TheStatican
Registered User
- Mar 14, 2012
- 1,731
- 1,512
There continues to be a misunderstanding that Lemieux's performance in 1988-89 was largely powerplay driven, specifically due to his team receiving a high number of powerplay opportunities than the norm that season. However this assumption is a flawed one as it does not consider two important factors - Lemieux's even strength scoring rate relative to his powerplay scoring rate and the fact that Lemieux also played in a large number of low scoring situations as the teams primary PK'er. When all of those considerations are taken into account the benefit of playing in a high scoring powerplay season was no where near as dramatic as some imagine it to have been.
Here are the conclusions directly to the point:
1. The number of powerplay points he loses from his totals if his team had the league average number of powerplays is 12 (12.5 specifically).
2. If Lemieux's scoring was at the same level as an average star player who typically has a two to one powerplay to even strength scoring rate, he already replaces half of those points at even strength, meaning his point totals drop by 6 (6.25 specifically).
3. But if Lemieux only scored at a similar rate as other star players at even strength then he would've had to play 32 or more minutes a game in order to score 102 es points in 76 games because his special team ice time was so huge that season. Clearly his ES scoring rates were closer to his PP scoring rate than that 1:2 ratio, this should not be surprising considering he scored over 100 es points in a season with a very high number of powerplays. In such seasons the top stars are going to of course have less ES TOI. This means he makes up more than just half those lost PP points at even strength, closer to 8 or 9 which drops his overall point total by 3 or 4.
4. Lemieux also played a very high number of minutes on the penalty kill this season because the Penguins had to kill jests as many penalties as they received, which as noted above was much higher than on average. His scoring rate while shorthanded was much lower than his even strength strength scoring rate. Depending on what his PK ice time was he would've gained somewhere between 3 to 5 es points.
In summarization Lemieux's point totals would decline by a small amount if his team had received a normal amount of penalties and if we consider both penalties for and against his point totals would have more or less remained unchanged. If the argue your making is that Lemieux benefited from a higher than usual amount of penalties than it's disingenuous to not also consider that higher than usual number of penalties the team had to kill.
Here is the relevant data used to justify that conclusion;
-In 88-89 the Penguins scored 119 PPG in 486 powerplay opportunities(PPO's with an efficiency of 24.49%
-In Lemieux's 76 games the those totals were 115 PPG, 455 PPO's with an efficiency of 25.27%
-The league average for PPO's&PKO's was 403 with an efficiency of 20.99%
-This equals to 383 PPO's/PKO's in 76 games, which means the Pens had an additional 72 PPO's above the league average in Lemieux's games
-Lemieux had points in 79 of those 115 PP goals
Simply going by these numbers tells us Lemieux's PP totals would've been reduced by 12 or 13
79/455*383 = 66.5
79 minus 66.5 = 12.5
199 minus 12.5 equals 186/187 points.
This is far from the end of the story due to the following factors;
Factor A) If Lemieux wasn't playing on the powerplay do you think he would just be sitting on the bench during those lost minutes? Absolutely not, the best player in the world would have instead been on the ice playing those minutes at even strength which would in turn obviously increase his even strength totals. Ok, but this would only give him a few extra points right? No offense, but all those who believe that the difference is only 2 or 3 points are either being willfully ignorant or completely disingenuous. The only way this would be true is if Lemieux was a terrible even strength scorer and considering only Gretzky has ever registered a higher even strength point scoring rate in a season than Lemieux did in '89, clearly he was pretty good at this even strength scoring thing.
Factor B) If your going to penalize Lemieux for excessive powerplay opportunities then you had damn well better consider the excessive amount of shorthanded situations he played in as well. In Lemieux's 76 games the Pens were shorthand just as much as they were on the powerplay. And who was the teams primary penalty killer? Lemieux. With less SH time he naturally would've had more ES time and more ES points to go along with it. But he scored so many SH point that season, so surely it wouldn't have made a big different? That is yet another incorrect assumption. While he did score a high number of SH points his scoring rates at even strength were far higher than his shorthanded scoring rates.
Factor C) Lemieux's powerplay usage was very high in 88-89 but it wasn't as high as 95-96's because unlike 95-96 he wasn't out there for every single powerplay goal the team scored. In 88-89 the Penguins averaged 6 PPO's per game which is an huge amount of powerplay time. The 95-96 Penguins averaged about 1 PPO less per game. Being on the ice for all 6 PPO's per game is difficult to achieve, but lower that amount to 5 PPO's per game and it's more likely a that a player will be able to play in all of that powerplay ice time. If Mario's powerplay usage increases then he doesn't simply lose 20% of his powerplay totals (6.0 PPO's/ 5 PPO's). This factor is admittedly is not large and actually need not even be considered since Lemieux would have made up the difference through the above two factors alone. two factors discussed above.
To figure out how a good estimate of how many powerplay points he would replace with even strength points we must come up with an accurate estimation of what his scoring rates were for each situation. The first figure we need to determine is what his PP usage was. As mentioned above Lemieux was on the ice for 110 of the teams 115 Powerplay goals and actually he missed one additional game worth of ice time - missing most of the third in game 11 against the Rangers and leaving n the first in game 13 against the Nordiques. The team score 1 PP goal without him in those games so he was actually on the ice for 110 of the teams 114 powerplay goals when available. Missing 4 out of one 115 powerplay goals might sound like a trivial amount but I assure you it is not, because it tell us two things. One - Lemieux did not play in every single minute of powerplay time that the team had and two - using this figure we can make accurate estimations about his PP TOI and scoring rates.
