the OP, Mario Lemieux isn't the greatest power play scorer of all-time. He has now been surpassed by Connor McDavid

Absolutely I would consider McDavid's season to be in the discussion for the greatest Powerplay performance of all time. There's no question at all he is rivaling Lemieux's 95-96 season. I mean the Oilers of this year are on pace to challenge the NHL record for powerplay efficiency, that doesn't just happen without a player having a historical seasons. We'll see if he's able to maintain until the end of the season, Lemieux's started off 95-96 at an absurd pace much higher than even McDavid's this year; A little past the halfway point of the season he was averaging 1.40 powerplay points a game with a team PP rate of 31.5%(when he played), but it declined as the season progressed.

But that's just hardily the only metric upon which I was measuring players. It's become quite clear that Lemieux was just as good as an even strength scorer as Gretzky by the fact that he had the highest adjusted EV scoring season between the two in 92-93 and his per/60 EV rate in 88-89 may very well have been even higher, quite possibly the best all time on a per/60 scoring basis.

 

[TheStatican]

Here's some even more interesting data, the minute by minute scoring of both Gretzky's and Lemieux best 60 game spans;

Game Time In mins​	WG 83-84​	ML 92-93​
1​	1​	3​
2​	3​	7​
3​	4​	9​
4​	5​	14​
5​	5​	18​
6​	6​	20​
7​	8​	21​
8​	13​	24​
9​	14​	25​
10​	18​	27​
11​	22​	29​
12​	25​	30​
13​	25​	32​
14​	28​	33​
15​	30​	36​
16​	33​	38​
17​	37​	41​
18​	39​	45​
19​	44​	47​
20​	51​	51​
21​	56​	53​
22​	59​	57​
23​	64​	59​
24​	66​	64​
25​	71​	65​
26​	72​	68​
27​	78​	72​
28​	80​	75​
29​	82​	78​
30​	83​	82​
31​	85​	82​
32​	89​	84​
33​	93​	88​
34​	97​	95​
35​	100​	101​
36​	105​	104​
37​	112​	107​
38​	116​	110​
39​	118​	114​
40​	119​	119​
41​	120​	120​
42​	124​	122​
43​	127​	125​
44​	127​	126​
45​	128​	127​
46​	131​	129​
47​	133​	131​
48​	135​	133​
49​	135​	134​
50​	137​	136​
51​	141​	137​
52​	146​	140​
53​	147​	143​
54​	147​	144​
55​	149​	148​
56​	153​	149​
57​	155​	151​
58​	158​	151​
59​	163​	156​
60​	174​	158​
61​	​	​
62​	​	​
63​	​	159​
64​	175​	160​

These charts will show things more clearly. Not that I need to remind anyone but Lemieux was getting radiation treatments during this time, it matters because there's no way Lemieux was at 100% in his first few games back. Total points by the minute;

Gretzky's tiny edge throughout most of the 3rd was solely because Lemieux played 59 and one third games, missing the 2ne & 3rd periods in one game verse a full 60 games for Gretzky. Right up the the 55th minute they were exactly tied in points per game. Points per game by the minute;

If we use the season long 74 game metrics for Gretzky, Lemieux has the clear edge all the way until the final
minute

Additionally Bowman clearly reduced Mario's playtime in the 3rd during blowouts, hence why his PPG was dropping throughout the third. Just compare the two teams overall scoring rates per period;

92-93 Pens(in Mario's 60)	Goals For​	GF% by Period​	Goals Against​	Ratio For/Against​
1st Period	100​	34.2%​	53​	1.9​
2nd Period	112​	38.4%​	71​	1.6​
3rd Period	77​	26.3%​	71​	1.1​
OT	3​	1.3%​	0​
Total	292​		195​	1.5​

83-84 Oilers(first 60)	Goals For​	GF% by Period​	Goals Against​	Ratio For/Against​
1st Period	102​	30.7%​	82​	1.2​
2nd Period	120​	36.1%​	98​	1.2​
3rd Period	107​	32.2%​	69​	1.6​
OT	2​	1.3%​	0​
Total	332​		249​	1.4​

Lemieux was matching Gretzky at his 175 point in 60 game pace. The only reason Gretzky came out ahead or instead of being tied is because he decided he was going to pad his numbers as much as possible at the end of blowout games in 1983-84. This chart attests to that - Their points in each minute of the game;

 

[daver]

Waaaaaaay too much info in this thread.

Let's just ask Chatbot to figure this out.

