Assists are a subtler point that need to be considered. Not many people realize that during the 1950's, the NHL awarded fewer assists per goal compared to the modern NHL. Leaguewide, there were 1.66 assists per goal in 1996, but only 1.50 assists per goal in 1953. Applying this adjustment and Howe is up to 154 adjusted points (75 goals, 78 assists).
That's a good point which is rarely considered. I do wonder about one thing though, perhaps the total number of assists have also gone up because there was more passing going on in higher scoring era's vs the low scoring 50's or simply as a matter of the game becoming more creative. It would be interesting to see how many unassisted goals were scored in a give year in the 50's vs the 80's & 90's. But I suspect that 'creativity' alone likely wouldn't account for that dramatic of a difference in assists per goal.
Another subtle point is the split between ES, PP and SH scoring. Surprisingly, 1996 only featured about 13% more goals per game at ES compared to 1953. But there was a massive difference in scoring on the powerplay (it was 90% higher in 1996 compared to 1953). Unfortunately we don't have enough data to tell if more powerplays were being called in 1996, if powerplays were more successful (ie a higher percentage resulted in a goal), or some combination of both. But for Lemieux, a player who by 1996 got a significant portion of his points on the powerplay, this is a huge advantage. He wouldn't have come close to putting up the same scoring totals in 1953 because PP goals were much less common back then. (Shorthanded offense was also far more common in 1996 - 121% more goals per game compared to 1953 - but since this is a relatively small part of offense, it doesn't have a huge impact).
Here's my best estimate of what Lemieux's production would have looked like, scaling everything back to 1953:
| SEASON | ES G | ES A | PP G | PP A | SH G | SH A | Goals | Assists | Points |
| Howe (1953) | 37 | 30 | 11 | 16 | 1 | 0 | 49 | 46 | 95 |
| Lemieux (1996) | 23 | 30 | 14 | 19 | 3 | 0 | 40 | 49 | 89 |
Of course I'm not saying that Lemieux would have scored exactly 89 points (no more, no less) in 1953. But it's a ballpark estimate that helps adjust for context (just like we need to do if we're comparing dollars to Euros, or prices from today to prices from 30 years ago).
The numbers seem a bit off are you making adjustments for some other factor(s)? Using separate even & powerplay scoring rates rather than the difference in overall scoring does drop Lemieux's numbers drop a fair bit bit mor but not as low as your numbers are.
Even Strength points > 73/2.10*1.85=64/65
Powerplay points > 79/0.90*0.48=42
Which gives us;
| SEASON | | ES G | ES A | PP G | PP A | SH G | SH A | Goals | Assists | Points |
| Howe (1953) | | 37 | 30 | 11 | 16 | 1 | 0 | 49 | 46 | 95 |
| Lemieux (1996) | Separate ev & pp rates | 26 | 37 | 16 | 26 | 3 | 0 | 45 | 63 | 108 |
| Lemieux (1996) | gpg rate only | 23 | 33 | 24 | 37 | 6 | 0 | 53 | 70 | 123 |
Are your also assuming Lemieux misses the same % of games(70/82=0.85*70=60)?
Under that scenario Howe does indeed best Lemieux's total's but that's a rather important distinction to make;
Lemieux 39-54-93 in 60 games
Howe 49-46-95 in 70 games
Still all these conversions don't take into consideration a couple of other important factors.
First off if a season is only 70 games long how many games would Lemieux actually skip? Taking games off wasn't solely about managing fatigue during portions of a season it was just as much about managing season-long fatigue and it wasn't automatically a given that Lemieux would skip back-to-backs like everyone seems to think it was; he played in
7 of the team's 16(44%) back-to-backs in 1995-96. He also played 20 games straight at one point, scoring 46 points in that stretch which is exactly equal to his seasonal average. Furthermore in his 00-01 comeback season he skipped just 2 of 8 back to back games and one of those was the last game of the season. What I'm getting at is the shorter the season the less games he seems to misses on a percentage basis(15% in 95-96 verse 6.5% in 00-01). Presuming his health issues are the same the number of games he'd miss in a 70-game season is likely to be in the range of 5 to 8 rather than 10 or 12.
Secondly there is another even more important factor that we must consider.
