Has Anyone Not Named Gretzky, Lemieux or Orr ever dominated a year like Sakic?

[Iain Fyffe]

Here's my theory on stats: They are what they are! You know what, instead of quantifying, adjusting, and debating what variables to add to the quadratic formula to figure out how many points so-and-so's point total is worth in whatever imagined statistical variable someone has decided is valid, WE COULD JUST STICK WITH REAL STATS.

But you don't stick with the real stats. You said so yourself: you consider the context. You realize that scoring levels change, that styles of play change. This means you are making mental adjustments when comparing stats from different eras.

The problem with adjusting stats is the philosophy of the 'adjuster' that stats require adjusting in order to form comparisons. They don't.

But this is exactly what you're doing when making your considerations of context. You might not do it explicitly, but you're doing it. You know that Anderson's 105 points are not more impressive than Howe's 95, which means you do not just stick with real stats. You apply mental adjustments.

 

[The Panther]

Since my point is being missed, let me put it in simpler terms: You can't compare stats across different eras. (Even comparing stats in the same season can be wonky in that no two players play exactly the same style, but anyway...)

That is basically what I'm saying. So, let's just leave the stats as they are. The players in each era earned their stats in their own era.

We don't need stats to compare Glenn Anderson to Gordie Howe. We could use stats to compare Glenn Anderson to Steve Larmer, but even then we need to consider the context.

Adjusting stats does not solve the problem of relativity and context. Stats are stats. Leave them as they are.

 

[Iain Fyffe]

That is basically what I'm saying. So, let's just leave the stats as they are. The players in each era earned their stats in their own era.

This suggests a common misunderstanding with respect to adjusted stats. Adjusted stats are not meant to indicate what a particular player would have scored in another era. They indicate a comparable level of performance, considering the contexts of each era, that's all. (Some do use it in the former sense, but that's not really defensible, while the latter sense is.)

When someone mentions the Consumer Price Index, do you tell them that you can't compare prices across eras? It's a pretty good analogy, because while the CPI may not work well for certain specific goods, on the whole it presents useful information that you would not otherwise have in a usable form.

We don't need stats to compare Glenn Anderson to Gordie Howe. We could use stats to compare Glenn Anderson to Steve Larmer, but even then we need to consider the context.

If you cannot compare stats across era, then you cannot compare stats across teams either. Anderson and Larmer had different teammates, different distributions of opponents, different home arenas, different travel schedules, different roles and expectations placed upon them, etc, etc. You're drawing an arbitrary line as to how much difference is too much difference in context.

Adjusting stats does not solve the problem of relativity and context.

No, but they do help.

 

[Bear of Bad News]

When someone mentions the Consumer Price Index, do you tell them that you can't compare prices across eras? It's a pretty good analogy, because while the CPI may not work well for certain specific goods, on the whole it presents useful information that you would not otherwise have in a usable form.

This is the analogy I prefer, too.

By way of example, goals are not useful in and of themselves - they're only useful in the context of producing team wins. So it's natural to ask how many goals does it take to purchase a win. The answer's quite different in 1984-85 than in 1998-99..

Yes, sometimes you get a bargain and can purchase a win for only one goal. Sometimes, you spend eight goals and can't get a win. But in terms of long-term value, there's a definite difference in how many wins you can "buy" with X goals in a year.

It's not about adjusting performances.

 

[Fantomas]

This suggests a common misunderstanding with respect to adjusted stats. Adjusted stats are not meant to indicate what a particular player would have scored in another era. They indicate a comparable level of performance, considering the contexts of each era, that's all. (Some do use it in the former sense, but that's not really defensible, while the latter sense is.)

When someone mentions the Consumer Price Index, do you tell them that you can't compare prices across eras? It's a pretty good analogy, because while the CPI may not work well for certain specific goods, on the whole it presents useful information that you would not otherwise have in a usable form.


If you cannot compare stats across era, then you cannot compare stats across teams either. Anderson and Larmer had different teammates, different distributions of opponents, different home arenas, different travel schedules, different roles and expectations placed upon them, etc, etc. You're drawing an arbitrary line as to how much difference is too much difference in context.


No, but they do help.

Excellent post.

 

[The Panther]

This suggests a common misunderstanding with respect to adjusted stats.

Actually, my post did not in any way suggest that misunderstanding.

Adjusted stats are not meant to indicate what a particular player would have scored in another era. They indicate a comparable level of performance, considering the contexts of each era, that's all.

Now, this starts to make some sense. OK then, but why then do we not -- instead of tweaking numbers of goals from 45 to 29 when 45 pucks actually went in the net -- simply rate goals, assists, and points (and everything else) on a scale of 1 to 100?

