Auston Matthews has more goals and a higher gpg than Ovechkin age for age. And the gap is about to grow. Can he also make a run at 894?

[Vilica]

This is one of the main criticisms of the VsX methodology. Those sorts of nonsensical outcomes are common and also easily avoidable. Why have a method where people spot check and fix, or not, based on whatever agenda is being served in the moment (not saying you have one here), as opposed to a repeatable standard? What is the advantage of that?

For examples, there are many periods where league scoring really didn't change much over a 4 or 5 year period. Why limit the comparison to 1 season in instances where there are 5 comparable seasons?

That's exactly what I'm showing that my methodology does. By Average GVsX, Bure's seasons are 126.39 and 124.80, because they are nearly the same year, and not 109.26 and 131.82. The actual GVsX for 99-00 ends up being 46.47 instead of 44, and 46.68 instead of 54 in 00-01. I am using the expanded sample of 05-06 through 18-19 to create an Average - some years will be high, some years low, some years correct, and taking an average will be more accurate than a 1 season sample. To give an example from Points instead of Goals, when scoring stagnated around 218/219 from 11-12 through 15-16, Average VsX comes out with a value of around 96/97 points for each of those years, like 11-12 and what 12-13 would have prorated to in 82 games, and not the 87, 86, and 89 from the actual VsX set those subsequent years.

 

[BraveCanadian]

I meant instead of looking at scoring versus 10th in scoring, the average of the 10 scorer will tend to be way less noisy.

Bure 58 goals VsX being what 25% higher in Vx2 type of model than the following season 59 goals season show how noisy it can be, second best scorer from year to year are quite different beast, it can be prime Mario-Gretzky or Jamie Benn. Yzerman 155 pts season is even a minus something in some Vx way to look at it.

I meant more that him claiming average scoring somehow wasn't ignoring surrounding seasons as well as a lot of other context. Because average scoring ignores all context whereas VsX contains it by default because it compares to peers.

But yes, outlier seasons are a problem which is why VsX drops extreme outliers.. the drawback of that is it is a but of a subjective line drawn but still vastly superior to going by average scoring when comparing players.

 

[Midnight Judges]

That's exactly what I'm showing that my methodology does. By Average GVsX, Bure's seasons are 126.39 and 124.80, because they are nearly the same year, and not 109.26 and 131.82. The actual GVsX for 99-00 ends up being 46.47 instead of 44, and 46.68 instead of 54 in 00-01. I am using the expanded sample of 05-06 through 18-19 to create an Average - some years will be high, some years low, some years correct, and taking an average will be more accurate than a 1 season sample. To give an example from Points instead of Goals, when scoring stagnated around 218/219 from 11-12 through 15-16, Average VsX comes out with a value of around 96/97 points for each of those years, like 11-12 and what 12-13 would have prorated to in 82 games, and not the 87, 86, and 89 from the actual VsX set those subsequent years.

Yeah I agree it appears that avg VsX mitigates volatility to a decent degree. Your outcomes are now much closer to hockeyreference.com's highest adjusted goal seasons than the other versions of VsX I've seen.

 

[MadLuke]

average scoring when comparing players.

Sure it is hard imagine a worst way than that one.

I was trying to simply one big rational of using a specific (2-7-10) player to compare instead of say the average top 7-10 players, would be that it is much easier-faster to do even if it look inferior (cannot see the argument for it outside ease of use).

It is not like there is money to be made here or that it matter much, something easier-faster to do is not a bad rational to choose between 2 options, specially that none will be perfect. A bit like looking at Top-10 finish, maybe it make no much sense to punish-reward a simple 1-2 pts either ways changing that number, but it is quick to look at, already available on some site, etc.. and still give some rough idea.

 

[BraveCanadian]

Sure it is hard imagine a worst way than that one.

I was trying to simply one big rational of using a specific (2-7-10) player to compare instead of say the average top 7-10 players, would be that it is much easier-faster to do even if it look inferior (cannot see the argument for it outside ease of use).

It is not like there is money to be made here or that it matter much, something easier-faster to do is not a bad rational to choose between 2 options, specially that none will be perfect. A bit like looking at Top-10 finish, maybe it make no much sense to punish-reward a simple 1-2 pts either ways changing that number, but it is quick to look at, already available on some site, etc.. and still give some rough idea.

Previously people had the idea of doing the average of 3 x the number of teams in the league to capture the average of all front liners who presumably have similar usage and PP time and all that.