Lemieux's adjusted totals for powerplay scoring under-estimates his even strength numbers because it uses Lemieux's actual even strength production in 95-96. Using his actual 95-96 production is flawed because if powerplay opportunities are dramatically lower in 52-53, which they almost assuredly were, then that means Lemieux is playing less powerplay and less short-handed minutes. What happens to those minutes? Lemieux would not just be sitting on the bench waiting for the next powerplay or shorthanded situation to happen.
No, he would be playing more minutes at even strength, meaning his even strength numbers would be higher.
Would he put up as many even strength points as powerplay points? Absolutely not. Players generally produce points at double the rate on the powerplay vs even strength and Lemieux's ratio in 2000-01(the closest season of his we have numbers for) was pretty close to this; 3.67 vs 6.98 = 1.90 But players also score at a much higher rate at even strength vs being shorthanded. His ev point rate in 00-01 was more than triple his shorthanded per 60 rate; 3.67 to 1.12 = 3.28 Though as I found out later, this ratio is much lower for superstar players who have exceptional shorthand scoring seasons. Nonetheless it's the change in the minutes played that matter most.
To figure out what adjustments we'd need to make we would need to know Lemieux's time on ice in each situation. Unfortunately the NHL didn't track TOI numbers until 97-98 so we don't know his exact numbers for 95-96. Fortunately we
do have data from a season of his which is similar and I believe we can make some reasonable approximations based on his numbers in 00-01, from using some lesser know stats like PGA and from going through the teams 95-96 game logs. Not wanted to bog things down in here with an overly long post I started a sperate thread on that matter here;
This thread is an addendum to this post; https://forums.hfboards.com/threads/was-lemieux-still-in-his-peak-in-1995-96.2906900/post-187317395 Using his TOI stats from 00-01 will help to figure out Lemieux's 95-96 powerplay TOI. Here's an analysis of the Penguins powerplay for each season. I...
forums.hfboards.com
Naturally even strength play is more tiring than time on the powerplay so the loss of one powerplay minute does not equate to the gain of one minute at even strength. Assuming Lemieux rests for a couple shifts after playing the entire powerplay which he almost always did in 95-96(as I discussed in that separate thread), that gives us 120 mins of PP TOI for every 240 mins of game time. While on even-strength he would be able to play a shift and then skip 3 comparatively shorter shifts by the other lines, except in certain high leverage situations, which generally also gives us 60 mins of EV TOI for every 240 mins of game time. So the conversation rate would be two minutes of powerplay time = one min of even strength time. Shorthanded minutes are somewhat more taxing but for the sake of simplicity we'll use the same 2 to 1 conversation rate which I think is fair when you consider those high leverage game situations where Lemieux would only skip 2 shifts vs 3.
I'm not going to spend as much time trying to estimate his shorthanded TOI since its much less likely to be accurate. Unlike his incredible powerplay scoring prowess Mario was never considered a Selke candidate defensively though I would say he wasn't a bad defender on the penalty kill. Certainly his reach and size alone would have made him more difficult to play against than the average player but his purpose out there was almost as much to look for openings to score as it was to defend. His NHL record 13 shorthanded goals in 1988-89 can attest to that. Lemieux played a significant role on the Penguins penalty killing unit in 1995-96, until at least about
two-thirds of the way into the season when back issues started to creep up again. The only absolute figures we have about his shorthanded play for 95-96 are his powerplay goals against(PGA) which was 28(0.4 per gm) and his shorthanded point totals which was a league leading 9. It's a small sample size but for what it's worth he went from 7 shorthanded points in 48 games(0.15 per gm) to 2 in the final 22 games (0.09 per gm).
Interestingly enough he was averaging about 2.50 ppg prior to the back issues flaring up and after that he was under 2 ppg for the rest of the season which leads me to believe that is another factor for his production decline in the second half aside from losing his most skilled line mate in Thomas Sandstrom and possibly general fatigue as well.
In 2000-01 he only notched a single shorthanded point in 43 games(0.02 per game) and his PGA total was 13(0.3 per game). Which leads me to think that perhaps it would be better to compare his SH numbers not to his 00-01 but his closest contemporaries. Here's a list of of superstar or highly skilled players who have scored at least 8 shorthanded points since the NHL started tracking TOI. The PPO figures for Alfredsson & Fleury are averages, I'm not about to go through their team logs to figure that out but their Team PGA(powerplay goals against) is accurate for their actually games played.