For example, in 1984 Gretzky scores a 100 for point/goal production and a 100 in 1986 for assist production? Then, everybody else in history has a lesser number based on how they compare to this standard? There's no need to pretend Gretzky scored 68 goals or whatever when he actually scored 87. We just score everyone's seasonal production out of 100.

That would make more sense to me, and would avoid a 12-year-old thinking that Anderson '83 had a better year than Howe '53.

When someone mentions the Consumer Price Index, do you tell them that you can't compare prices across eras? It's a pretty good analogy

Well, I think this is a poor analogy because the Consumer Price Index is rating the overall value of a car, but how many points Anderson scored in 1983 is not an evaluation of his overall value.

If you cannot compare stats across era, then you cannot compare stats across teams either. Anderson and Larmer had different teammates, different distributions of opponents, different home arenas, different travel schedules, different roles and expectations placed upon them, etc, etc.

That's pretty much what I already said in my previous post: "We could use stats to compare Glenn Anderson to Steve Larmer, but even then we need to consider the context."

Stats are not sufficient as player-comparisons in themselves, as you point out. Therefore, why adjust them at all?

Stats to me are counts of how many times something happened. Gretzky put the puck into the net 87 times in 74 games. Therefore, he scored 87 goals. End of.

You're drawing an arbitrary line as to how much difference is too much difference in context.

I'm actually not. I'm saying, don't use adjusted stats for comparison. There's nothing arbitrary there.

 

[Iain Fyffe]

Now, this starts to make some sense. OK then, but why then do we not -- instead of tweaking numbers of goals from 45 to 29 when 45 pucks actually went in the net -- simply rate goals, assists, and points (and everything else) on a scale of 1 to 100?

You could certainly present the information that way, but that would just be a change in presentation rather than methodology. Based on H-R's calculations, Howie Morenz in 27/28 has the highest adjusted points total of all time, at 190. So if we make 190 the 100 on the scale, this means that Howe in 52/53 would be at 69 and Anderson in 82/83 would be at 43.

There's no need to pretend Gretzky scored 68 goals or whatever when he actually scored 87.

Despite your protestations to the contrary, this is exactly the misunderstanding I was referring to. Adjusted stats does not pretend that a player scored a different number of goals than he actually did. It translates the stat into a comparable level of production given certain contextual assumptions.

If I have 100 Canadian dollars in my pocket, that translates into 80 US dollars today. In no way does that mean I am pretending to actually have 80 US dollars in my pocket.

Well, I think this is a poor analogy because the Consumer Price Index is rating the overall value of a car, but how many points Anderson scored in 1983 is not an evaluation of his overall value.

Yes but actual scoring stats have exactly the same problem. In this context we're only discussing scoring stats, overall player valuation is another topic entirely.

That's pretty much what I already said in my previous post: "We could use stats to compare Glenn Anderson to Steve Larmer, but even then we need to consider the context."

No, you missed my point. You say it's valid to use stats to compare stats from the same season, but not from different eras. My point is that even players playing in the same season play under some very different contexts. As such, you believe it is possible to account for context when comparing player stats, and as such there is no reason why you could not account for this context across eras, unless you draw an arbitrary line across, saying "this much contextual different is too much, you just can't do it."

Stats are not sufficient as player-comparisons in themselves, as you point out. Therefore, why adjust them at all?

Why count them at all?

 

[Rhiessan71]

The problem with using the consumer index as comparable is that consumers are all equal.
There are not different tiers of consumers or consumers that are more "skilled" at spending money than other consumers.

So while the idea of AS's being about value is appropriate, the idea that that value is equal amongst all players is not.

As has been said before, top tier scoring has not dropped by the same level as tier 2 scoring, tier 2 scoring has not dropped by the same level that tier 3 scoring has and tier 3 scoring has not dropped by the same level as tier 4 scoring.

The overall size of the pie is smaller today but how that pie is divided up is very different with the Elite player taking a much bigger slice than they did in say 1985.

It's still "easier" for Greztky to get closer to 73 goals than it would be for Chris Nilan to get to 21.
According to AS's, the totals for both of these players goes to 58 goals for Gretzky and 17 for Nilan for a total of 19 goals lost between them.
The more likely outcome would be that Gretz only scores 65 while Nilan only garners 10. The same 19 goals are missing but the distribution of those missing 19 goals are not equal.
Just a rough example to make my point.

 

[Iain Fyffe]

The more likely outcome would be that Gretz only scores 65 while Nilan only garners 10. The same 19 goals are missing but the distribution of those missing 19 goals are not equal.
Just a rough example to make my point.

Yes, that's a valid criticism of the specific methodology used by Hockey-Reference and other to calculated adjusted stats. It may be more accurate to use a sliding scale rather than a multiplier.

