Comparing Players across Eras

[penbeast0]

Oh another one:

Pace is just another factor to think about across eras, and you shouldn't adjust for it without adjusting a slew of other things like field goal percentage.

So, you are saying Pace is the only factor to think about during eras? Or much more important than other changes? Not sure why you think this is a myth.

Pace is one factor. It is not the only factor. To compare players across eras I think you have to compare how they did relative to their competition and that means if you are going to adjust for pace, you should also adjust for offensive/defensive efficiency like fg%. (if, if player A, call him Bill Russell, is in the top 5 in the league in efg% with 45% in an era where league average is below 40% that is a better efficiency than if player B, call him Jermaine O'Neal, averages 45% in an era where efg% is closer to 50%

[Doctor MJ]

So, you are saying Pace is the only factor to think about during eras? Or much more important than other changes? Not sure why you think this is a myth.

Pace is one factor. It is not the only factor. To compare players across eras I think you have to compare how they did relative to their competition and that means if you are going to adjust for pace, you should also adjust for offensive/defensive efficiency like fg%. (if, if player A, call him Bill Russell, is in the top 5 in the league in efg% with 45% in an era where league average is below 40% that is a better efficiency than if player B, call him Jermaine O'Neal, averages 45% in an era where efg% is closer to 50%

Not saying it's the only factor, but it's something that is a relatively easy adjustment that absolutely must be done. There's a contingent who honestly thinks that it's a failing that more modern players aren't playing as fast, and they've got it totally backwards. The goal has always been to do more per possession than your opponent, and whatever speed lets you do that. To begin even the most preliminary analyses without factoring in pace is a mistake.

Adjusting for field goal percentage on the other hand is a far more tricky issue. While it's true that one can adjust % based on standing relative to contemporaries, to do this without having a clear reason for doing so is completely pointless. Thus, while either that straight forward adjustment or a more complex one may end up done in an analysis, it's not the same as the pace adjustment.

[Vinsanity420]

Pace? No need to discredit Oscar for that. How would Lebron fare without crab dribbling, open lanes, rules that allowed players to play extremely physical defense? Without players getting star calls? Or the fact that assists weren't awarded as easily as they are today?

Take pace into account, but do take also take into account how vastly different 60's ball is from 00's ball.

[Doctor MJ]

Pace? No need to discredit Oscar for that. How would Lebron fare without crab dribbling, open lanes, rules that allowed players to play extremely physical defense? Without players getting star calls? Or the fact that assists weren't awarded as easily as they are today?

Take pace into account, but do take also take into account how vastly different 60's ball is from 00's ball.

Well this is what I'm talking about, and it's clearly a pretty common opinion.

The whole point is not to discredit, it's to get more accurate. Refusing to do something incredibly easy to do, and just glaringly correct, because some other really hard thing to do correctly isn't done is foolishness. I'm not even saying we shouldn't take other things into account, but pace is a pure numerical thing - and "extreme physical defense" absolutely is not.

[penbeast0]

However relative efficiency (as opposed to absolute efficiency), the thing you particularly mentioned, is just as easy to calculate or see intuitively as pace.

It's also just as era dependent as pace. Unlike some of the clueless teens who think there has been an evolutionary shift in the human genome in the last 50 years, most intelligent people realize that the apparent greater athleticism of the modern player is, in fact, due almost completely to environmental factors such as better training, health care, equipment, etc. (not to mention steroids again). Well, so is league wide efficiency at least from the point where they opened the league up racially. Different shooting styles, rules, weight work, etc. made it much harder to shoot high percentages in the very early 60s just like the higher pace and lower efficiency increased raw rebounding.

You can only play the teams you face and you can only play by and with the rules/era differences that were around in your time. Yes you have to adjust for pace but adjusting for relative efficiency is just as easy (easier actually since you only need to adjust one number not several) and failing to do so is just as prejudicial to certain players -- just in this case you are favoring the more modern players rather than the ealier players. Very surprised that you are falling into this fallacy Doc.

[Doctor MJ]

However relative efficiency (as opposed to absolute efficiency), the thing you particularly mentioned, is just as easy to calculate or see intuitively as pace.

It's also just as era dependent as pace. Unlike some of the clueless teens who think there has been an evolutionary shift in the human genome in the last 50 years, most intelligent people realize that the apparent greater athleticism of the modern player is, in fact, due almost completely to environmental factors such as better training, health care, equipment, etc. (not to mention steroids again). Well, so is league wide efficiency at least from the point where they opened the league up racially. Different shooting styles, rules, weight work, etc. made it much harder to shoot high percentages in the very early 60s just like the higher pace and lower efficiency increased raw rebounding.