 

[The Panther]

It's become quite clear that Lemieux was just as good as an even strength scorer as Gretzky by the fact that he had the highest adjusted EV scoring season between the two in 92-93 and his per/60 EV rate in 88-89 may very well have been even higher, quite possibly the best all time on a per/60 scoring basis.

Wait, what???

No, no it isn't clear. In fact, the exact opposite of what you're (obsessively) arguing is 100% clear.

At this point, I can't remember if you detailed your mathematical determination of "adjusting" ES points, but as you surely know any particular method is going to have its flaws. We can't depend on one system to give us the be-all / end-all "adjusted" points. (I might add -- do we really need to adjust points when the two players' careers overlapped??) Then, how in the world can you mathematically calculate the two players' per/60 when we don't know how much ES (or total) ice-time each player had??

Here's what we do know: The raw numbers of ES points / game and in (parenthesis) the rank of each player (among the top-10 ES overall producers):

1979-80
1.27 Gretzky (1)
1980-81
1.3 Gretzky (1)
1981-82
1.84 Gretzky (1)
1982-83
1.65 Gretzky (1)
1983-84
1.82 Gretzky (1)
1984-85
1.83 Gretzky (1)
0..92 Lemieux
1985-86
1.79 Gretzky (1)
0.95 Lemieux (5)
1986-87
1.57 Gretzky (1)
1.20 Lemieux (2)
1987-88
1.42 Gretzky (1)
0.96 Lemieux (3)

(Gretzky traded to L.A.)

1988-89
1.34 Lemieux (1)
1.28 Gretzky (2)
1989-90
1.32 Gretzky (1)
1.20 Lemieux (2)
1990-91
1.32 Gretzky (1)
(Lemieux barely played and thus isn't eligible, but was at 1.19.)

(Gretzky's prime decisively ends)

1991-92
1.16 Lemieux (1)
0.85 Gretzky (4)
1992-93
1.60 Lemieux (1)
(Gretzky played only half the season and isn't eligible, but was at 0.84.)
1993-94
0.77 Gretzky (7)
(Lemieux barely played and thus isn't eligible, but was at 1.00.)
1995
0.48 Gretzky
1995-96
1.04 Lemieux (2)
0.68 Gretzky
1996-97
1.06 Lemieux (2)
0.79 Gretzky (8)
1997-98
0.73 Gretzky (3)

So, first of all, forget 'adjusting' anything and let's just look at how each player finished in ES points-per-game (again, these rankings are only of the top-10 overall ES scorers):
Gretzky -- 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1
Lemieux -- 1, 1, 1, 2, 2, 2, 2, 3, 5

Top raw ES/game seasons:
1.84 - Gretzky
1.83 - Gretzky
1.82 - Gretzky
1.79 - Gretzky
1.65 - Gretzky
1.60 - Lemieux
1.57 - Gretzky
1.42 - Gretzky
1.34 - Lemieux
1.32 - Gretzky
1.32 - Gretzky
1.30 - Gretzky
1.28 - Gretzky
1.27 - Gretzky
1.20 - Lemieux
1.16 - Lemieux
1.05 - Lemieux
1.04 - Lemieux

So, Gretzky has 7 of the top 8 (88%), and 12 of the top 14 (86%).

But you will argue, we have to adjust those points!! Well, we can't do that any reliable formula. But how about we look at the only four seasons when the two players' primes directly overlapped---namely, 1987-88 to 1990-91 (for most of which, Gretzky was in Los Angeles):
1.33 -- Gretzky
1.17 -- Lemieux

But you will argue, this thread is about peak, not about four seasons!

Right you are! So, whose peak is more impressive?: The guy who averaged 1.71 ES points-per-game for 7 consecutive seasons (while appearing in 96% of his club's games), or the guy who averaged 1.60 ES points-per-game for sixty games?

 

[TheStatican]

At this point, I can't remember if you detailed your mathematical determination of "adjusting" ES points, but as you surely know any particular method is going to have its flaws. We can't depend on one system to give us the be-all / end-all "adjusted" points. (I might add -- do we really need to adjust points when the two players' careers overlapped??) Then, how in the world can you mathematically calculate the two players' per/60 when we don't know how much ES (or total) ice-time each player had??
So, first of all, forget 'adjusting' anything and let's just look at how each player finished in ES points-per-game (again, these rankings are only of the top-10 overall ES scorers):
Gretzky -- 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1
Lemieux -- 1, 1, 1, 2, 2, 2, 2, 3, 5

So, Gretzky has 7 of the top 8 (88%), and 12 of the top 14 (86%).

But how about we look at the only four seasons when the two players' primes directly overlapped---namely, 1987-88 to 1990-91 (for most of which, Gretzky was in Los Angeles):
1.33 -- Gretzky
1.17 -- Lemieux

But... but... but... you will argue, this thread is about peak, not about four seasons!