 

[The Panther]

on H-R's calculations, Howie Morenz in 27/28 has the highest adjusted points total of all time, at 190. So if we make 190 the 100 on the scale, this means that Howe in 52/53 would be at 69 and Anderson in 82/83 would be at 43.

Perfect!

Adjusted stats does not pretend that a player scored a different number of goals than he actually did.

Yes it does, because in the box that says "goals" for Gretzky in 1984, it says on Adjusted Stats "69" not "87". So clearly it is pretending he scored a different number of goals than he actually did. Hence my suggestion of using another standard besides "goals/assists/points counted" as the variable in Adjusted Stats.

As such, you believe it is possible to account for context when comparing player stats, and as such there is no reason why you could not account for this context across eras

I'm not sure where you're getting this from, but I think I made it obvious in my (last two) posts that I do not believe this.
Please read my sentence again: "We could... BUT... (i.e., we can't)."

Why count them at all?

To count how many times the puck went into the net. That's all.

 

[Iain Fyffe]

Perfect!

Note that this changes absolutely nothing about the calculations. Just read the "Adjusted Points" column as being on a scale out of 190, and you get exactly the same thing.

To be clear: Adjusted Points are useless, Adjusted Points times 100/190 are not. This is what you're telling me.

Yes it does, because in the box that says "goals" for Gretzky in 1984, it says on Adjusted Stats "69" not "87".

Those are "Adjusted Goals", not goals. So no, what you're saying is not accurate. The definition of Adjusted Goals is not "the number of goals we pretend the player would have scored in another era." You're reading something that's not there.

To count how many times the puck went into the net. That's all.

But why? What's the point of counting them? Individual player goals, that is, not goals scored by the team. What does is matter which players are credited with scoring goals, so long as we count how many there were so that we can tell who won?

 

[Fish on The Sand]

What? He shattered bossy's old rookie goal record by over 20 goals.

His rookie year is considered an unbreakable feat.

http://www.nhl.com/ice/news.htm?id=534625

This thread isn't in the context of rookies though. It is in the context of all players. Selanne didn't have the most dominant season in 1993.

 

[Brooklanders*]

Sakic was never a dominate force like the aforementioned.

 

[The Panther]

To be clear: Adjusted Points are useless, Adjusted Points times 100/190 are not. This is what you're telling me.

What I'm telling you is that 'Adjusted Stats' are useless, but since certain fans seem to need them, I personally think it would be better (that doesn't mean I need it or that it's ideal) to have a comparative figure that doesn't impersonate 'goals scored in a season' -- which seems to lead people to silly conclusions based on 'adjusted stats'.

I don't need to know how many 'adjusted Home Runs' Babe Ruth hit in 1920. I know the 54 he hit was a s***load more than anybody else. I don't need to know Wilt Chamberlain's 'adjusted points-per-game' in 1962 -- I know his 50 pts.-per-game was way more than anybody else. Likewise, I don't need to know how many 'adjusted points' Wayne Gretzky scored between 1981 and 1987, because I know his 1219 points in this period was twice as many as the next guy.

What I'm interested in vis-a-vis stats is not comparisons across eras, but simply real, actual numbers measuring how many times something happened. I don't want or need those numbers adjusted. Apparently some people do -- that's fine. I don't.

I hope it is now clear what I'm saying. (Feel free to have the last word.)

 

[martz11]

Not to mention Lindros barely beat Jagr out for the Hart.

Jagr in 1998-99 was more dominant. He ran away with the Art Ross and Hart and won the Pearson as well.

127 Pts (82 evenstrength points in 81 games), first in assists, 2nd in goals.

Lindros had 10 of 15 first place votes. Jagr had 2 votes. How did he barely beat Jagr out?

 

[Iain Fyffe]

What I'm interested in vis-a-vis stats is not comparisons across eras

Well if you're not interested in that, there's nothing to say, since that's just personal preference. Plenty of people value cross-era comparisons, and there's nothing wrong with that either.

But your characterization that Adjusted Stats are "pretending" is inaccurate. And it's a common mischaracterization, I see it around here all the time. You can disagree with the value of doing something without misrepresenting the methods used to do so, and indeed your perception of the value of cross-era comparisons may be coloured by your apparent misunderstanding of this method. There's no pretending involved, it's merely a re-scaling of actual scoring stats.

 

[Brooklanders*]

What I'm telling you is that 'Adjusted Stats' are useless, but since certain fans seem to need them, I personally think it would be better (that doesn't mean I need it or that it's ideal) to have a comparative figure that doesn't impersonate 'goals scored in a season' -- which seems to lead people to silly conclusions based on 'adjusted stats'.