You can only play the teams you face and you can only play by and with the rules/era differences that were around in your time. Yes you have to adjust for pace but adjusting for relative efficiency is just as easy (easier actually since you only need to adjust one number not several) and failing to do so is just as prejudicial to certain players -- just in this case you are favoring the more modern players rather than the ealier players. Very surprised that you are falling into this fallacy Doc.

It's not that I don't look at context at all to evaluate a player, but when people adjust efficiency before applying thought they lose a ton of information. Guys in the 40s shot 25% on average, but a few guys shot more like 40%, much further above the average than anyone is now. Does that make those guys the best shooters/players in history? Of course not. One shouldn't assume they couldn't have done better with later strategies and techniques, but he certainly wouldn't have shot 90% in today's league.

More practically, I see people dismiss Wilt's efficiency issues based on era, and that's wrong - just as it would be wrong to completely ignore the era in which he did it. The Wilt-West-Kareem comparison, where West played through both guys' volume scoring periods while himself shooting similar volume and efficiency the whole way through, really makes clear how a simplistic statistical adjustment won't work.

[penbeast0]

Yes, you can't just make simplistic adjustments and expect accuracy . . . but that's just as true of pace as it is of relative fg efficiency. Why is pace a more useful adjustment than relative efficiency which is after all what you seemed to claim in your first post?

[Doctor MJ]

Yes, you can't just make simplistic adjustments and expect accuracy . . . but that's just as true of pace as it is of relative fg efficiency. Why is pace a more useful adjustment than relative efficiency which is after all what you seemed to claim in your first post?

The goal of basketball is to score more points per possession than your opponent. Whatever pace you choose to play at to accomplish that is as good as any other pace - so it's to treat entire leagues that played at faster pace as superior simply because of that pace. By contrast, efficiencies aren't all the same - in fact they're pretty much definition of what's better and worse. Adjust for pace, then take the tougher to determine stuff into account.

[Doctor MJ]

Sorry, I don't understand your point here. (a)Pace is very relative to player/team styles and varies both team to team and year to year . . . some players play better in open court and some have advantages in half court. (b) relative efficiency isn't tougher to determine than pace; it's easier (mathematically anyway -- not sure if you mean something else since this makes no sense to me).

I'll try to be more clear. I'll focus on the efficiency adjustment since that seems what we disagree on.

Really, the crux of it is that you handed me a single "era adjusted efficiency" of two guys and said "So player X was more efficient", I'm not going to be satisfied with that, and I kind of doubt you would be either. That's not to say I don't do something like an adjustment in my head when doing a comparison between two players - but it's too fuzzy a process to fully codify. That's not to say we shouldn't use a relative efficiency statistic, just that it's got to be used to augment what we have, not replace the raw data. Whereas, I actually want to effectively replace the unadjusted volume stats with pace adjusted ones.

Something else I'll mention: If in an all-time ranking type context, you want to go solely by how guys did relative to their peers, I'm not saying that's wrong. However, I'm also not going to say it's necessarily wrong for someone to say that the shooters of today are light years better than the shooters of 60 years ago because of the numbers they see. There is a range of reasonable ways to look at things - the key point is that it's never reasonable to think that players of today are doing something inherently wrong because they aren't looking to shoot the basketball as quickly as players in other eras.

[penbeast0]

I think you can say "pace adjusted, X scored/rebounded etc. at a higher rate than Y"

I don't think it's any more reasonable than saying "era adjusted, X scored at a higher efficiency than Y"

You seem to. But I can't see any valid reason for using one and ignoring the other.

I also use other quick adjustment figures where I can quantify them easily. I use an adjustment to pace differentials for the change in the assist rule. I discount 3 point shooting a bit during the era of the shorter 3 point line. I think you use the adjustments that you can quantify and that seem reliable to the best extent possible.

I even posted an ABA/NBA scoring and rebounding volume conversion chart in the RPOY thread . . . not that it's perfect but it gives at least a first cut at estimating the statistical difference betweent the two leagues for the various years.

[kabstah]

I think you can say "pace adjusted, X scored/rebounded etc. at a higher rate than Y"

I don't think it's any more reasonable than saying "era adjusted, X scored at a higher efficiency than Y"

You seem to. But I can't see any valid reason for using one and ignoring the other.

I also use other quick adjustment figures where I can quantify them easily. I use an adjustment to pace differentials for the change in the assist rule. I discount 3 point shooting a bit during the era of the shorter 3 point line. I think you use the adjustments that you can quantify and that seem reliable to the best extent possible.