There are 100 ways to look at this it just depends how fast you can do it and how many numbers you have to crunch. I'm quite confident that most methods of comparing to peers and then comparing standing vs. peers across time are superior to adjusting based on average scoring. This can be seen very easily when looking at outlier seasons like 1993, 1996, 2005 etc where average scoring does an abysmal job.

 

[MadLuke]

I'm quite confident that most methods of comparing to peers and then comparing standing vs. peers across time are superior to adjusting based on average scoring.

Yes for what we are trying to do comparing with average scoring make no sense, (I feel like we both agree on this on the beginning about that), I feel like I really unclear maybe, by average scoring I mean if you are interesting to compare to the 7th best scorer type, you probably better to calculate the average top x scorer (excluding the person of interest) instead of an specific rank finish (or always look at a bunch of them, just to avoid a strange exceptional finish misleading you).

But there is an obvious good rational not to and do a simple VsX instead, saving time and headache, specially if you are going by hands at it.

 

[Video Nasty]

I've given it some thought and I don't think it's even fair to label Matthews as injury prone at this point. People are making way too big of a deal about it. He is still very durable.

Since missing a month due to shoulder injury early in the 2018-2019 campaign, this is what he has suited up for:

Remainder of 2018-2019: 57 GP of 57
2019-2020: 70 of 70
2020-2021: 52 of 56
2021-2022: 73 of 82
2022-2023: 74 of 82
In progress 2023-2024: 46 of 47

He has played 372 out of the Leafs past 394 games, 94.4%, or 77.42 games/82.

@ImporterExporter mentioned this within the first few comments, but it also helps we're currently in a league that is doing what they can to make sure their best players have the opportunity to play the available games, as well as a really nice scoring environment that I think the NHL has figured out needs to remain and be at a somewhat stable level (at least 6 GPG).

Matthews is certainly in the driver's seat right now. Ovechkin had another weak year during his age 26 season and then a lockout, before collecting 104 more goals at age 28 and 29, to enter his age 30 season with 475. He didn't pot his 600th until late into his age 32 campaign.

I'll absolutely become a believer if Matthews pots 70 this season,and then collects somewhere around 150-170 goals during his age 27-29 seasons, to enter his age 30 campaign with somewhere between 520-540 goals. If he's somehow flirting with 600 near the end of his age 30 season or early into his age 31 season, I think he has a great shot.

Ovechkin bucked a lot of expectations, but regardless of the reasons, the fact remains that he "only" had 1 season past the age of 23 where he topped the 52 he had as a rookie, and it was with 53. The 48-51 goal seasons not being 58-61 goals seasons give someone like Matthews a shot.

 

[blundluntman]

No way for sure to know until we see how he ages. Ovi's 2nd goalscoring prime wasn't something most saw coming, and I don't think many expected it to last as long as it did. There's no real reason to expect anybody, let alone Matthews could maintain that level of play for any stretch of time let alone in his older years

 

[WarriorofTime]

I'm quite confident that most methods of comparing to peers and then comparing standing vs. peers across time are superior to adjusting based on average scoring.

It's almost certainly the opposite. The 'peers' method requires a lot of work to try and get apples to apples based on rotating deployments and injuries. League-wide scoring is league-wide scoring. The same guy may not play 1st Line Center/PP1 all 82 games for a given team, very relevant for 'peer' study but we know someone on that team played 1st Line Center/PP1 in each of the 82 games and the cumulative efforts there are reflected in the League-wide scoring as a result. It doesn't rely on circumstances of a few players situation in a particular year to have a massive effect on outcomes. Trying to compare to "peers" is a fundamentally backwards system.

 

[MadLuke]

League-wide scoring is league-wide scoring.

Is there anyone that simple use league-wide scoring over time that seem improbable a bit.

A HR method is still doing quite a bit of work to adjust for deployments, but it use roster size (maybe not a bad proxy in baseball, but for hockey with how much more ice stars can get versus the last forward added not so much and adding D changing virtually nothing for star forward ice time) for it and you need to adjust for the assist by goals change over time when it come to points.

. It doesn't rely on circumstances of a few players situation in a particular year to have a massive effect on outcomes.

It can depend on what we mean by a few someone that compare to the leader or the second place or the average top 3, but once you average say the top 15 in the 06 and 25+ type of players in the larger league, the average season to season does not seem to move much without a change that we also see in the league average (say 04 to 06 season or 95 to 96) and when the average goes down.

Trying to compare to "peers" is a fundamentally backwards system.