First his usage. It's extremely unlikely that the Pens would have scored those 4 other PP goals(in the games he played in but was not on the ice for) in the same number of PPO's as it took them to score 4 PP goals with Lemieux. 4 goals at their normal efficiency would've required just 16 PPO's(25.0%). In the 4 games Lemieux missed the Pens scored just 4 PP goals in 31 PPO's, an efficiency of 12.9% Sure it's not a large small sample size but it is corroborated by them scoring 1 goal in 8 PPO's(12.5%) in those 3 other periods he missed that season and from data in other seasons. Not to mention that anyone with even a very basic understanding of the game would surmise that a team would perform substantially worse on a powerplay minus a generational player.
There is little doubt the Pens powerplay without him(in short spans) was atrocious. Yes their powerplay was much better than that in the seasons where he missed a lot of games(22.94 in 90-91 & 18.81 in 93-94 & 19.00 in 94-95) but that's entirely expected. It only makes sense that their PP unit operated better in those situations considering coaches will naturally make adjustments over the course of a season to figure out what works and what doesn't work. But when you lose the anchor of your powerplay just one or two games at a time or when he skips a powerplay in any individual game, it's much more difficult to adequately compensate for that loss as the above data proves.
Additionally the team did not have the same level of talented players in 87-88/88-89 to fill in Lemieux's void as they did in the 90-91, 93-94 or 94-95 seasons. Though having a talented rooster in 1995-96 yet that didn't prevent the team from posting a meager 11.5% efficiency during the 12 separate games Lemieux skipped. Therefore It's logical to assume the team would've had a similarly poor efficiency, if not worse, when he wasn't on the powerplay in the games he played in. In most situations Lemieux was probably out there the entire powerplay (120 seconds) without the team scoring, most powerplays are unsuccessful afterall. In many others he would've only been out there for whatever number of seconds it took before the team scored(about 25% of the time). But in a small but not insignificant number of them he would have left off the ice before the conclusion of the powerplay and another player would take his spot for mop up duties in the final 15, 20 or 30 seconds of an ongoing powerplay. This would naturally be a very unlikely situation for success. In this context it makes perfect sense that the team would have had a very low PP efficiency, probably even lower than that 12.9% figure they had in the games he didn't play in as at least in those games the team would have had more opportunities to make good zone entries and set the powerplay up for the entire length of the PP.
Here's an example of what I mean, taken from actual known TOI numbers;
In 2009-10 Steven Stamkos lead the league with 41 PP points and he was on the ice for 54 of the teams 63 PP goals = an involvement percentage of 85.7%. His PP TOI was 381:18 while the teams total PP ice time was 524:56 This means his usage was much lower at 72.6% The team had 325 PPO's that season(19.38% 3.96 per gm, 96.9s per PP). His PP ice time equates to 236 PPO's. Meaning that without him on the ice the team scored only 9 PP goals in 89 PPO's. That's an efficiency of 22.88% with him and only 7.87% without - The team was 2.9 times better with him than without on the man advantage.
At 12.9% we are surmising that the Penguins were only twice as good with Lemieux as they were without him. Imo this is a conservative estimate and it aids in creating an unlikely scenario in his stats - if the Penguins powerplay was not that bad with him then it means that his usage numbers are higher. The higher his usage numbers are, the more powerplay TOI he has. This has two effects one: the higher his PP TOI is the lower his PP/60 scoring rate is and two: if he played more minutes on the powerplay than that correspondingly decreases his ES minutes. Furthermore players do not get more overall TOI in high power play seasons;
forums.hfboards.com
Going back to the numbers,
-If we suppose that 12.8% usage figure, the team would've needed 31 PPO's worth of ice time to score those 4 other PP goals
-This gives an estimate of Lemieux having played in 416 of the 447 powerplay opportunities worth of ice time the team had when he was available, a usages figure of 93%
Obviously this doesn't mean Lemieux skipped 31 PPO's entirely but rather that the cumulative powerplay time he didn't play in was equal to 31 PPO's worth of powerplay ice time
-The Pens powerplay efficiency when he way playing was 25.5%
-The average length of a powerplay that operates at that efficiency would be on average about 95.5 seconds.
Higher PP efficiencies = shorter overall powerplay lengths and from that we can make a good estimate of what the average length of any teams powerplay was based on their powerplay efficiency. In 1995-96 the Penguins had an overall PP efficiency of 25.95% which equated to an average powerplay length of 95.4 seconds(info here). With a slightly lower efficiency, powerplay length would be slightly longer, hence the 95.5s figure I came to for the Pens in 88-89. However this figure can vary by about 3 - 13 teams have had similar PP efficiencies, the average of those 13 seasons was 96 seconds. Most of them are either from this season or the covid shortened season(s). The shorter the sample size the greater the variance. Of the full seasons the range was between 92.0s and 98.6s a little more than 3 seconds either way. The difference amounts to at most 3.5% aside from the Capitals of 20-21 which was an outliner at 5.5% But longer PP lengths means more PP minutes which if you've read through to this point actually means Lemieux makes up more ES points. In any case 3 to 4 % makes very little difference in the numbers. His PP usage has a much more drastic effect as it changes the numbers by 10% or more.