I even posted an ABA/NBA scoring and rebounding volume conversion chart in the RPOY thread . . . not that it's perfect but it gives at least a first cut at estimating the statistical difference betweent the two leagues for the various years.

IMO the reason is obvious. Pace is often a matter of playing style and it's not necessarily beneficial to play at the highest possible pace. On the other hand, it is always always always beneficial to score as efficiently as possible. For example, Bill Russell and the 60's Celtics could've played at a faster or slower pace had they chosen to. They couldn't, however, arbitrarily raise their shooting percentages -- otherwise they would've done so and been a better team for it.

In other words, adjusting for pace is normalizing for stylistic differences. Adjusting for efficiency is normalizing for actual shortcomings in ability. The former makes sense when comparing players and teams, the latter does not.

[Doctor MJ]

IMO the reason is obvious. Pace is often a matter of playing style and it's not necessarily beneficial to play at the highest possible pace. On the other hand, it is always always always beneficial to score as efficiently as possible. For example, Bill Russell and the 60's Celtics could've played at a faster or slower pace had they chosen to. They couldn't, however, arbitrarily raise their shooting percentages -- otherwise they would've done so and been a better team for it.

In other words, adjusting for pace is normalizing for stylistic differences. Adjusting for efficiency is normalizing for actual shortcomings in ability. The former makes sense when comparing players and teams, the latter does not.

Thank you kabstah. I seem to be having trouble writing coherently this weekend - you said it well.

[Vinsanity420]

In other words, adjusting for pace is normalizing for stylistic differences. Adjusting for efficiency is normalizing for actual shortcomings in ability. The former makes sense when comparing players and teams, the latter does not.

Well said.

In general, teams used to RUN in the 60s... and its simply because there was more offensive talent on each team. Back then there were around 8 teams... now there are 30. Since there is less O talent, teams figured slowing it down and looking for the best shot possible was the way to go.

[penbeast0]

And the choice of running or walking it up, that doesn't affect player efficiency somehow to say nothing of the many other coaching decisions about where to get the ball to players, who posts and who plays on the perimeter, etc.? I think you are drawing false distinctions.

Take Bill Russell -- the classic case for using relative fg%. He joined a Celtic team where the bulk of the scoring was done by jump shooters with excellent range (Cousy, Sharman, Heinsohn, even Ramsey). With those players, Red had him playing low post and he was in the top 5 in the league in FG% for his first few years in the league -- of course that was only shooting 45% but in a league where the average fg% was under 40%.

Then the Celtics turned over their personnel with the above players being replaced by players who liked to work much closer to the basket -- Bailey Howell (postup), Havlicek and Sam Jones (midrange slashers). They moved Russell out to the high post and made him a passing hub. He didn't have a good shot from there (poor FT shooter) and as the league's shooting efficiency went up 5-10%, his actually dropped 1-2%. But the real efficiency drop was much greater than that as the league was now shooting 45%+ with low post centers even higher.

Ignoring the league change masks the huge damage done to Russell's offensive game by this change -- just as ignoring the monster pace in the early 60s inflates his numbers. Same goes for guys like Elgin Baylor or Bob Pettit who were among the league's all-time greats in the late 50/early 60s but get badly underrated because of a failure to adjust for leaguewide efficiency changes.

[Doctor MJ]

In other words, adjusting for pace is normalizing for stylistic differences. Adjusting for efficiency is normalizing for actual shortcomings in ability. The former makes sense when comparing players and teams, the latter does not.

Well said.

In general, teams used to RUN in the 60s... and its simply because there was more offensive talent on each team. Back then there were around 8 teams... now there are 30. Since there is less O talent, teams figured slowing it down and looking for the best shot possible was the way to go.

Wait, no that's completely wrong. Regardless of the talent level at any particular time, there is never any reason to NOT figure out the best shot possible. And it's not like efficiency stayed constant over time (which would fit with the idea that earlier eras saw no reason to slow it down) - efficiency's gone way up.

[Doctor MJ]

And the choice of running or walking it up, that doesn't affect player efficiency somehow to say nothing of the many other coaching decisions about where to get the ball to players, who posts and who plays on the perimeter, etc.? I think you are drawing false distinctions.

Take Bill Russell -- the classic case for using relative fg%. He joined a Celtic team where the bulk of the scoring was done by jump shooters with excellent range (Cousy, Sharman, Heinsohn, even Ramsey). With those players, Red had him playing low post and he was in the top 5 in the league in FG% for his first few years in the league -- of course that was only shooting 45% but in a league where the average fg% was under 40%.