Right you are! So, whose peak is more impressive?: The guy who averaged 1.71 ES points-per-game for 7 consecutive seasons (while appearing in 96% of his club's games), or the guy who averaged 1.60 ES points-per-game for sixty games?

I mean, this is silly.

If you just want to go by raw totals you already have your answer. I do not see raw totals alone as presenting the 'truth' of a matter. It's as simple as that.

Most of you points are about career metrics. If we talk about their careers as a whole I agree, absolutely there's no question that Gretzky on average performed better at even strength than Lemieux did. No one has the kind of peak duration that Gretzky had, that's not what's in dispute here.

Wait, what???

No, no it isn't clear. In fact, the exact opposite of what you're (obsessively) arguing is 100% clear.

At this point, I can't remember if you detailed your mathematical determination of "adjusting" ES points, but as you surely know any particular method is going to have its flaws. We can't depend on one system to give us the be-all / end-all "adjusted" points. (I might add -- do we really need to adjust points when the two players' careers overlapped??) Then, how in the world can you mathematically calculate the two players' per/60 when we don't know how much ES (or total) ice-time each player had??

None of their absolute peak seasons overlap, there in lies the problem. Keep in mind some argue that they did in fact overlap in 1988-89(or even 87-88), the latter season one in which Lemieux was clearly ahead of Gretzky, laughably bad Hart voting notwithstanding, but for the record I'm not one of them.

The only seasons that matter(imo) for this precise debate are these;

Top raw ES/game seasons:
1.84 - Gretzky
1.83 - Gretzky
1.82 - Gretzky
1.79 - Gretzky
1.65 - Gretzky
1.60 - Lemieux
1.57 - Gretzky
1.34 - Lemieux
1.04 - Lemieux

We DO know something important about those seasons - the overall league-wide even strength scoring levels.

Top raw ES/game seasons:
2.98 / 1.84 - Gretzky
2.82 / 1.83 - Gretzky
2.88 / 1.82 - Gretzky
2.81 / 1.79 - Gretzky
2.86 / 1.65 - Gretzky
2.44 / 1.60 - Lemieux
2.64 / 1.57 - Gretzky
2.53 / 1.34 - Lemieux
2.10 / 1.04 - Lemieux

They were all notably higher during in Gretzky seasons. If Player A's season(s) features more even strength scoring than player B's season(s) all else being equal player A is going to accumulate a higher raw total of ES points. I would agree that all else is not entirely equal as there are some other factors to account for, but I don't see how anyone can disagree that this is not the biggest single factor which effects how many points a player will accumulate at even strength in a give season. We have almost 90 years of data which proves that point.

Lets forget about what the exact adjustments come out to, it's already clear from the above that Lemieux has at least one season that is at the same exceptionally high level as any of Gretzky's top seasons. In truth he really has two.

Then, how in the world can you mathematically calculate the two players' per/60 when we don't know how much ES (or total) ice-time each player had??

Sure, we can never accurately calculate such a thing. But it's easy to hide behind a veil of "we don't know for sure, so there's no point in discussing it." I simply choose not to pursue the easy path.

When you compensate for factors which are obvious it becomes clear that you are greatly underestimating Lemieux's performance at even strength, specifically in 1988-89. What other obvious factor? Ice time.

We know for a fact that seasons in which players have more powerplay ice time do not play more overall. There of course can be individual variances but the correlation is very weak at best at the individual level. My opinion? I'm sure Lemieux logged more total minutes in 88-89 than he ever had in any year prior to that. But that is in comparison with himself, not Gretzky. There's no indication anywhere that his overall time on ice was clearly more than Gretzky's during his best season(s). The only thing of note that I ever recall hearing mentioned about their ice times is an acknowledgment by Gretzky that both he and Lemieux logged more overall ice time than the other players in the league.

It's also been clearly established that players who are on the ice for a larger number of powerplay goals for and against have correspondingly high powerplay and shorthand minutes. I mean obviously that's common sense and not something anyone needed to research to prove. What is useful is an understand of how closely those percentages align. And they trend to align very closely for players who are on the ice for a majority of the powerplay goals a team scores(80-90% or more). Overall their time on ice figures are slightly lower than what you would expect for how many goals they were on the ice for which is understandable. Because the players who are on the ice for a high number of powerplay goals are the most skilled players on a team and the ones who replace them for the remainder of that time are less skilled. Penalty killing times are more difficult to precisely pin down, largely because no one ever plays the PK so much that they one the ice for an extremely high percentage of their teams powerplay goals against, around 60% being the upper limit. But you can certainly make some common sense deductions from the numbers.