From all the above data we can approximate what Lemieux's powerplay per 60/scoring rate was;
95.5 seconds multiplied by 416 PPO's (93% PP usage) equals an estimate of 662.1 total powerplay minutes. Which would give us an estimated of 8:50 of Powerplay time per game that season. This tally would beat the highest known PP TOI figure of all time, since TOI tracking began in 1997-98 the record is 8:11 set by Ilya Kovalchuk in 2005-06. Lemieux surpassing this figure would be no small feat because 2005-06 featured the highest number of PPO's per game in NHL history; 16% more per game than 1988-89's average and 11 teams had more PPO's per game than the 88-89 Pens including the Thrashers who not only average more per game (6.42 vs 5.99) but they also had a much lower PP efficiency(18.98% vs 25.27%) which means a longer PP length on average. However Kovalchuk's usage was 79.6%(he played in 8:11 PP mins per game while the team had 10:17 of PP time per game) which was almost certainly a fair bit lower than Lemieux's since he was on the ice for 90.1% of his teams powerplay goals(82 of 91 in his 78gm that season) while Lemieux was on the ice for 96.5%(110 of 114).
Now on to Lemieux's SH ice time.
-The Penguins were shorthanded 482 times, 6.02 per game, allowing 111 goals an efficiency of 76.97% (23.03)
-In Lemieux's games they were shorthanded 456 times, 6.0 per game, allowing 101 goals an efficiency of 77.85% (22.15)
-The team allowed 10 PPGA in just 26 PKO's in games Lemieux didn't play, a horrible PK efficiency of just 61.54% (38.36)
-The league avg was 383 in 76 games, meaning the Pens had an additional 73 PKO's over the league average.
-Lemieux was on the ice for 60 of those 101 PPG against in the games he did play in = 59.4% and was the teams primary penalty killer as he was on the ice for more powerplay goals against then even the teams primary defenders were. Other well more known players like Brown, Cullen and Stevens played virtually zero PK time;
The Penguins penalty kill wasn't great overall that year but it was absolutely putrid without Lemieux as it allowed teams to score with a 38.4% PP efficiency. While undoubtedly Lemieux's main purpose on the PK was to counterattack and score I believe he was a better than average penalty killer - though obviously no Selke candidate by any means, but teams had to be more cautious and respectful of his ability to score shorthanded. Of course it's also possible that even if he was a better than average PK'er teams may have still scored at a higher rate with him on the ice due to factors outside his control like that quality of his fellow PK'ers or even random variances. Since I cannot prove any of the for fact I've made no adjustments for any of those considerations. His SH TOI estimates are based on the assumption that scoring rates for the opposition with him on the ice were the same as team average. Either way it is unquestionable that Lemieux played a HUGE roll on the Pens PK that season, in fact his 60 PPGA are tied for the NHL record for most powerplay goals against for forwards. The highest total Gretzky had was 43. If Lemieux was allowing powerplay goals to be scored against the team at an alarming rate I highly doubt the team would've had him out there to the detrimental of the team especially when he could have otherwise played more minutes at ES providing much more scoring in that situation.
-If we assume he was basically a team average PK'er that means his PK usage would've been 59.4%
-This means his PK minutes would be equal to 59.4% of the teams 456 PKO's, which equates to 270 PKO's
-The teams PK % was 77.85 when he played, inverted that's a powerplay of 22.15% for the opposition. This is lower than the team's powerplay and therefor means longer PK's on average. It's not too far off the Capitals 2009 to 2022 figure of 21.89% in the chart above which equates to a PP length of 97.58s Hence I'll use an estimate of 97.5 seconds for the average length of the 88-89 Penguins PK.
97.5 seconds multiplied by 270 PKO's (59.4% PP usage) equals 438.8 total Penalty Killing minutes. Per game that's 5:46 of PK time.
That leaves one question to be answered, what was his even strength and overall time on ice? Perhaps the most comprehensive analysis of past TOI figures ever provided was done by Iain Fyffe. He estimate Lemieux's ice time to be 26:00 per game in 1988-89; https://web.archive.org/web/20060526131616/http://www.puckerings.com/research/1989pa.html
With that estimate in place we can finally calculate his per 60 scoring rates for 1988-89;
The formula for Points per 60 is
Points divided by (TOI mins a gm multiple by total games) x 60
199 / (26:00 x 76) x 60
199 / (1976) x 60 = 6.04
His even strength and powerplay scoring rates are identical. Meaning if it those numbers are accurate he would entirely make up for a huge amount of that lost PP production at even strength. But I agree a scoring ratio of 1 to 1 seems unrealistic, however it HAS happened in the past - in 1997-98 Teemu Selanne scored at a higher rate at even strength than he did on the powerplay and others have come close as well.
Player season PP/60 to EV/60:
Selanne 97-98 2.97 to 3.20 ratio of 0.92 to 1
Crosby 09-10 5.12 to 4.20 ratio of 1.22 to 1
Bure in 99-00 3.71 to 3.38 ratio of 1.10 to 1
Jagr in 99-00 4.83 to 3.81 ratio of 1.27 to 1
Keep in mind that if you think his PP usage was EVEN HIGHER than 93% that means his powerplay minutes were HIGHER yet and his PP per/60 would then go lower while his EV per/60 would be even higher.