Age | Player | season | Gm | Totals G/PT | | | Opp PP% | Opp PPO/Gm | Opp PPO | Team PGA | PGA | SH G/Pt's | PGA/Gm | SH Pt's per 60 | EV Pt's per 60 | EV ppg | SH TOI Adjusted for 5.13PPO & 20.3% |
30 | Lemieux | 95-96 | 70 | 69/161 | 25:24 | 3:15 | 17.8% | 5.86 | 410 | 73 | 28 | 8/9 | 0.40 | 2.37 | 4.35 | 1.04 | 3:15 |
35 | Lemieux | 00-01 | 43 | 35/76 | 23:59 | 1:15 | 21.2% | 4.72 | 203 | 43 | 13 | 1/1 | 0.30 | 1.12 | 3.67 | 1.00 | 1:51 |
26 | Bure | 97-98 | 82 | 51/90 | 22:14 | 3:01 | 20.6% | 4.56 | 374 | 77 | 34 | 6/9 | 0.41 | 2.18 | 2.79 | 0.68 | 4:29 |
28 | St. Louis | 03-04 | 82 | 38/94 | 20:34 | 1:36 | 15.1% | 3.39 | 278 | 42 | 14 | 8/11 | 0.17 | 5.03 | 2.60 | 0.65 | 2:21 |
31 | St. Louis | 06-07 | 82 | 43/102 | 24:09 | 1:43 | 21.6% | 3.72 | 305 | 66 | 13 | 5/11 | 0.16 | 4.69 | 2.54 | 0.74 | 3:17 |
26 | Lecavalier | 06-07 | 82 | 52/108 | 22:36 | 1:31 | 18.4% | 4.56 | 374 | 69 | 14 | 5/9 | 0.17 | 4.34 | 2.82 | 0.77 | 2:01 |
35 | Alfredsson | 07-08 | 70 | 40/89 | 22:17 | 2:48 | 18.9% | 4.63 | 324.4 | 62 | 22 | 7/9 | 0.31 | 2.76 | 3.11 | 0.79 | 3:48 |
27 | Nolan | 99-00 | 78 | 44/84 | 20:55 | 2:31 | 16.6% | 4.56 | 356 | 59 | 25 | 4/8 | 0.32 | 2.44 | 2.41 | 0.56 | 3:01 |
26 | Hossa | 05-06 | 82 | 39/92 | 21:41 | 2:36 | 20.8% | 5.85 | 480 | 100 | 18 | 7/8 | 0.22 | 2.25 | 2.60 | 0.55 | 3:02 |
32 | Fleury | 00-01 | 62 | 30/74 | 21:47 | 2:16 | 21.5% | 4.88 | 302.4 | 65 | 26 | 7/9 | 0.42 | 3.42 | 2.44 | 0.55 | 3:05 |
26 | Sullivan | 00-01 | 81 | 34/75 | 20:32 | 1:55 | 15.9% | 3.95 | 320 | 51 | 18 | 8/10 | 0.22 | 3.86 | 2.49 | 0.60 | 2:32 |
| Average | | | | | 2:07 | | | | | | 8.5 | | | | | 2:57 |
One thing that jumps out is St. Louis's unbelievable SH per 60, it was actually higher than both his ev and powerplay per 60! in 06-07. In fact 6 of the 9 had a higher per/60 shorthanded than at even strength. But none of these players are close to Lemieux offensively. The average SH TOI for all these players, not including Lemieux 00-01 season where it was clear he was playing less SH minutes, is 2:07 The adjusted average, to 95-96 PPO's level and a slightly better penalty kill %, which would means more minutes playing shorthanded, is 2:22
But there are two players who's number most closely corelate to Lemieux's. Goals scored against per game on the penalty kill is one metric to look at. But this figure doesn't tell the whole story as other factors like having poor quality penalty killing teammates might be inflating this number. That's where the penalty killing percentage(inverted to Opponents powerplay % on the list above). If the penalty kill % and the number of powerplay goals against in addition to their shorthand scoring are all in line with each other I would say Lemieux TOI figures would be more stronger correlation to those particular players
and in all those instances Bure & Fleury's seasons are the closest. And for that reason I'll be using their numbers for Lemieux's estimate. The average SH TOI between the two is 2:39 while adjusted it's 2:54. But that last number seems a little high so I'll stick with the unadjusted figured.