I don't need to know how many 'adjusted Home Runs' Babe Ruth hit in 1920. I know the 54 he hit was a s***load more than anybody else. I don't need to know Wilt Chamberlain's 'adjusted points-per-game' in 1962 -- I know his 50 pts.-per-game was way more than anybody else. Likewise, I don't need to know how many 'adjusted points' Wayne Gretzky scored between 1981 and 1987, because I know his 1219 points in this period was twice as many as the next guy.

What I'm interested in vis-a-vis stats is not comparisons across eras, but simply real, actual numbers measuring how many times something happened. I don't want or need those numbers adjusted. Apparently some people do -- that's fine. I don't.

I hope it is now clear what I'm saying. (Feel free to have the last word.)

I agree.

Adjusted stats aren't real stats and I hate when people bring them up.

Just compare players to their peers who played during the same era like you said.

Frankly who cares if Crosby would put up Sakic like numbers or better in that Era.
We will never truly know.
Just like how good is Gordie Howe or Wayne or Mario when we put them up against today's game.

Maybe they dominate like they did in their time wish I predict is likely but by how much?
Surely they wouldn't put up 200 point seasons.

 

[Iain Fyffe]

Just compare players to their peers who played during the same era like you said.

Adjusted stats do not change a player's rank amongst his peers. If Player A scored 10% more goals than Player B in the same season, then he will also have 10% more adjusted goals than Player B.

But I notice you said that you compare players who played during the same era, not the same season. So now, you have to define how you determine if two players played in the same era or not. How much difference in time is enough to create different eras? This is another arbitrary line: same eras okay, different eras not okay, without addressing the point that "era" is a vague term.

Frankly who cares if Crosby would put up Sakic like numbers or better in that Era.
We will never truly know.

That's not the question that adjusted stats are intended to answer. Some people certainly do use them that way, which isn't valid. But if someone uses a hammer to drive in a screw, where does the fault lie, with the person or with the hammer? We don't say hammers are useless because you can't use them to drive in screws.

 

[BenchBrawl]

Lafleur in 77 stands out but was his season really that dominant compared to what Marcel Dionne did in LA that year and with the type of support Marcel had?

Regardless of how you look at it Lafleur's 76-77 is one of the greatest season by a single player.He won the Hart,Smythe,Pearson and Ross.1st in playoff scoring.2nd in goal in the regular season.

Maybe Dionne could've done it with a better supporting cast, Lafleur actually did it.

 

[livewell68]

Lindros had 10 of 15 first place votes. Jagr had 2 votes. How did he barely beat Jagr out?

Jagr got all the 2nd place votes and the actual voting record was very close.

 

[Wrath]

Jagr got all the 2nd place votes and the actual voting record was very close.

http://www.hockey-reference.com/awards/voting-1995.html#hart

Lindros with 10 first place votes, 4 second place votes, 1 third place vote, total of 65 voting points.

Jagr with 2, 4, and 5 first, second and third place votes for a total of 27 voting points.

Hasek with 3, 2 and 2 first, second, and third place votes for a total of 23 voting points.

First place vote = 5 points, Second place vote = 3 points, Third place vote = 1 point.

 

[Voight]

Congrats, Lindros won the Hart in a lockout half season when Gretzky was past his prime and Lemieux didn't play. Big freaking accomplishment.

 

[Bear of Bad News]

Congrats, Lindros won the Hart in a lockout half season when Gretzky was past his prime and Lemieux didn't play. Big freaking accomplishment.

As easy as it apparently was, there were hundreds of NHL players who didn't accomplish it.

 

[Iain Fyffe]

As easy as it apparently was, there were hundreds of NHL players who didn't accomplish it.

C'mon Doc, you know how incredibly easy it is to win the Hart most seasons! I mean, just because the greatest players of all time won the scoring title just the season before and was still playing, that doesn't mean there was any sort of competition. And that chump Jagr, he was nothing, right?

 

[livewell68]

C'mon Doc, you know how incredibly easy it is to win the Hart most seasons! I mean, just because the greatest players of all time won the scoring title just the season before and was still playing, that doesn't mean there was any sort of competition. And that chump Jagr, he was nothing, right?

This was one of Jagr's weaker Art Ross wins. Jagr hadn't even hit his peak yet, it's quite possible that Lindros hit his peak that season.

Either way, he was barely better than Jagr was. When you are barely better than the second best player in the NHL (especially when that second best player has had far more dominant seasons), Lindros does not belong in this discussion.

Also Lindros' supporting cast, especially his own first line (the Legion of Doom line anyone), was very, very good; both his linemates were over PPG.

BTW based on the stats, there wasn't much that seperated Lindros and Jagr in 1994-95, of course I don't dare bring up the Canadian vs European bias do I when it came to Hart voting?

http://www.hockey-reference.com/leagues/NHL_1995_leaders.html