From all that you can extrapolate what a teams PP & PP % could have been while a player was on the ice. You can makes arguments that a players actual powerplay and shorthanded ice times differed greater from what they would have expected to be but after a certain point the numbers make no logical sense. For example, for Lemieux to only be playing 4 minutes of PK time in 88-89 it would mean that the team was allowing more than twice as many powerplay goals against while he was on the ice than then when he was off it(his SH scoring doesn't come anywhere near to making up for that kind of deficit) and additionally that they had the best PK unit in the league when he was off the ice. I'm not going to discuss that matter anymore here, I'll leave it to others to draw their conclusions from.

We know for a fact that Lemieux was one the ice for far more powerplay goals for and against in his seasons, especially in 1988-89;

​	Season​	Team totals PP for & against​	Player on ice for(both)​	Player per gm​	Est. PP+PK TOI mins per game​
Gretzky​	81-82​	155​	89​	1.11​	7.1​
Gretzky​	82-83​	175​	112​	1.40​	9.0​
Gretzky​	83-84​	144​	99​	1.34​	8.6​
Gretzky​	84-85​	150​	98​	1.23​	7.8​
Gretzky​	85-86​	155​	111​	1.39​	8.9​
Gretzky​	86-87​	142​	91​	1.15​	7.4​
​	​	​	​	​	​
Lemieux​	88-89​	216​	170​	2.24​	14.3​
Lemieux​	92-93​	146​	105​	1.75​	11.2​
Lemieux​	95-96​	175​	130​	1.86​	11.9​

I mean it's double the number Gretzky had in 81-82. By all means don't take my Estimated PP & PK TOI mins per game accurately. It could be off by a minute or so either way, but go any further than that and you start introducing bizarre possibilities like that one I posted about Lemieux. Point being the exact numbers don't matter. But it should be clear what the numbers we have mean.

Let me ask you a simple question;
How could Lemieux possible have had anywhere near as much even strength ice time in 1988-89 as Gretzky had in ANY of his of best even strength seasons?

He obviously couldn't have and it would've been a difference of several minutes. Clearly your metric, of continually referencing raw even strength totals is too simplistic a measure to use to compare each player as it does not account for even strength scoring levels or even strength TOI. These are fundamentally factors which have a significant impact on a players raw totals.

Fact is, Lemieux put up 102 ES points in a season where ES scoring levels were lower(we know exactly how much) and where he played significantly less ES TOI(we don't know exactly how much). When you account for the combined deficit in both it places this season(along with his 92-93 season) at the same level as Gretzky's best ES seasons.

 

[The Panther]

If you just want to go by raw totals you already have your answer. I do not see raw totals alone as presenting the 'truth' of a matter. It's as simple as that.

Most of you points are about career metrics. If we talk about their careers as a whole I agree, absolutely there's no question that Gretzky on average performed better at even strength than Lemieux did. No one has the kind of peak duration that Gretzky had, that's not what's in dispute here.



None of their absolute peak seasons overlap, there in lies the problem. Keep in mind some argue that they did in fact overlap in 1988-89(or even 87-88), the latter season one in which Lemieux was clearly ahead of Gretzky, laughably bad Hart voting notwithstanding, but for the record I'm not one of them.

The only seasons that matter(imo) for this precise debate are these;



We DO know something important about those seasons - the overall league-wide even strength scoring levels.



They were all notably higher during in Gretzky seasons. If Player A's season(s) features more even strength scoring than player B's season(s) all else being equal player A is going to accumulate a higher raw total of ES points. I would agree that all else is not entirely equal as there are some other factors to account for, but I don't see how anyone can disagree that this is not the biggest single factor which effects how many points a player will accumulate at even strength in a give season. We have almost 90 years of data which proves that point.

Lets forget about what the exact adjustments come out to, it's already clear from the above that Lemieux has at least one season that is at the same exceptionally high level as any of Gretzky's top seasons. In truth he really has two.


Sure, we can never accurately calculate such a thing. But it's easy to hide behind a veil of "we don't know for sure, so there's no point in discussing it." I simply choose not to pursue the easy path.

When you compensate for factors which are obvious it becomes clear that you are greatly underestimating Lemieux's performance at even strength, specifically in 1988-89. What other obvious factor? Ice time.