Edit* I mistakenly used the Penguins PPO(power play opportunities) number which was 359 instead of his opponents PPO number which was much higher at 410. In the games he played the Pens were short handed 5.86 times per game! Curiously enough in the games he didn't play that number drops dramatically to 4.75 It's hard to understand why the Pens would be called for more than 1 penalty less per game when Mario wasn't playing, that's a
23% difference! It's not like Mario himself was the cause for this huge spike, he had 54PIM that year and was called for 27 minor penalties in 70 games and 6 of these were offsetting minors which leaves us with 21 PPO's against or exactly 0.3 per game. Who knows how many were called against whomever took his place in the lineup when he wasn't playing too. But even if that player didn't get called for a single penalty there's still a difference of 0.81 per game to account for, an increase of 17%. Either the Penguins played a much more physical game with Mario in the lineup, which clearly was not the case
or something else was going on... I think the evidence clearly speaks for itself, this was a case of referee's seeking to equalize the playing field and reduce the Penguin's massive advantage on the powerplay by calling them for more calls overall.
| | For | | | | | | Against | | | | |
| | PPG's | PPO | % | per/Gm | PP TOI | Avg length | PPG's | PPO | % | per/Gm | SH TOI |
With | 70 | 102 | 359 | 28.4% | 5.13 | 08:01:00 | 93.7s/1:34 | 73 | 410 | 17.8% | 5.86 | ? |
Without | 12 | 7 | 61 | 11.5% | 5.08 | 08:55:00 | 108.8s/1:49 | 5 | 57 | 8.8% | 4.75 | ? |
The jury find the referee's of the 1995-96 season
Guilty of the Charge of
referring to the score.(All the above stats and how they were derived can be found in the
Lemieux's 1995-96 powerplay stats and figures thread)
Getting back to the point, the Penguins clearly played a lot of minutes shorthanded and my original estimate of 2:39 SH TOI per game for Mario was an underestimate and we really don't even need to look at his peers for the answer as there's actually a pretty straight forward way to deduce a reasonable estimate of what his SH TOI was. Powerplay time per powerplay is lower for teams which have better powerplays because they score more often during them and once they score the powerplay is of course over. We see this clearly in the Penguins numbers listed above, those numbers also took into consideration calls against the Penguins while they were on the powerplay which is the other means by which powerplays end early. Shorthanded the Penguins penalty-kill operated at a rate in-between the Penguins PP rate with and without Lemieux so the average length of their powerplay would also likely be in-between the two. Somewhere within a few seconds of 100 seconds would be a safe assumption, though even that number may be high; Some
figures calculated in this article gives us totals within a few seconds of 100s for powerplays operating around 20% efficiency. The Penguins opponents were operating at 17.8% with him in the lineup which was essentially the same as the league average that season. There's absolutely no chance that they would have been anywhere near the lower figure of 93.7 seconds like the penguins supremely proficient powerplay was. Now we multiply that figure by the total number of opportunities - 5.86 per game. This gives us a total of 9:46 minutes per game. This is where
the PGA comes to play. No, not
that PGA, I'm talking about powerplay goals against. Lemieux was on the ice for 28 of the 73 goals powerplay goals scored against the team, or rather 38.4% See what I'm getting at here? 38.4% of 9:46 is 3:45, which is significantly more than my earlier estimate of 2:39. If Lemieux was better than the average Penguins penalty killer that year than he would have spent even more time on the ice shorthanded as goals would be scored at a lower rate with him out there. If he was a worse penalty killer on average than his short handed time would be less as goals would be scored at a higher rate against him. My guess is he probably was a little worse than average since he was focusing more on offensive, which mind you did pay off considering he had 9 SH points - For every 3 goals teams scored against him on the powerplay he basically got one back, not a bad trade off imo. With all that in mind I'm going to readjust his totals yet again and go with a figure of him being out on the ice 33% of SH time rather than 38.4% meaning goals scored against his team while he was on the ice was at least 16% higher which shorthanded. This gives us a SH TOI of 3:15 (33% of 9:46)
First one more look at his actual TOI figures from 2000-01;
| Total | EV | PP | SH |
| 2000-01 TOI | 23:59 | 16:20 | 6:23 | 1:15 |
| Points | 76 | 43 | 32 | 1 |
| PPG | 1.77 | 1.00 | 0.74 | 0.02 |
| Points Per 60 | 4.42 | 3.67 | 6.98 | 1.12 |
| League Averages | 2.76 | 1.89 | 0.76 | 0.11 |
| P60 adjusted for 95-96 gpg | 4.93 | 4.08 | 8.27 | 1.43 |
League averages are goals per game figures not points per 60.