Do you really think there is absolutely nothing that can make taking scoring distribution into account, maybe it is just talking about 2 different things.

Someone can be interested into the value of a goals and adjust for that (i.e. league wide scoring), which can be used after that to know the value of average starting goaltender versus their replacement over time or scoring d vs forward output overtime or something like that.

But we tend to do it to compare individual scorer over time, if someone played in a era where defenceman did not score goal why would that make them better scorer versus someone that did ?, which seem to be what is proposed.

 

[WarriorofTime]

Is there anyone that simple use league-wide scoring over time that seem improbable a bit.

A HR method is still doing quite a bit of work to adjust for deployments, but it use roster size (maybe not a bad proxy in baseball, but for hockey with how much more ice stars can get versus the last forward added not so much and adding D changing virtually nothing for star forward ice time) for it and you need to adjust for the assist by goals change over time when it come to points.


It can depend on what we mean by a few someone that compare to the leader or the second place or the average top 3, but once you average say the top 15 in the 06 and 25+ type of players in the larger league, the average season to season does not seem to move much without a change that we also see in the league average (say 04 to 06 season or 95 to 96) and when the average goes down.


Do you really think there is absolutely nothing that can make taking scoring distribution into account, maybe it is just talking about 2 different things.

Someone can be interested into the value of a goals and adjust for that (i.e. league wide scoring), which can be used after that to know the value of average starting goaltender versus their replacement over time or scoring d vs forward output overtime or something like that.

But we tend to do it to compare individual scorer over time, if someone played in a era where defenceman did not score goal why would that make them better scorer versus someone that did ?, which seem to be what is proposed.

Fundamentally if the top 2 percentile guy was closer to leading scorer but the era is lower scoring that means we’re gonna see someone be further away eventually, wherever it be, the 5th percentile guy, the 10th percentile guy, the 50th percentile, whatever it may be. We get lots of noise if we are just looking at a specific, very high percentile number. We are placing disproportionate influence on circumstances involving the very top if we look at those very high. Looking at the wider numbers across the whole league spectrum will paint a clearer picture over time to be analyzed. Who the top liners are on a given night, how and which defensemen are contributing to scoring and if they are even supposed to be, these are superfluous questions. What we know is how many scores on average get scored in a given game, and how often is a particular player the one doing that.

 

[daver]

It's almost certainly the opposite. The 'peers' method requires a lot of work to try and get apples to apples based on rotating deployments and injuries. League-wide scoring is league-wide scoring. The same guy may not play 1st Line Center/PP1 all 82 games for a given team, very relevant for 'peer' study but we know someone on that team played 1st Line Center/PP1 in each of the 82 games and the cumulative efforts there are reflected in the League-wide scoring as a result. It doesn't rely on circumstances of a few players situation in a particular year to have a massive effect on outcomes. Trying to compare to "peers" is a fundamentally backwards system.

How many top 10/20/30 scorers fall into this category in any given year? One or two?

IMO, when talking about offensive numbers by all-star calibre forwards, deployment is secondary, but not necessarily irrelevant.

IMO, Matthews' goalscoring numbers speak for themselves, and a comparison of how their respective goalscoring compares using the peers method doesn't need context.

Question for you

If comparing using league GPG is significantly different from comparing how they finished vs. the #5, #10, #20 and #40 scorer in a season, are you seriously believing the league GPG should be the default metric?

Mike Bossy's 147 point season in 81/82, 2nd that year to Wayne and 3rd best non Wayne/Mario point total of the '80s "adjusts" to a 6th place showing in 98/99.

Does that sound reasonable to you?

 

[MadLuke]

. What we know is how many scores on average get scored in a given game, and how often is a particular player the one doing that.

But I imagine even you would be ready to consider the fact that when Joe Malone scored 2.2 goal a night in a league where the average team scored 4.77 goals a night, should not automatically adjust the same to someone that would have scored 69 goals the first 49 game of this season.

That the fact that they was 9-10 regular player on team will change (for the same talent, importance, everything) the proportion of goal a scorer will have relative to the team league average scoring.
Or maybe you disagree with hockey reference adjustment even more than the average hfboard user ? saying that it should not be taken into account ?

We are placing disproportionate influence on circumstances involving the very top if we look at those very high.

I am not sure it is possible to take the circumstance of the elite players too much, when that it is the only thing that interest us usually.

 

[WarriorofTime]

Mike Bossy's 147 point season in 81/82, 2nd that year to Wayne and 3rd best non Wayne/Mario point total of the '80s "adjusts" to a 6th place showing in 98/99.