The total time on ice figure is probably low for Lemieux so lets give him 30 minutes a game in total, in his 75 full games - remember he missed 3 periods, pretty much a games worth of ice time in game 11 and 13. citations for that here and here. He also only played on the powerplay in his first game back after that second injury but things are complicated enough as is so I'm ignoring that in the calculations.
That gives us a per 60 PP/EV scoring rates of 7.16 to 5.30 and a much more realistic PP/EV scoring ratio of 1.35 to 1
If you look at the numbers of most superstar players in recent seasons you might think that's still too high of a ratio but none of those players were ever on Lemieux's level, none of them (aside from Jagr that one season) ever came close to 102 ES points and yet Selanne, Crosby, Bure and Jagr all managed to have even higher ES to PP ratios in what was arguably the peak season for all of them. I see little reason to believe that Lemieux wouldn't have been be able to score at those rates and that ratio in what was one of the greatest seasons in NHL history. If Lemieux's PP/EV ratio was 1.35 to 1 that means we just need to divide his lost Power play point totals by this number > 12.5 / 1.35 = 9.3
In conclusion, converting his lost powerplay production to even strength would result in a net loss of...
3 points!
That just covers factor A, now lets look at factor B; Less SH TOI = more EV TOI.
First off this is how many SH points Lemieux would lose when reducing his PKO's to the league average;
The formula is; SH points / team PKO's x league avg PKO's
18/456*383 = 15.1
18-15.1 = loss of 2.9 SH PT's
Here's how many EV points he would gain
EV per 60 = 6.52
SH per 60 = 2.46
Ratio = 2.65 to 1
2.65 x 2.9 = gain of 7.7 EV PT's
Difference; 7.7 -2.9 = net gain of 4.8 PT's
Combining factors A&B
-3+4.8 = +1.8
Normalizing Lemieux powerplay and shorthand opportunities to the league average would actually increase Lemieux's point totals.
I know exactly the response I'll get from Gretzky proponents who are the ones that continue to propitiate this entire argument; "That's ridiculous, Mario wouldn't gain points by playing in less powerplays"
Answer; No he wouldn't, he would gain points by playing in less powerplays AND shorthanded situations.
Let's breaking it down again;
FACT - The Penguins were involved in a extraordinary amount of PPO's AND PKO's that season.
FACT - Lemieux was on the ice for 95.6% of the teams powerplay goals for and 59.4% of the teams powerplay goals against.
FACT - Lemieux would play ES minutes in place of decreased PP minutes. There is zero or at most a very weak correlation between between increased powerplay time and increased overall TOI.
Estimate based on thousands of games worth of data - PP & PK lengths of 95.5 seconds(@25.27%) & 97.5 seconds (@22.15%)
Estimate - powerplay usage of 86%
Estimate - shorthanded usage of 59.4%
Estimate - I've recalculated everything above using a TOI figure of 30 minutes per game.
Considering we've established the baseline figures for his PP & SH TOI, if his total TOI minutes were lower than 26 minutes, personally I wouldn't suggest anything below 25, it means his even strength scoring rate was even higher. If his total TOI minutes were higher, also a possibility, then it drops his even strength scoring rates. But how much higher do you think they could possibly have been? Do really think Lemieux's was playing what... 28? 29? 30 minutes a game? Lemieux was a beast to be sure but there's NO WAY in h ell he was playing 30 freaking minutes a game - THIS IS NOT THE 50's PEOPLE. Also as mentioned three times before their is no correlation between increase powerplay time and increase overall TOI. If you think Lemieux was playing 30 freaking minutes a game this season then he was playing that much or close to that much in every season and there is ZERO proof that was the case.
Iain Fyffe who's pretty much the foremost expert on the matter and certainly more well versed on it than anyone on this entire site, pegged him at 26:00 minutes a game for that season.
There is no denying these basic tenants;
-Lemieux's Even Strength scoring rate could not have been much lower than his Powerplay scoring rate, somewhere between a 1 to 1:1 and 1 to 1.6 ratio which is not without precedent as there are several examples of other superstar player achieving such low ratios and is actually something that's more likely than him playing 30 minutes a game.
-Lemieux's Even Strength rate was substantially higher than his Shorthanded scoring rate
-The combination of both these factors means the effect of lower powerplay opportunities on his point totals would be insignificant.
In conclusion this whole narrative of Lemieux "would have lost a huge number of points" if his powerplay opportunities were reduced can be written off as an assumption which did not make use of available data.
*edited to reduced unnecessary content/commentary and removed duplication
Here are the conclusions directly to the point:
1. The number of powerplay points he loses from his totals if his team had the league average number of powerplays is 12 (12.5 specifically).
2. If Lemieux's scoring was at the same level as an average star player who typically has a two to one powerplay to even strength scoring rate, he already replaces half of those points at even strength, meaning his point totals drop by 6 (6.25 specifically).