And now with everything taken into consideration these are the TOI numbers I have estimated Lemieux to have played in 1995-96;
| Total | EV | PP | SH |
| 1995-96 TOI estimates | 25:24 | 14:24 | 7:45 | 3:15 |
| Points | 161 | 73 | 79 | 9 |
| Points per 60 | 5.34 | 4.35 | 8.74 | 2.37 |
I have him as playing as extra 1:23 on the powerplay in 95-96 which is a very safe assumption based on
my extensive research and an extra 2:00 minutes shorthanded based on the calculations I made above. Meanwhile his EV time is down 1:56 His EV PPG was very close for both years 1.04 vs 1.00 only a 4% difference but I doubt a 35 year old Mario was an equal EV point producer to a 30 year old Mario in a lower scoring league; ev scoring was much less in 00-01; 2.10 vs 1.89 a 11% difference, it only makes sense that he played less time at ev strength in 95-96. Adjusted for scoring rates his 00-01 numbers are actually very close to my estimated 95-96 per 60 figures for him with the sole exception of shorthanded scoring. If you increase his even strength time to anything above 15:21 for 95-96(making his points per 60 less than 4.08 the scoring level adjusted figure for 00-01) than you're essentially saying that Mario was a better EV score at age 35 than at age 30.
Regardless, whatever number you choose to estimate for his actual EV TOI in 95-96 it makes no difference to the EV adjustment. Whether you have him as playing 12 minutes or 18 minutes, his EV time would increase beyond that because of the decreased special team time. The only thing you can argue about is whether that 2 minutes of special team time to 1 minute of EV time is an appropriate ratio.
Some may say it's unlikely Lemieux could have played over 25 minutes a game as forwards typically only average in the low 20's now. But as any hockey historian will tell you TOI figures have been consistently dropping for the last several decades, though the seem to have finally stabilized now. It was hardly unusual for a star forward to average 25 minutes per game in the 80's or early 90's and while that may no longer have been the case by the mid 90's there were clearly some exceptions dropping like Kariya in 98-99 and if anyone else could do it player like Gretzky and Lemieux would be at the top of that list. Plus much of this estimated time is spent playing less taxing special team minutes. Hey, if
35 year old Mario can play 24 minutes a game I don't think it's too much of a stretch to think the 95-96 version of him could have played about minute and half more overall while spending two minutes less at even strength.
So now that we have a rough estimate of his numbers we can do a more accurate conversation to 1952-53 scoring levels. Unfortunately we're also lacking a lot of data on that end considering we don't know how many powerplay opportunities teams had in 52-53. But there's obviously no way powerplay opportunities would be the same as 1995-96 which featured the
forth highest rate in NHL history. If it actually did that would mean powerplay efficiency would have been about half of what it was in 95-96. The lowest powerplay efficiency in history since the NHL began recording in 63-64 was 15.08% in 97-98. Though it is true that powerplay efficiency in the NHL's early days was thrend much lower than average. The second lowest figure on record was in 63-64, the first season for which the NHL started tracking powerplay data with a rate of 15.75%. But the powerplay efficiency in 95-96 was only modestly better than this at 17.93% Meanwhile the number of pp opportunities per team in 63-64 was 3.64 Since the average number of powerplay goals was only a little higher in 63-64 vs 52-53 (0.57 to 0.48) it's hard to believe that the efficiency rate would've been significantly different. If the number of opportunities stayed the same the efficiency rate would've been 13.26% by far the lowest in history and if the efficiency rate was the same the number of opportunities would've been 3.07 - there are several seasons with figures even lower than this. Most likely it was a combination of the two. I'm going to go with a nice round estimate of 3.5 PPO's per game which gives us an efficiency rate of 15.14% largely because powerplay efficiency has a much lower divergence from the mean compared to the average number of PPO's per game. A 3.5 PPO per game rate represents a 46.6% reduction from 95-96 and that's the percentage I used to reduce his PP and SH TOI numbers.