We know for a fact that seasons in which players have more powerplay ice time do not play more overall. There of course can be individual variances but the correlation is very weak at best at the individual level. My opinion? I'm sure Lemieux logged more total minutes in 88-89 than he ever had in any year prior to that. But that is in comparison with himself, not Gretzky. There's no indication anywhere that his overall time on ice was clearly more than Gretzky's during his best season(s). The only thing of note that I ever recall hearing mentioned about their ice times is an acknowledgment by Gretzky that both he and Lemieux logged more overall ice time than the other players in the league.

It's also been clearly established that players who are on the ice for a larger number of powerplay goals for and against have correspondingly high powerplay and shorthand minutes. I mean obviously that's common sense and not something anyone needed to research to prove. What is useful is an understand of how closely those percentages align. And they trend to align very closely for players who are on the ice for a majority of the powerplay goals a team scores(80-90% or more). Overall their time on ice figures are slightly lower than what you would expect for how many goals they were on the ice for which is understandable. Because the players who are on the ice for a high number of powerplay goals are the most skilled players on a team and the ones who replace them for the remainder of that time are less skilled. Penalty killing times are more difficult to precisely pin down, largely because no one ever plays the PK so much that they one the ice for an extremely high percentage of their teams powerplay goals against, around 60% being the upper limit. But you can certainly make some common sense deductions from the numbers.

From all that you can extrapolate what a teams PP & PP % could have been while a player was on the ice. You can makes arguments that a players actual powerplay and shorthanded ice times differed greater from what they would have expected to be but after a certain point the numbers make no logical sense. For example, for Lemieux to only be playing 4 minutes of PK time in 88-89 it would mean that the team was allowing more than twice as many powerplay goals against while he was on the ice than then when he was off it(his SH scoring doesn't come anywhere near to making up for that kind of deficit) and additionally that they had the best PK unit in the league when he was off the ice. I'm not going to discuss that matter anymore here, I'll leave it to others to draw their conclusions from.

We know for a fact that Lemieux was one the ice for far more powerplay goals for and against in his seasons, especially in 1988-89;

​	Season​	Team totals PP for & against​	Player on ice for(both)​	Player per gm​	Est. PP+PK TOI mins per game​
Gretzky​	81-82​	155​	89​	1.11​	7.1​
Gretzky​	82-83​	175​	112​	1.40​	9.0​
Gretzky​	83-84​	144​	99​	1.34​	8.6​
Gretzky​	84-85​	150​	98​	1.23​	7.8​
Gretzky​	85-86​	155​	111​	1.39​	8.9​
Gretzky​	86-87​	142​	91​	1.15​	7.4​
​	​	​	​	​	​
Lemieux​	88-89​	216​	170​	2.24​	14.3​
Lemieux​	92-93​	146​	105​	1.75​	11.2​
Lemieux​	95-96​	175​	130​	1.86​	11.9​

I mean it's double the number Gretzky had in 81-82. By all means don't take my Estimated PP & PK TOI mins per game accurately. It could be off by a minute or so either way, but go any further than that and you start introducing bizarre possibilities like that one I posted about Lemieux. Point being the exact numbers don't matter. But it should be clear what the numbers we have mean.

Let me ask you a simple question;
How could Lemieux possible have had anywhere near as much even strength ice time in 1988-89 as Gretzky had in ANY of his of best even strength seasons?

He obviously couldn't have and it would've been a difference of several minutes. Clearly your metric, of continually referencing raw even strength totals is too simplistic a measure to use to compare each player as it does not account for even strength scoring levels or even strength TOI. These are fundamentally factors which have a significant impact on a players raw totals.

Fact is, Lemieux put up 102 ES points in a season where ES scoring levels were lower(we know exactly how much) and where he played significantly less ES TOI(we don't know exactly how much). When you account for the combined deficit in both it places this season(along with his 92-93 season) at the same level as Gretzky's best ES seasons.

This is all very nice supposition based on your statistical guesses and unverified assumptions, but how do you account for the fact that, head-to-head, from Oct,. 1987 to April 1991, post-peak Gretzky handily out-produced peak Lemieux in ES points-per-game?

 

[I Hate Blake Coleman]

Gretzky knew the game better than everyone. 86 is THE Gretzky season in terms of player profile, 163 assists. The only player to ever score more than 163 points is Lemieux. Wild.

 

[TheStatican]

This is all very nice supposition based on your statistical guesses and unverified assumptions, but how do you account for the fact that, head-to-head, from Oct,. 1987 to April 1991, post-peak Gretzky handily out-produced peak Lemieux in ES points-per-game?

That's a wide stretch of games covering 4 seasons. When I said "Lemieux was just as good as an even strength scorer as Gretzky" I meant that in the context of this thread i.e. in a specific single season, not a stretch of seasons or their overall careers.