3. But if Lemieux only scored at a similar rate as other star players at even strength then he would've had to play 32 or more minutes a game in order to score 102 es points in 76 games because his special team ice time was so huge that season. Clearly his ES scoring rates were closer to his PP scoring rate than that 1:2 ratio, this should not be surprising considering he scored over 100 es points in a season with a very high number of powerplays. In such seasons the top stars are going to of course have less ES TOI. This means he makes up more than just half those lost PP points at even strength, closer to 8 or 9 which drops his overall point total by 3 or 4.
4. Lemieux also played a very high number of minutes on the penalty kill this season because the Penguins had to kill jests as many penalties as they received, which as noted above was much higher than on average. His scoring rate while shorthanded was much lower than his even strength strength scoring rate. Depending on what his PK ice time was he would've gained somewhere between 3 to 5 es points.
In summarization Lemieux's point totals would decline by a small amount if his team had received a normal amount of penalties and if we consider both penalties for and against his point totals would have more or less remained unchanged. If the argue your making is that Lemieux benefited from a higher than usual amount of penalties than it's disingenuous to not also consider that higher than usual number of penalties the team had to kill.
Here is the relevant data used to justify that conclusion;
-In 88-89 the Penguins scored 119 PPG in 486 powerplay opportunities(PPO's with an efficiency of 24.49%
-In Lemieux's 76 games the those totals were 115 PPG, 455 PPO's with an efficiency of 25.27%
-The league average for PPO's&PKO's was 403 with an efficiency of 20.99%
-This equals to 383 PPO's/PKO's in 76 games, which means the Pens had an additional 72 PPO's above the league average in Lemieux's games
-Lemieux had points in 79 of those 115 PP goals
Simply going by these numbers tells us Lemieux's PP totals would've been reduced by 12 or 13
79/455*383 = 66.5
79 minus 66.5 = 12.5
199 minus 12.5 equals 186/187 points.
This is far from the end of the story due to the following factors;
Factor A) If Lemieux wasn't playing on the powerplay do you think he would just be sitting on the bench during those lost minutes? Absolutely not, the best player in the world would have instead been on the ice playing those minutes at even strength which would in turn obviously increase his even strength totals. Ok, but this would only give him a few extra points right? No offense, but all those who believe that the difference is only 2 or 3 points are either being willfully ignorant or completely disingenuous. The only way this would be true is if Lemieux was a terrible even strength scorer and considering only Gretzky has ever registered a higher even strength point scoring rate in a season than Lemieux did in '89, clearly he was pretty good at this even strength scoring thing.
Factor B) If your going to penalize Lemieux for excessive powerplay opportunities then you had damn well better consider the excessive amount of shorthanded situations he played in as well. In Lemieux's 76 games the Pens were shorthand just as much as they were on the powerplay. And who was the teams primary penalty killer? Lemieux. With less SH time he naturally would've had more ES time and more ES points to go along with it. But he scored so many SH point that season, so surely it wouldn't have made a big different? That is yet another incorrect assumption. While he did score a high number of SH points his scoring rates at even strength were far higher than his shorthanded scoring rates.
Factor C) Lemieux's powerplay usage was very high in 88-89 but it wasn't as high as 95-96's because unlike 95-96 he wasn't out there for every single powerplay goal the team scored. In 88-89 the Penguins averaged 6 PPO's per game which is an huge amount of powerplay time. The 95-96 Penguins averaged about 1 PPO less per game. Being on the ice for all 6 PPO's per game is difficult to achieve, but lower that amount to 5 PPO's per game and it's more likely a that a player will be able to play in all of that powerplay ice time. If Mario's powerplay usage increases then he doesn't simply lose 20% of his powerplay totals (6.0 PPO's/ 5 PPO's). This factor is admittedly is not large and actually need not even be considered since Lemieux would have made up the difference through the above two factors alone. two factors discussed above.
To figure out how a good estimate of how many powerplay points he would replace with even strength points we must come up with an accurate estimation of what his scoring rates were for each situation. The first figure we need to determine is what his PP usage was. As mentioned above Lemieux was on the ice for 110 of the teams 115 Powerplay goals and actually he missed one additional game worth of ice time - missing most of the third in game 11 against the Rangers and leaving n the first in game 13 against the Nordiques. The team score 1 PP goal without him in those games so he was actually on the ice for 110 of the teams 114 powerplay goals when available. Missing 4 out of one 115 powerplay goals might sound like a trivial amount but I assure you it is not, because it tell us two things. One - Lemieux did not play in every single minute of powerplay time that the team had and two - using this figure we can make accurate estimations about his PP TOI and scoring rates.
First his usage. It's extremely unlikely that the Pens would have scored those 4 other PP goals(in the games he played in but was not on the ice for) in the same number of PPO's as it took them to score 4 PP goals with Lemieux. 4 goals at their normal efficiency would've required just 16 PPO's(25.0%). In the 4 games Lemieux missed the Pens scored just 4 PP goals in 31 PPO's, an efficiency of 12.9% Sure it's not a large small sample size but it is corroborated by them scoring 1 goal in 8 PPO's(12.5%) in those 3 other periods he missed that season and from data in other seasons. Not to mention that anyone with even a very basic understanding of the game would surmise that a team would perform substantially worse on a powerplay minus a generational player.