As a result this is Lemieux's TOI and production adjusted when using a 2 to 1 PP/SH to EV TOI conversation;
| Total | EV | PP | SH |
| 1995-96 TOI estimates | 25:24 | 14:24 | 7:45 | 3:15 |
| Totals | 69-91-161 | 30-43-73 | 31-48-79 | 8-1-9 |
| TOI net decrease/increase | -1:45 (-7.7%) | +1:45 (+12.1%) | -2:28 | -1:02 |
| 1952-53 adjusted TOI = | 23:39 | 16:09 | 5:17 | 2:13 |
| EV PT's after TOI adjustment | | 34-48-81.8 | | |
The only relevant number here is knowing the increased EV strength points totals. Why not calculate the adjusted PP & SH numbers here as well? There's no point in doing it in the above step as the decreased PP and SH totals will be adjusted based on 1952-53 PP & SH scoring levels vs 1995-96 levels(0.90 vs 0.48). The above decreased PP & SH time is indeed the primary reason why these totals were slow in the 50's and other early era's of the NHL. The other importantly factor is of course decreased powerplay efficiency but it accounts for less of the decrease than lower PP opportunities do i.e. lower PP TOI totals. PP goals per game takes both these contributing factors into play.
Now we can adjust for the differing scoring levels at EV, PP & SH in each season;
EV points > 82/2.10*1.85=73
PP points > 79/0.90*0.48=42
SH points > 9/0.14*0.05 =3
Now we can finally have a true adjustment in scoring rates;
| SEASON | | ES G | ES A | PP G | PP A | SH G | SH A | Goals | Assists | Points |
| Howe (1953) | | 37 | 30 | 11 | 16 | 1 | 0 | 49 | 46 | 95 |
| Lemieux (1996) | EV TOI adjusted + Separate EV, SH & PP scoring rate adjustments, 70 games | 30 | 42 | 16 | 26 | 3 | 0 | 49 | 68 | 117 |
| Lemieux (1996) | Separate EV, SH & PP scoring rate adjustments | 26 | 37 | 16 | 26 | 3 | 0 | 45 | 63 | 108 |
| Lemieux (1996) | gpg rate only | 23 | 33 | 24 | 37 | 6 | 0 | 53 | 70 | 123 |
If we reduce Lemieux's true converted numbers to xx number of games played he would still win the scoring title as we see below;
64 games 45-62-107
60 games 42-58-100
58 games 40-56-96
Imo all this makes perfect sense if you take into consideration the level of competition. They may have had comparable VsX PPG numbers but Mario was competing against 75 first line players while Howe was competing against 15. It's no longer a metric that makes sense at that point. The talent pool was simply much larger in 95-96 vs 52-53 because the population pool to draw from was that much larger. Canada's population had doubled from 10 million to 20 million from the time of Howe's birth to Lemieux's and that's before we take into account whether or not there was greater interest in hockey in the 70's/80's vs the 30's/40's to groom that talent(there most certainly was!) AND all the American and European talent that would enter the league by 95-96, compared to basically none in 52-53.
One other factor that I haven't accounted for is that shift lengths, which were far longer in the 50's. In the 90's superstars played minute-long shifts but in the 50's the stars would play 2 minute shifts and be on the ice for half the game. Meaning Lemieux's ev strength ice time would likely be much higher. But as we all know now performance also degrades with increased ice time. Considering those are two opposing factors lets call it a wash, plus I wouldn't even know where to begin to make adjustments for that so I'll be skipping it thank you very much.
In conclusion I firmly believe that the above should clearly and definitively prove that 95-96 Lemieux would have out-pointed 52-53 Howe. All that said I respect that Howe brought a lot more to his game then just scoring prowess alone. Not to mention there's a good chance that he would've still won the 'rocket' while providing a level of physically few could match(this reminds me of the early career Ovechkin vs Crosby debates...) Considering that even for me it would be a coin toss as to whom I'd rather have for my team; '53 Howe or '96 Lemieux.
This post also serves to show the flaws of using any type of adjusted scoring metric when comparing seasons with vastly different scoring rates and TOI distributions.
www.nhl.com