Does that sound reasonable to you?

This is what you need to train your brain away from. 100 guys had 60 or more points in 1981-82 in a 21 team nhl. 38 guys had 60 or more points in 1998-99 in a 27 team nhl.

 

[WarriorofTime]

But I imagine even you would be ready to consider the fact that when Joe Malone scored 2.2 goal a night in a league where the average team scored 4.77 goals a night, should not automatically adjust the same to someone that would have scored 69 goals the first 49 game of this season.

Obsessing over war years and seasons over 100 years ago is not particularly relevant if we’re discussing usefulness of the last 50 years or so of adjusted goal scoring. It’s letting perfect be the enemy of good in order to prop up an even worse and less mathematically based method, which is just odd.

 

[MadLuke]

Obsessing over war years and seasons over 100 years ago is not particularly relevant if we’re discussing usefulness of the last 50 years or so of adjusted goal scoring

It make just clear that the fundamental are shaky, we can do the exact same for the 16 player roster size versus 18 (excluding goaltender) of some seasons in the 60s, it will just pop-up less.

It is rarely used only for the post 06 era, if you are against roster size adjustment you are "with us " united against using hockey-reference values. If you are using HF values, you are doing roster size and other tricks and not simply using league scoring average, maybe all along you are having a different conversation.

prop up an even worse and less mathematically based method,

Too be clear I am really not a fan to comparing the Rocket Richard winner with the second position if you think that what I am propping up, I simply want to adjust for scoring distribution caused for example by having 3v3 overtime or a change in powerplay opportunity and not just taking for granted that everyone happen to have their career year because they were better at hockey in 1993.

 

[WarriorofTime]

It is rarely used only for the post 06 era, if you are against roster size adjustment you are "with us " united against using hockey-reference values. If you are using HF values, you are doing roster size and other tricks and not simply using league scoring average

I don’t subscribe to “us” versus them and strive to be a free thinker. I need more data than “this is what everyone here has decided” when it seems grounded in pseudo-science.

 

[WarriorofTime]

Too be clear I am really not a fan to comparing the Rocket Richard winner with the second position if you think that what I am propping up, I simply want to adjust for scoring distribution caused for example by having 3v3 overtime or a change in powerplay opportunity and not just taking for granted that everyone happen to have their career year because they were better at hockey in 1993.

This requires a much more precise game by game inquiry across an entire league into line usage/deployment across all teams, not using shorthands of top so and so players. It’s a bold and ambitious project what you describe but not one that should be made up for by using faultier methods like looking at whomever was 15th in scoring across the 21 team nhl and comparing it to 15th in scoring across a 32 team nhl and trying to proxy that out. Especially when you’re trying to roundabout your way back to the original question - how hard was it to score in that season?

 

[MadLuke]

I don’t subscribe to “us” versus them and strive to be a free thinker. I need more data than “this is what everyone here has decided” when it seems grounded in pseudo-science.

We all love to think that

But I am still unsure, you too are against HF ways too adjust right ? To come back to the original conversation.

Especially when you’re trying to roundabout your way back to the original question - how hard was it to score in that season?

That where I think you are trying to achieve something else than the usual way adjustment are made and their goals, we usually adjust to compare the best scorer of all time that played in different seasons.

And the question that interest us, is more how hard was it to score for a Top forward playing on the first unit of the powerplay that season, not how much it was hard in general, because we are trying to compare people that were in that situation. You can disagree to be interested in this, but it would explain the mathematic going on for that goal.

If starting with the 2025 season, the league become 4x4 hockey with 4x3 powerplay becoming the norm and coach start to use mostly 8 forward, and all of a sudden using simply just league average or HR adjustement way if look like the 20 top forward scored more adjusted goals in 2025 than the previous season, would you be open to use something else to adjust ?

 

[bobholly39]

I've given it some thought and I don't think it's even fair to label Matthews as injury prone at this point. People are making way too big of a deal about it. He is still very durable.

Since missing a month due to shoulder injury early in the 2018-2019 campaign, this is what he has suited up for:

Remainder of 2018-2019: 57 GP of 57
2019-2020: 70 of 70
2020-2021: 52 of 56
2021-2022: 73 of 82
2022-2023: 74 of 82
In progress 2023-2024: 46 of 47

He has played 372 out of the Leafs past 394 games, 94.4%, or 77.42 games/82.

@ImporterExporter mentioned this within the first few comments, but it also helps we're currently in a league that is doing what they can to make sure their best players have the opportunity to play the available games, as well as a really nice scoring environment that I think the NHL has figured out needs to remain and be at a somewhat stable level (at least 6 GPG).

Matthews is certainly in the driver's seat right now. Ovechkin had another weak year during his age 26 season and then a lockout, before collecting 104 more goals at age 28 and 29, to enter his age 30 season with 475. He didn't pot his 600th until late into his age 32 campaign.

I'll absolutely become a believer if Matthews pots 70 this season,and then collects somewhere around 150-170 goals during his age 27-29 seasons, to enter his age 30 campaign with somewhere between 520-540 goals. If he's somehow flirting with 600 near the end of his age 30 season or early into his age 31 season, I think he has a great shot.

Ovechkin bucked a lot of expectations, but regardless of the reasons, the fact remains that he "only" had 1 season past the age of 23 where he topped the 52 he had as a rookie, and it was with 53. The 48-51 goal seasons not being 58-61 goals seasons give someone like Matthews a shot.

I agree about the durability.

Also - I'm trying to be a bit more conservative than you and not project too high goal totals for him outside of this season and next, and I still very easily get to 800+ goals.

Hypothetically:

2023-2024 65 goals (on pace for 71)
2024-2025 57 goals

421 career goals (50 more than Ovechkin at same age)

2025-2026 until 2027-2028 (so 3 years, ages 28-30). Average of 45 goals per year. 556 goals by age 30

2028-2029 to 2030-2031 (3 years, ages 31-33) average of 40 goals per year. 676 goals by age 33

2031-2032 to 2035-2036 (5 years, ages 34-38) averages of 30 goals per year. Total = 826 goals

I think my estimates for ages 28-38 are actually quite conservative for a goal-scorer of his stature, and even allows for injuries and some decent decline. So long as he avoids major injuries or an unexpected sharp decline, I think he's very well positioned to do 800+ career goals minimum. If he ages better than even this, he could make a run for 894.

 

[WarriorofTime]

And the question that interest us, is more how hard was it to score for a Top forward playing on the first unit of the powerplay that season, not how much it was hard in general, because we are trying to compare people that were in that situation.

This naturally discounts the team that has the "rotating cast of fake 1st liners/PP1" players that get thrown into that role up and down the lineup throughout the year. But in some ways that's almost the exact scenario you should be thinking about to answer the question. If we say replacement level players really struggling in that same role versus still scoring a good amount in whatever sample they are in that role, it speaks volumes to what the top players are doing in a relative sense.

If starting with the 2025 season, the league become 4x4 hockey with 4x3 powerplay becoming the norm and coach start to use mostly 8 forward, and all of a sudden using simply just league average or HR adjustement way if look like the 20 top forward scored more adjusted goals in 2025 than the previous season, would you be open to use something else to adjust ?

If that happened it would be a fundamental change that has not happened to a level since Gordie Howe was playing. If we just look at the landscape of the League of former players that are still alive we see less change that doesn't make comparisons so hard using HR adjustment, comparing to World War II or whatever.

 

[Vilica]

How many top 10/20/30 scorers fall into this category in any given year? One or two?

IMO, when talking about offensive numbers by all-star calibre forwards, deployment is secondary, but not necessarily irrelevant.

IMO, Matthews' goalscoring numbers speak for themselves, and a comparison of how their respective goalscoring compares using the peers method doesn't need context.

Question for you

If comparing using league GPG is significantly different from comparing how they finished vs. the #5, #10, #20 and #40 scorer in a season, are you seriously believing the league GPG should be the default metric?

Mike Bossy's 147 point season in 81/82, 2nd that year to Wayne and 3rd best non Wayne/Mario point total of the '80s "adjusts" to a 6th place showing in 98/99.

Does that sound reasonable to you?

Name	Year	Team	Games	Goals	Assists	Points	Team GF	LA GF	% LA	G%	P%		Avg VsX	LA GF	New GF	Goals	Assists	Points
Wayne Gretzky	81-82	EDM	80	92	120	212	417	321	1.299	0.221	0.508		150.66	216	280.60	61.91	80.75	142.65
Mike Bossy	81-82	NYI	80	64	83	147	385	321	1.199	0.166	0.382		104.46	216	259.07	43.07	55.85	98.92
Peter Stastny	81-82	QUE	80	46	93	139	356	321	1.109	0.129	0.390		98.78	216	239.55	30.95	62.58	93.53
Dennis Maruk	81-82	WAS	80	60	76	136	319	321	0.994	0.188	0.426		96.65	216	214.65	40.37	51.14	91.51
Jaromir Jagr	98-99	PIT	81	44	83	127	242	216	1.120	0.182	0.525		134.12	321	359.64	65.39	123.35	188.74
Teemu Selanne	98-99	ANA	75	47	60	107	215	216	0.995	0.219	0.498		113.00	321	319.51	69.85	89.17	159.01
Paul Kariya	98-99	ANA	82	39	62	101	215	216	0.995	0.181	0.470		106.67	321	319.51	57.96	92.14	150.10
Peter Forsberg	98-99	COL	78	30	67	97	239	216	1.106	0.126	0.406		102.44	321	355.18	44.58	99.57	144.15
Joe Sakic	98-99	COL	73	41	55	96	239	216	1.106	0.172	0.402		101.39	321	355.18	60.93	81.74	142.67

Here's the top four point scorers in 81-82 converted to a 216 level in 98-99, and the top 5 scorers in 98-99 converted to the 321 level in 81-82. Because the scoring levels are so far apart, we're talking more than 100 goals added or subtracted over a full season - Gretzky and Bossy lose 137 and 126, Stastny and Maruk 117 and 105, while Jagr gains 118 goals, the Anaheim players 105, and the Colorado players 116.

Bossy's adjusted total of 99 points means he finishes 4th instead of 6th, but that's just an artifact of single-season variance. Bossy finishes 1st in scoring in 99-00, 3rd in 00-01, 1st in 01-02, 4th in 02-03, and 1st in 03-04. The 98-99 year just happens to be a season where multiple players had high-end seasons. Even when he finishes 4th in 02-03, the point totals go 106 104 101 100 98 97 for Forsberg Naslund Thornton "Bossy" Hejduk Bertuzzi. There's certainly an argument that Bossy would have competed harder for points with 1st place up for grabs, rather than being 60 points behind 1st, but we can only go by the lines they put up in their actual seasons.

Is there any real difference between '3rd best non-Wayne/Mario point total of the 80s' and 'would have won the Art Ross in 50% (3 of 6 seasons) in the sample'?

 

[jigglysquishy]

Do you have the top 50 seasons using the average VsX method?

 

[MadLuke]

If we just look at the landscape of the League of former players that are still alive we see less change that doesn't make comparisons so hard using HR adjustment, comparing to World War II or whatever.

I think it is a bit why the conversation can get frustrating, we go from talking about using league scoring average to something way more complex like HR adjustment as if it is the same and being used interchangeably.

In the 1982 season, the average team was scoring 321 goals
In the 1983 season, the average team was scoring 309 goals

Mike Bossy scored 147 points in 1982, HR adjust for 108 pts
Mike Bossy scored 118 points in 1983, HR adjust for 95 pts

In 1982 it was .7347 adjusted points for each actual points (108/147)
In 1983 it was .80508 adjusted points for each actual points

.7347 / .80508 = .91258
309 / 321 = .962617

Why such a difference between adjusted points ratio between the 2 seasons and the average scoring, it is because in 1983 the league added a roster spot and could be (did not check) that the average assist per goals ratio changed a bit, arguing for Hockey reference is arguing against simply adjusting using league wide scoring average for era of players still alive today as well.

You seem to be arguing for and against HR reference way of adjusting at the same time. And if we go back to what started the argument, a top 5 season of all time it is not like it did not involve what pre-expansion player like Hull-Richard-Howe did.

 

[WarriorofTime]

I think it is a bit why the conversation can get frustrating, we go from talking about using league scoring average to something way more complex like HR adjustment as if it is the same and being used interchangeably.

In the 1982 season, the average team was scoring 321 goals
In the 1983 season, the average team was scoring 309 goals

Mike Bossy scored 147 points in 1982, HR adjust for 108 pts
Mike Bossy scored 118 points in 1983, HR adjust for 95 pts

In 1982 it was .7347 adjusted points for each actual points (108/147)
In 1983 it was .80508 adjusted points for each actual points

.7347 / .80508 = .91258
309 / 321 = .962617

Why such a difference between adjusted points ratio between the 2 seasons and the average scoring

Don't follow the complaint. League-wide scoring dipped, so each point becomes harder to obtain. So when we try and adjust we'd give more reverence to a 1983 point compared to a 1982 point. 4.01 goals per game compared to 3.86 goals per game is a fairly noticeable difference that a closer observer would notice with their own eyes without any spreadsheets while watching games.