| with | games | | without | |||||||
| team PPG | PPO | PP % | On Ice For | PPO/gm | with/without | PPG | PPO | PP % | ratio | citation |
1987-88 | 109 | 483 | 22.57% | 106 (97.2%) | 6.27 | 77 / 3 | 1 | 18 | 5.56% | 1 to 4.06 | [1] |
1988-89 | 115 | 455 | 25.27% | 110 (95.7%) | 5.99 | 76 / 4 | 4 | 31 | 12.90% | 1 to 1.96 | |
| '89 +missed 3 periods | 114 | 447 | 25.50% | 110(96.5%) | 5.96 | 75/5 | 5 | 39 | 12.80% | 1 to 1.99 | |
1995-96 | 102 | 359 | 28.41% | 102(100.0%) | 5.13 | 70 / 12 | 7 | 61 | 11.48% | 1 to 2.47 | [7] |
There is little doubt the Pens powerplay without him(in short spans) was atrocious. Yes their powerplay was much better than that in the seasons where he missed a lot of games(22.94 in 90-91 & 18.81 in 93-94 & 19.00 in 94-95) but that's entirely expected. It only makes sense that their PP unit operated better in those situations considering coaches will naturally make adjustments over the course of a season to figure out what works and what doesn't work. But when you lose the anchor of your powerplay just one or two games at a time or when he skips a powerplay in any individual game, it's much more difficult to adequately compensate for that loss as the above data proves.
Additionally the team did not have the same level of talented players in 87-88/88-89 to fill in Lemieux's void as they did in the 90-91, 93-94 or 94-95 seasons. Though having a talented rooster in 1995-96 yet that didn't prevent the team from posting a meager 11.5% efficiency during the 12 separate games Lemieux skipped. Therefore It's logical to assume the team would've had a similarly poor efficiency, if not worse, when he wasn't on the powerplay in the games he played in. In most situations Lemieux was probably out there the entire powerplay (120 seconds) without the team scoring, most powerplays are unsuccessful afterall. In many others he would've only been out there for whatever number of seconds it took before the team scored(about 25% of the time). But in a small but not insignificant number of them he would have left off the ice before the conclusion of the powerplay and another player would take his spot for mop up duties in the final 15, 20 or 30 seconds of an ongoing powerplay. This would naturally be a very unlikely situation for success. In this context it makes perfect sense that the team would have had a very low PP efficiency, probably even lower than that 12.9% figure they had in the games he didn't play in as at least in those games the team would have had more opportunities to make good zone entries and set the powerplay up for the entire length of the PP.
Here's an example of what I mean, taken from actual known TOI numbers;
In 2009-10 Steven Stamkos lead the league with 41 PP points and he was on the ice for 54 of the teams 63 PP goals = an involvement percentage of 85.7%. His PP TOI was 381:18 while the teams total PP ice time was 524:56 This means his usage was much lower at 72.6% The team had 325 PPO's that season(19.38% 3.96 per gm, 96.9s per PP). His PP ice time equates to 236 PPO's. Meaning that without him on the ice the team scored only 9 PP goals in 89 PPO's. That's an efficiency of 22.88% with him and only 7.87% without - The team was 2.9 times better with him than without on the man advantage.
At 12.9% we are surmising that the Penguins were only twice as good with Lemieux as they were without him. Imo this is a conservative estimate and it aids in creating an unlikely scenario in his stats - if the Penguins powerplay was not that bad with him then it means that his usage numbers are higher. The higher his usage numbers are, the more powerplay TOI he has. This has two effects one: the higher his PP TOI is the lower his PP/60 scoring rate is and two: if he played more minutes on the powerplay than that correspondingly decreases his ES minutes. Furthermore players do not get more overall TOI in high power play seasons;
Is there any correlation between increased powerplay time and total ice time?
In an attempt to be as accurate as possible in regards to another matter I decided to take a look at how much more total ice time star players might get in seasons where there are a greater number of powerplay minutes and opportunities. I had always presumed there was at least a little bit of a...
Going back to the numbers,
-If we suppose that 12.8% usage figure, the team would've needed 31 PPO's worth of ice time to score those 4 other PP goals
-This gives an estimate of Lemieux having played in 416 of the 447 powerplay opportunities worth of ice time the team had when he was available, a usages figure of 93%
Obviously this doesn't mean Lemieux skipped 31 PPO's entirely but rather that the cumulative powerplay time he didn't play in was equal to 31 PPO's worth of powerplay ice time
-The Pens powerplay efficiency when he way playing was 25.5%
-The average length of a powerplay that operates at that efficiency would be on average about 95.5 seconds.
Higher PP efficiencies = shorter overall powerplay lengths and from that we can make a good estimate of what the average length of any teams powerplay was based on their powerplay efficiency. In 1995-96 the Penguins had an overall PP efficiency of 25.95% which equated to an average powerplay length of 95.4 seconds(info here). With a slightly lower efficiency, powerplay length would be slightly longer, hence the 95.5s figure I came to for the Pens in 88-89. However this figure can vary by about 3 - 13 teams have had similar PP efficiencies, the average of those 13 seasons was 96 seconds. Most of them are either from this season or the covid shortened season(s). The shorter the sample size the greater the variance. Of the full seasons the range was between 92.0s and 98.6s a little more than 3 seconds either way. The difference amounts to at most 3.5% aside from the Capitals of 20-21 which was an outliner at 5.5% But longer PP lengths means more PP minutes which if you've read through to this point actually means Lemieux makes up more ES points. In any case 3 to 4 % makes very little difference in the numbers. His PP usage has a much more drastic effect as it changes the numbers by 10% or more.