Then the Celtics turned over their personnel with the above players being replaced by players who liked to work much closer to the basket -- Bailey Howell (postup), Havlicek and Sam Jones (midrange slashers). They moved Russell out to the high post and made him a passing hub. He didn't have a good shot from there (poor FT shooter) and as the league's shooting efficiency went up 5-10%, his actually dropped 1-2%. But the real efficiency drop was much greater than that as the league was now shooting 45%+ with low post centers even higher.

Ignoring the league change masks the huge damage done to Russell's offensive game by this change -- just as ignoring the monster pace in the early 60s inflates his numbers. Same goes for guys like Elgin Baylor or Bob Pettit who were among the league's all-time greats in the late 50/early 60s but get badly underrated because of a failure to adjust for leaguewide efficiency changes.

beast, when you objected to what I said earlier, Russell came to mind and I actually think he's a good example to think on this from.

The thing is - doing the efficiency adjustment doesn't help him all that much for the reasons you specified. The fact of the matter is that for a large chunk of his career, he wasn't very impressive efficiency wise even relative to his peers. The defense of Russell you gave is right on point, but it's not something that is really compatible with an Adjust'em All!!! approach. This is part of why adjusting for pace by simple math is a definite yes, and doing it for other things is not necessarily as clear cut.

Re: Baylor, Pettit. This is something where we probably have a legit disagreement. Baylor and Jerry West came into the league two years apart. Baylor's efficiency is terrible by modern standards, while West's is basically right with modern standards. Dismissing Baylor's efficiency issues based on his era, only makes sense to me if we then give West an effectively comparable boost - which puts him at a jaw dropping level relative to modern standards once he hits his prime. Then the league keeps getting more efficient, and West basically stands still, and by pure adjustment logic is getting worse when in reality he's just staying the same.

I can understand the thought process that says that player's ability to score with efficiency is a product of when they start, and so if you truly want to know how impressive a guy is, you've got to adjust accordingly. But literally here, rigid adherence to this logic puts us in the situation where a guy is worse than a guy who is 5 years younger than him, and we pretend he isn't.

[penbeast0]

Most players have their best statistical season before age 25 (though not necessarily their best season) and as for Russell:

Your way, there is no indication that he moved to the high post or was appreciably less efficient there. My way gives a much better snapshot of his actual career arc. It's like ignoring games played for a guy like Yao.

[kabstah]

penbeast-

Are you saying that Russell was more efficient than a modern player like Jermaine O'Neil who shot roughly the same % from the floor? Shooting 45% might have been good for top 5 in the league in 1960, but in the modern era we have low post players who regularly shoot in the mid 50's or higher while being the primary option on offense (Dwight, Yao, Stoudemire, Gasol, and if you go back a few years Shaq). Do you think that if you replaced Dwight on the 2010 Magic with 1960 Russell, that Russell could replicate Dwight's numbers on offense?

I can see the case for adjusting perimeter players' efficiencies if they played before the advent of the 3 point line, but that doesn't apply to guys like Russell or Petit.

[penbeast0]

Simple adjusted numbers for 1960 Russell -- he starts at 18.2pts on .467 shooting (might be his best offensive season btw). Average team scored 115.3ppg and shot .410 from the field (no 3 point line so efg is .410 too). In 2010, the average team scored only 104.0 ppg on .461 from the field, but with the prevalence of the 3 point shot, league-wide efg% was a spectacular .501. Russell's rough equivalence would be 16.4ppg on .571 efg.

And remember, a player shooting .467 in 1960 is a high percenage player who you want to get more shots, a player shoooting an efg of .467 in 2010 is an extremely poor percentage player in your lineup (ignoring foul draw and other similar factors). How can these be equivlanet?

More efficient than Jermaine O'Neal . . . hell yes, though if you adjust for his whole career instead of just one of his best years Russ still comes off pretty mediocre offensively.

More efficient than Dwight (18.2 on .613fg%) . . . no, Russell's best year was barely better than Dwight at age 20 and significantly worse than every year since; and yet people slam Dwight here for his "weak" offensive game.

[Doctor MJ]

Most players have their best statistical season before age 25 (though not necessarily their best season) and as for Russell:

Your way, there is no indication that he moved to the high post or was appreciably less efficient there. My way gives a much better snapshot of his actual career arc. It's like ignoring games played for a guy like Yao.

Honestly not sure exactly what your point is about "before age 25".

Re: Russell. I don't think either approach, yours or mine, really has a case for really explaining Russell. As you say, players typically have their best statistical season at a young age, so seeing a drop of after Russell turns 27 could easily lead someone to the incorrect conclusion that he was simply following a normal curve.