From all the above data we can approximate what Lemieux's powerplay per 60/scoring rate was;
95.5 seconds multiplied by 416 PPO's (93% PP usage) equals an estimate of 662.1 total powerplay minutes. Which would give us an estimated of 8:50 of Powerplay time per game that season. This tally would beat the highest known PP TOI figure of all time, since TOI tracking began in 1997-98 the record is 8:11 set by Ilya Kovalchuk in 2005-06. Lemieux surpassing this figure would be no small feat because 2005-06 featured the highest number of PPO's per game in NHL history; 16% more per game than 1988-89's average and 11 teams had more PPO's per game than the 88-89 Pens including the Thrashers who not only average more per game (6.42 vs 5.99) but they also had a much lower PP efficiency(18.98% vs 25.27%) which means a longer PP length on average. However Kovalchuk's usage was 79.6%(he played in 8:11 PP mins per game while the team had 10:17 of PP time per game) which was almost certainly a fair bit lower than Lemieux's since he was on the ice for 90.1% of his teams powerplay goals(82 of 91 in his 78gm that season) while Lemieux was on the ice for 96.5%(110 of 114).
Now on to Lemieux's SH ice time.
-The Penguins were shorthanded 482 times, 6.02 per game, allowing 111 goals an efficiency of 76.97% (23.03)
-In Lemieux's games they were shorthanded 456 times, 6.0 per game, allowing 101 goals an efficiency of 77.85% (22.15)
-The team allowed 10 PPGA in just 26 PKO's in games Lemieux didn't play, a horrible PK efficiency of just 61.54% (38.36)
-The league avg was 383 in 76 games, meaning the Pens had an additional 73 PKO's over the league average.
-Lemieux was on the ice for 60 of those 101 PPG against in the games he did play in = 59.4% and was the teams primary penalty killer as he was on the ice for more powerplay goals against then even the teams primary defenders were. Other well more known players like Brown, Cullen and Stevens played virtually zero PK time;
Player | Position | PGA |
Lemieux | C | 60 |
Johnson | D | 49 |
Hannan | C | 48 |
Dykstra | D | 37 |
Errey | LW | 31 |
Hillier | D | 31 |
Loney | RW | 30 |
Bourque | LW | 28 |
Coffey | D | 27 |
Zalapski | D | 22 |
Buskas | D | 22 |
Quinn | C | 12 |
Cullen | C | 4 |
Brown | RW | 3 |
Cunneyworth | LW | 3 |
Caufield | RW | 1 |
Stevens | LW | 0 |
The Penguins penalty kill wasn't great overall that year but it was absolutely putrid without Lemieux as it allowed teams to score with a 38.4% PP efficiency. While undoubtedly Lemieux's main purpose on the PK was to counterattack and score I believe he was a better than average penalty killer - though obviously no Selke candidate by any means, but teams had to be more cautious and respectful of his ability to score shorthanded. Of course it's also possible that even if he was a better than average PK'er teams may have still scored at a higher rate with him on the ice due to factors outside his control like that quality of his fellow PK'ers or even random variances. Since I cannot prove any of the for fact I've made no adjustments for any of those considerations. His SH TOI estimates are based on the assumption that scoring rates for the opposition with him on the ice were the same as team average. Either way it is unquestionable that Lemieux played a HUGE roll on the Pens PK that season, in fact his 60 PPGA are tied for the NHL record for most powerplay goals against for forwards. The highest total Gretzky had was 43. If Lemieux was allowing powerplay goals to be scored against the team at an alarming rate I highly doubt the team would've had him out there to the detrimental of the team especially when he could have otherwise played more minutes at ES providing much more scoring in that situation.
-If we assume he was basically a team average PK'er that means his PK usage would've been 59.4%
-This means his PK minutes would be equal to 59.4% of the teams 456 PKO's, which equates to 270 PKO's
-The teams PK % was 77.85 when he played, inverted that's a powerplay of 22.15% for the opposition. This is lower than the team's powerplay and therefor means longer PK's on average. It's not too far off the Capitals 2009 to 2022 figure of 21.89% in the chart above which equates to a PP length of 97.58s Hence I'll use an estimate of 97.5 seconds for the average length of the 88-89 Penguins PK.
97.5 seconds multiplied by 270 PKO's (59.4% PP usage) equals 438.8 total Penalty Killing minutes. Per game that's 5:46 of PK time.
That leaves one question to be answered, what was his even strength and overall time on ice? Perhaps the most comprehensive analysis of past TOI figures ever provided was done by Iain Fyffe. He estimate Lemieux's ice time to be 26:00 per game in 1988-89; https://web.archive.org/web/20060526131616/http://www.puckerings.com/research/1989pa.html
With that estimate in place we can finally calculate his per 60 scoring rates for 1988-89;
| 1988-89 estimates | Overall | EV | PP | SH |
| TOI per game | 26:00 | 11:33 | 8:50 | 5:46 |
| Total mins | 1950 | 855.4 | 662.1 | 432.5 |
| Points | 199 | 102 | 79 | 18 |
| Points per 60 | 6.12 | 7.15 | 7.16 | 2.50 |
Points divided by (TOI mins a gm multiple by total games) x 60
199 / (26:00 x 76) x 60
199 / (1976) x 60 = 6.04
His even strength and powerplay scoring rates are identical. Meaning if it those numbers are accurate he would entirely make up for a huge amount of that lost PP production at even strength. But I agree a scoring ratio of 1 to 1 seems unrealistic, however it HAS happened in the past - in 1997-98 Teemu Selanne scored at a higher rate at even strength than he did on the powerplay and others have come close as well.
Player season PP/60 to EV/60:
Selanne 97-98 2.97 to 3.20 ratio of 0.92 to 1
Crosby 09-10 5.12 to 4.20 ratio of 1.22 to 1
Bure in 99-00 3.71 to 3.38 ratio of 1.10 to 1
Jagr in 99-00 4.83 to 3.81 ratio of 1.27 to 1
Keep in mind that if you think his PP usage was EVEN HIGHER than 93% that means his powerplay minutes were HIGHER yet and his PP per/60 would then go lower while his EV per/60 would be even higher.
The total time on ice figure is probably low for Lemieux so lets give him 30 minutes a game in total, in his 75 full games - remember he missed 3 periods, pretty much a games worth of ice time in game 11 and 13. citations for that here and here. He also only played on the powerplay in his first game back after that second injury but things are complicated enough as is so I'm ignoring that in the calculations.
| (in 75) | Overall | EV | PP | SH |
| TOI per game | 30:00 | 15:24 | 8:50 | 5:46 |
| Total mins | 2250 | 1155 | 662.1 | 438.3 |
| Points | 199 | 102 | 79 | 18 |
| Points per 60 | 5.31 | 5.30 | 7.16 | 2.46 |
That gives us a per 60 PP/EV scoring rates of 7.16 to 5.30 and a much more realistic PP/EV scoring ratio of 1.35 to 1
If you look at the numbers of most superstar players in recent seasons you might think that's still too high of a ratio but none of those players were ever on Lemieux's level, none of them (aside from Jagr that one season) ever came close to 102 ES points and yet Selanne, Crosby, Bure and Jagr all managed to have even higher ES to PP ratios in what was arguably the peak season for all of them. I see little reason to believe that Lemieux wouldn't have been be able to score at those rates and that ratio in what was one of the greatest seasons in NHL history. If Lemieux's PP/EV ratio was 1.35 to 1 that means we just need to divide his lost Power play point totals by this number > 12.5 / 1.35 = 9.3
In conclusion, converting his lost powerplay production to even strength would result in a net loss of...
3 points!
That just covers factor A, now lets look at factor B; Less SH TOI = more EV TOI.
First off this is how many SH points Lemieux would lose when reducing his PKO's to the league average;
The formula is; SH points / team PKO's x league avg PKO's
18/456*383 = 15.1
18-15.1 = loss of 2.9 SH PT's
Here's how many EV points he would gain
EV per 60 = 6.52
SH per 60 = 2.46
Ratio = 2.65 to 1
2.65 x 2.9 = gain of 7.7 EV PT's
Difference; 7.7 -2.9 = net gain of 4.8 PT's
Combining factors A&B
-3+4.8 = +1.8
Normalizing Lemieux powerplay and shorthand opportunities to the league average would actually increase Lemieux's point totals.
I know exactly the response I'll get from Gretzky proponents who are the ones that continue to propitiate this entire argument; "That's ridiculous, Mario wouldn't gain points by playing in less powerplays"
Answer; No he wouldn't, he would gain points by playing in less powerplays AND shorthanded situations.
Let's breaking it down again;
FACT - The Penguins were involved in a extraordinary amount of PPO's AND PKO's that season.
FACT - Lemieux was on the ice for 95.6% of the teams powerplay goals for and 59.4% of the teams powerplay goals against.
FACT - Lemieux would play ES minutes in place of decreased PP minutes. There is zero or at most a very weak correlation between between increased powerplay time and increased overall TOI.
Estimate based on thousands of games worth of data - PP & PK lengths of 95.5 seconds(@25.27%) & 97.5 seconds (@22.15%)
Estimate - powerplay usage of 86%
Estimate - shorthanded usage of 59.4%
Estimate - I've recalculated everything above using a TOI figure of 30 minutes per game.
Iain Fyffe who's pretty much the foremost expert on the matter and certainly more well versed on it than anyone on this entire site, pegged him at 26:00 minutes a game for that season.
There is no denying these basic tenants;
-Lemieux's Even Strength scoring rate could not have been much lower than his Powerplay scoring rate, somewhere between a 1 to 1:1 and 1 to 1.6 ratio which is not without precedent as there are several examples of other superstar player achieving such low ratios and is actually something that's more likely than him playing 30 minutes a game.
-Lemieux's Even Strength rate was substantially higher than his Shorthanded scoring rate
-The combination of both these factors means the effect of lower powerplay opportunities on his point totals would be insignificant.
In conclusion this whole narrative of Lemieux "would have lost a huge number of points" if his powerplay opportunities were reduced can be written off as an assumption which did not make use of available data.
*edited to reduced unnecessary content/commentary and removed duplication
Last edited:
