RealGM 2017 Top 100 List #56 (Willis Reed)

Moderators: Doctor MJ, trex_8063, penbeast0, PaulieWal, Clyde Frazier

trex_8063
Forum Mod
Forum Mod
Posts: 12,684
And1: 8,322
Joined: Feb 24, 2013
     

Re: RealGM 2017 Top 100 List #56: RUNOFF! Reed vs Harden 

Post#41 » by trex_8063 » Sat Oct 28, 2017 8:50 pm

Owly wrote:My questions would be as follows then:
Was your interjection (post 34) in defense or labelling '12 non-prime, in support of Dr P's arguments or both?


Mostly it was in support of labelling '12 non-prime, though I suppose it's marginally in the same vein as Dr P's arguments about portability, too.


Owly wrote:Have you looked at converting scaled WS/48 back to wins? Do they total to the correct number of available wins. If so was the correlation with team performance (mainly points dif) better or worse than conventional Win Shares?

Fwiw, I (personally) would want the answers to be yes, yes and "it was better" before I'd use "Scaled Win Shares" over Win Shares.


No, actually I haven't looked at that, but now I know what my next statistical project is going to be.
I suspect it will come out at least as good as regular Win Shares. When deducing the standard deviations each year, I wasn't just sampling data points ABOVE the mean, I was also sampling from BELOW the mean. While doing so I noted that in most of those "greater parity" seasons it wasn't just that players weren't distancing themselves above the mean (in raw terms) as much as we've seen in recent years.......there was also less deviation below the mean. At any rate, the scaling within a given year will apply equally to those above the mean and those below it, so I imagine it will be nearly identical to the correlation of regular WS.

I'll have to get back to you at a later date about that....
"The fact that a proposition is absurd has never hindered those who wish to believe it." -Edward Rutherfurd
"Those who can make you believe absurdities, can make you commit atrocities." - Voltaire
iggymcfrack
RealGM
Posts: 11,999
And1: 9,454
Joined: Sep 26, 2017

Re: RealGM 2017 Top 100 List #56: RUNOFF! Reed vs Harden 

Post#42 » by iggymcfrack » Sat Oct 28, 2017 11:12 pm

Owly wrote:
trex_8063 wrote:
Owly wrote:1) He's 56th in total minutes, 71st in mpg. He's playing above median starter minutes. A bit less than you'd like from a star. But is that "a reduced minute role".


Is it a "reduced minute role" may be debatable from a semantic standpoint, but for me you hit on the gist of it with the underlined portion. That it's significantly reduced vs every other year of his prime is the other consideration that makes me label it as such.


Owly wrote:2) How does "a reduced minute role" mean that he wasn't fitting well with Westbrook (and per my later point, if you believe that to be the case, how amazing must he have been when Westbrook was off court).


He may well have been remarkable (and up to the standards of the rest of his prime) during those times Westy was off the court; but if that's the case, within the context of talking specifically about portability, that wouldn't really help him.

Owly wrote:3) But why would you scale it? How would you scale it in a manner that remains fair (not an attack here, what's the methodology)?


Same basic methodology as Doc used for his Scaled RAPM sheets: it's based on standard deviations from the mean, and noting that the standard deviation [in PER and WS/48] isn't the same year-to-year; and that in some years/eras there appears to be a lot more parity (smaller standard deviation)---->that being my motive to scale it.

Here is the thread for the rs numbers. Here is the thread for playoff numbers. Neither has been updated to include the '17 season, btw.

My questions would be as follows then:
Was your interjection (post 34) in defense or labelling '12 non-prime, in support of Dr P's arguments or both?
Have you looked at converting scaled WS/48 back to wins? Do they total to the correct number of available wins. If so was the correlation with team performance (mainly points dif) better or worse than conventional Win Shares?

Fwiw, I (personally) would want the answers to be yes, yes and "it was better" before I'd use "Scaled Win Shares" over Win Shares. Ditto for scaled versions of PER/EWA, though that would be messier (EWA is wins over replacement - with replacement at circa 10 teams wins OTOH, but would have to look up - and Hollinger used different replacement levels at different positions, though if one doesn't mind the marginal redistrubution of credit, using an average of them wouldn't affect it's use when done at the team level, as each team notionally plays 48 minutes of each position in each game).

Fwiw, as I said I kind of get PER and WS comparisons across seasons not feeling perfect. I think we've discussed the disparity in the range of outcomes before. But as stated before (in this thread) I think the formulas were done as they were for reasons, and so, per the immediately above, I'd want to see that the scaled numbers were "better" to justify the tinkering.

In terms of response on the Westbrook on/off portability issue (and at this point I really wish whatever happened to photobucket hadn't happened and LA Bird's NBAWOWY Venn Diagrams were still up, so I/we could just look at it) I'm not wedded to the idea Harden is massively portable, so much as the idea as I've put it above, i.e. if he can have the RAPM, the On/Off type numbers he did, for him to be below average to poor portability that just means blowing the roof off the positive impact when put in optimal circumstances. And where optimal circumstances probably means playing with "only" one in-prime* player (all wing-scorers) who by themselves (or not in Durant's case) will go on to be able to put up MVP numbers/seasons, rather than two, I think that it's harsh to penalize Harden for, given how rare that circumstance is. To be honest, fwiw, in the broader discussion of player goodness, and player goodness is the ultimate point here, I'd probably rather have the higher upside guy than the one more able to blend into such an unlikely (and difficult) scenario.

* fwiw, prime used here is maybe contentious for Westbrook. I'm using it here to indicate an absolute standard of performance, Westbrook is north of (roughly) 20 PER and .150 WS/48, point being, this isn't "Magic's getting to play with McAdoo". No it's not peak Westbrook, but it is a very good (and quite ball-dominant) player. I can see not putting this as prime - perhaps especially for you where higher peaks mean a higher cut-off for what's included in prime - but you get my point.


I feel the same way about scaled PER. PER is a metric that's already done relative to league average. If fewer players are excelling far above average due to an even split in roles at a given time, I don't see why that makes what the players who did take on heavy roles for their team any easier of a task.

Furthermore, this seems like a stat that's inevitably used to prop up older players from a much weaker era, and it seems like any adjustment for it being harder to excel in a league where the roles are split evenly should be counteracted x10 for the fact that the average player in the 70s was much worse at basketball than the average player now. If two people exceeded the average by the same amount, one in 1972 and one today, I think it takes some real mental gymnastics to think the task accomplished by the older player is the one that represents more skill at basketball.
trex_8063
Forum Mod
Forum Mod
Posts: 12,684
And1: 8,322
Joined: Feb 24, 2013
     

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#43 » by trex_8063 » Sun Oct 29, 2017 8:23 pm

iggymcfrack wrote:I feel the same way about scaled PER. PER is a metric that's already done relative to league average. If fewer players are excelling far above average due to an even split in roles at a given time, I don't see why that makes what the players who did take on heavy roles for their team any easier of a task.


It's not an "easier" task; that's not what I'm saying. It's about not holding circumstances against one group. As you say in the underlined portion: if there were circumstances that prevent or disallow the same kind of primacy that stars enjoy in more disparate eras, that should not be held against them. And there almost surely was something about the circumstances contributing to this. I don't know how we can look at such wildly differing historical trends, and suggest that there wasn't.


I can only speculate on all the factors that contribute to this more "even split in roles", but one which is certainly a contributing factor is pace. Over a certain league avg pace threshold, faster pace [relative to league avg] has a negative correlation with team ORtg. This isn't speculation; I've done the studies, this appears concrete. I'm not exactly sure where that threshold is, but I've looked at two arbitrarily-selected thresholds of all seasons where league avg pace is >107 and years where it's >115. The correlation is already emerging in the >107 grouping, is more apparent/definitive in the >115 grouping.
I can talk to you about the appearance of trend lines and correlation coefficients, but if that's not your cup of tea, some general observations within the data will likely be illuminating:

Within the years of >115 league avg pace (145 data points or teams in this sample)....
*There were 19 teams with a rORTG of +3.0 or better........only ONE out of those 19 had a rPace >0 (+1.3).
**There were 4 teams with a rORTG of -5.0 or worse.......THREE of the 4 had a rPace that was not only greater than avg, but >+2.0 rPace; the other was basically right at avg with -0.1 rPace.
***Of the 13 fastest paced teams in this entire sample, ALL THIRTEEN had a rORTG <0.
****Of the 7 slowest paced teams in this sample, ALL SEVEN had a rORTG >0.


Bottom line is that over a certain pace threshold (I suspect somewhere around 105), teams are actively engaging in a certain degree of "offensive indiscretion" (which would include not always getting shots from their best options) to further increase the pace. And this would, in part, account for why we don't see [in these eras] as wide a margin from the low-end players either. Because it's not just about the guys on the top-side of average who aren't distancing themselves as far from the mean; BELOW average players are also not distancing themselves from the mean as much as the below avg players in other years (likely because they're getting more touches in these circumstances of "offensive indiscretion").

And where Reed specifically is concerned, league avg pace was >107 every year of his career, >115 in seven of the 10 (twice >120).


iggymcfrack wrote:Furthermore, this seems like a stat that's inevitably used to prop up older players from a much weaker era.....


No, this was not the direct intent (even if it appears to be the result). It was initiated in the spirit of the purple statement above, and meant to level the field slightly for circumstance if definitive trends exist....which they did. And then---once establishing that trends exist----hopefully get myself and others to think about why they exist.
"Truth is hidden only to the masses who are not inquisitive."


iggymcfrack wrote:......and it seems like any adjustment for it being harder to excel in a league where the roles are split evenly should be counteracted x10 for the fact that the average player in the 70s was much worse at basketball than the average player now. If two people exceeded the average by the same amount, one in 1972 and one today, I think it takes some real mental gymnastics to think the task accomplished by the older player is the one that represents more skill at basketball.


This too I have not directly said, nor have I suggested all eras should be considered equal in level of competitiveness.

That said, I feel you take opinions of past eras to extemes. A modicum of discounting of some eras is likely appropriate; near-total disregard is inappropriate.
I'd like to walk thru and draw you out in a more Socratic manner as to why you feel this way, but it just takes too long; and I’ve had enough of these conversations that I already know all the reasons….

Modern players are taller…..
Negligible. The average height of NBA players has increased by ~1” or less in the last half-century. And frankly I wouldn’t be surprised if that 1” change is more the result of factors like not being raised in an era where practically every adult smoked (possibly stunting growth of some individuals), or improved perinatal nutrition.


.....and bigger/stronger/faster….
Modern players have many more resources (by way of equipment, facilities, trainers and improved training methods, dieticians, etc) available to them than past players did. Additionally, physical training is now geared toward strength, quickness, explosiveness, leaping ability, etc [i.e. the very qualities that seem improved]; whereas it was once geared largely toward just cardio.
At one point, weight training was actively discouraged among basketball players, on the theory that extra weight being carried would cause players to fatigue quicker, and also that increased upper body mass/strength would [negatively] effect one’s shot. The latter myth was STILL sporadically present even around the time I was in highschool in the 1990’s.

Footwear (and floor conditions) are other underappreciated aspects of the quickness we perceive in past-era players.
Have you ever played on cruddy, poorly-kept courts? You can’t always explode laterally or change direction quite like you’d like to. The vectors of force you apply with your legs to move need to be slightly less lateral and slightly more vertical (to avoid slipping). Now let’s further complicate things by not providing you with sturdy footwear with good ankle, adequate cushioning, or a nice contoured sole…..instead, you get to wear something like the old-school Converse All-Star: a canvas shoe with a thin, FLAT sole and minimal ankle support. How are you inclined to move now? Probably slightly more cautiously. I don’t see Russell Westbrook, for example, flying around at a million mph in these conditions like he normally does (or I see him busting an ankle fairly early on if he tried).

Explosive moves with the dribble have a whole other layer of difficulty associated with them (more on that below)....


Past players dribble funny, like they barely know how…..
As an experiment, next time you’re in the gym, attempt to make an explosive dribble-drive move to the basket, or perhaps a sudden change of direction [with the dribble], or even simply dribbling in a straight line at a full sprint….but WITHOUT putting your hand under or even on the side of the ball.
If you’re doing it “legally” by the rules of the 50’s/60’s (and to a lesser degree the early 70’s), you likely now look very similar to how they did. They didn’t dribble that way because they couldn’t figure out how to do it better; they dribbled that way because that was what was allowed at the time. Players of the early 1960’s, for example, would have about as much use for a modern cross-over or spin move as they would for learning to dribble with their butt cheeks.
People like to think someone like Kyrie or Curry would be doing all kinds of fancy dribbles all over everyone in past eras. No. If raised in that era, they’d be dribbling like everyone else. At best they could be pioneers pushing the envelope like a Cousy or a Bob Davies in the 1950’s, or like a Frazier or Monroe in the late 60’s/early 70’s…..because that’s the limit of what was allowed at the time(s).


Past players often have goofy shooting form….
Again, hate to break it to you, but if reared in the same environment (with the same coaching, mentoring, visual influences to model your game from, etc), today’s players would have had shots that looked basically the same. Mechanics have improved, no question. But these kinds of changes happen(ed) slowly over time, just like improvements in nearly any other field (particularly pre-Information Age).


Past players seem to poorer bball IQ’s (slower on rotations, poorer shot selection, etc)....
True, but not because they’re stupid individuals. Perhaps almost NO ONE 50 years ago had the basketball IQ that basically any assistant coach in the league has today. Game theory and analytics have brought us an awful long way, and provided a ton of nuance, which professional teams attempt to be cognizant of it ALL. This too is a function of the slow progression/development of game analytics and the gradual dissemination of information, theory, and practice. Once again, if past players were availed some of this which is now commonplace knowledge among much of the professional basketball community, most would have drastic spikes in their bball IQ (this isn’t rocket-science, after all; it’s a game).
Likewise, modern players, if reared in a different time period would have been just as ignorant [of developments which had not yet occurred] as past players.


In short, precious little of the differences you perceive between the old game and the new have anything to do with the players as individuals, and have A LOT more to do with circumstance. Criticizing a past player for not having a modern quality to his game makes about as much sense as declaring Thomas Edison was a moron because he wouldn’t know how to use a smart phone.

Humans haven’t inherently changed in a span of 2-3 generations. Any actual physical changes are the result of environmental influences; you level the playing field for environment, and you see the players of then and now are very very similar.
Just about the only truly valid argument pertaining to strength of era stems directly from size of player pool. Just about everything else is environment/circumstance.
"The fact that a proposition is absurd has never hindered those who wish to believe it." -Edward Rutherfurd
"Those who can make you believe absurdities, can make you commit atrocities." - Voltaire
Frank Mathis
Ballboy
Posts: 6
And1: 3
Joined: Oct 29, 2017

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#44 » by Frank Mathis » Sun Oct 29, 2017 11:11 pm

only 7 votes because the moderators kick everyone off and aren't inclusive of anyone.
User avatar
Clyde Frazier
Forum Mod
Forum Mod
Posts: 20,241
And1: 26,118
Joined: Sep 07, 2010

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#45 » by Clyde Frazier » Mon Oct 30, 2017 12:30 am

Frank Mathis wrote:only 7 votes because the moderators kick everyone off and aren't inclusive of anyone.


This isn't even remotely true.
Frank Mathis
Ballboy
Posts: 6
And1: 3
Joined: Oct 29, 2017

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#46 » by Frank Mathis » Mon Oct 30, 2017 12:51 am

It's true. You have some offish and nongregarious on this board Clyde. You're not one of them but the others are.
User avatar
PaulieWal
Forum Mod
Forum Mod
Posts: 13,909
And1: 16,218
Joined: Aug 28, 2013

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#47 » by PaulieWal » Mon Oct 30, 2017 12:58 am

Frank Mathis wrote:It's true. You have some uptight moderators on this board Clyde. You're not one of them but the others are.


You have a great read on the mod team for a "new member" :D . ANyway, attacking other members on the board (mods or not) is not acceptable. Consider this a verbal warning.
JordansBulls wrote:The Warriors are basically a good college team until they meet a team with bigs in the NBA.
Frank Mathis
Ballboy
Posts: 6
And1: 3
Joined: Oct 29, 2017

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#48 » by Frank Mathis » Mon Oct 30, 2017 12:59 am

See what I mean, They can't take a little criticism.
User avatar
PaulieWal
Forum Mod
Forum Mod
Posts: 13,909
And1: 16,218
Joined: Aug 28, 2013

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#49 » by PaulieWal » Mon Oct 30, 2017 1:01 am

Frank Mathis wrote:See what I mean, They can't take a little criticism.


If you have issues with the mod team, PM anyone of us. Publicly criticizing and attacking others members is not tolerated from any member.
JordansBulls wrote:The Warriors are basically a good college team until they meet a team with bigs in the NBA.
Frank Mathis
Ballboy
Posts: 6
And1: 3
Joined: Oct 29, 2017

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#50 » by Frank Mathis » Mon Oct 30, 2017 1:04 am

Why do I get attacked all the time on here. Also this site should have over 2,000,000 users everyday. but it's dead and isn't making 1/100th the money it should be.
Last edited by Clyde Frazier on Mon Oct 30, 2017 1:07 am, edited 1 time in total.
Reason: Locking post, basically admitted to being a PBP
Frank Mathis
Ballboy
Posts: 6
And1: 3
Joined: Oct 29, 2017

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#51 » by Frank Mathis » Mon Oct 30, 2017 1:05 am

You laugh at new members? That's not very inclusive.
User avatar
PaulieWal
Forum Mod
Forum Mod
Posts: 13,909
And1: 16,218
Joined: Aug 28, 2013

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#52 » by PaulieWal » Mon Oct 30, 2017 1:06 am

Frank Mathis wrote:Why do I get attacked all the time on here. Also this site should have over 2,000,000 users everyday. but it's dead and isn't making 1/100th the money it should be.


Please provide examples of where you got attacked.
JordansBulls wrote:The Warriors are basically a good college team until they meet a team with bigs in the NBA.
iggymcfrack
RealGM
Posts: 11,999
And1: 9,454
Joined: Sep 26, 2017

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#53 » by iggymcfrack » Mon Oct 30, 2017 2:01 am

trex_8063 wrote:In short, precious little of the differences you perceive between the old game and the new have anything to do with the players as individuals, and have A LOT more to do with circumstance. Criticizing past players for not having a modern game makes about as much declaring Thomas Edison was a moron because he wouldn’t know how to use a smart phone.

Humans haven’t inherently changed in a span of 2-3 generations. Any actual physical changes are the result of environmental influences; you level the playing field for environment, and you see the players of then and now are very very similar.
Just about the only truly valid argument pertaining to strength of era stems directly from size of player pool. Just about everything else is environment/circumstance.


Read the whole post, (just didn't want to quote it all) and there were some interesting points there, specifically about the pace. That players playing a heavy minute load at a fast pace will be unable to put up the same kind of per-minute stats as players playing less minutes at a slower pace. I would actually prefer to see something specifically tied to pace and minute load than just seeing standard deviations of top and bottom to league average though which seems like a very sloppy rough solution to a very specific problem. For instance, Steve Nash's 7 seconds or less Suns team played at a faster pace than some of the slower paced teams in the 70s and 80s, but he's still going to face the same massive scaled PER hit as someone who had lots of rest at a leisurely pace. Is that fair? Is it fair to penalize Iverson relative to Jerry West when Iverson's playing 44 MPG against opposition starters averaging in the mid-30s while West is up against starters playing the same minute load he is?

I also still think you're underestimating some of the factors for why modern players have it more difficult against better competition. For instance, you said that the only valid reason that older players would perform worse in the modern era if they played now is due to the smaller player pool. This isn't really true. The fact is that especially in the 60s, and to some extent all the way up until Magic and Bird, basketball just wasn't that popular of a sport. There may well have been a lot of American players with the potential to be all-NBA talent that just didn't bother because it didn't seem intriguing enough for them with a professional league that wasn't televised live with relatively low salaries.

Also, the players that did play it didn't take it anywhere near as seriously. I mean if Kevin Durant spends every waking moment trying to make himself the best basketball player possible and Larry Bird drinks and eats fast food and smokes cigarettes, should KD just get no credit for all that work he put in? Even though that work made him actually better at basketball should we just rank Larry Bird higher because if he played now, he might have chosen to work harder because he was in a different culture? How far do you take that? Should we say that if DeMarcus Cousins had been drafted by the Spurs and Kawhi Leonard had been drafted by the Kings, Cousins would have been the better player with the benefit of the Spurs culture so we might as well just say they're the same? Should we say Giannis is better than LeBron right now since he would have developed his skills more if he had been working on his skills from a younger age with the benefit of growing up with AAU ball in a heavy basketball culture in the US? And if not, then why shouldn't Durant get credit for working on his body more than Bird or Westbrook get credit for working on his skills more than Oscar?

Also, the lack of skills from older players aren't all just the limitations of the rules and equipment and the knowledge of those around them. There are plenty of skills that older players had every chance to work on that were just nowhere near as developed. For instance, I watched half of a Warriors/Celtics Finals game from the 60s the other day, and I must have seen at least 5 or 6 times where someone had a relatively easy shot and missed the rim completely, often not even coming close to hitting it. I'm talking about hook shots from 8 feet over a defender that was standing still on the ground, layups in transition at speed 3 feet from the basket, stuff like that. That might happen to a team a few times a year now, and when it does whoever has the miss will end up on Shaqtin' a fool. This was a regular occurrence in a half of an NBA Finals game, and the play-by-play man was talking about how it was one of the greatest games ever played! If the Suns and Bulls played that sloppy in a preseason game now, the announcers now would be openly mocking both teams. They were also terrible at moving their feet on defense, and even if you say that's a coached thing, it's still much easier to score against someone who doesn't move their feet than someone who does.

Now I'm not saying we need to completely dismiss every older player and the contributions they've made to the game. I'm just saying that in general, the skill level differences between now and then even on things within the players' control are still great enough that at the very least, they should be weighted heavier than the bench players being fresher against the top starters at a different leaguewide average pace and minute load which is a very small difference compared to going against opposing players not never put in the effort to master basic skills.
trex_8063
Forum Mod
Forum Mod
Posts: 12,684
And1: 8,322
Joined: Feb 24, 2013
     

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#54 » by trex_8063 » Mon Oct 30, 2017 2:53 am

iggymcfrack wrote:
trex_8063 wrote:In short, precious little of the differences you perceive between the old game and the new have anything to do with the players as individuals, and have A LOT more to do with circumstance. Criticizing past players for not having a modern game makes about as much declaring Thomas Edison was a moron because he wouldn’t know how to use a smart phone.

Humans haven’t inherently changed in a span of 2-3 generations. Any actual physical changes are the result of environmental influences; you level the playing field for environment, and you see the players of then and now are very very similar.
Just about the only truly valid argument pertaining to strength of era stems directly from size of player pool. Just about everything else is environment/circumstance.


Read the whole post, (just didn't want to quote it all) and there were some interesting points there, specifically about the pace. That players playing a heavy minute load at a fast pace will be unable to put up the same kind of per-minute stats as players playing less minutes at a slower pace. I would actually prefer to see something specifically tied to pace and minute load than just seeing standard deviations of top and bottom to league average though which seems like a very sloppy rough solution to a very specific problem. For instance, Steve Nash's 7 seconds or less Suns team played at a faster pace than some of the slower paced teams in the 70s and 80s, but he's still going to face the same massive scaled PER hit as someone who had lots of rest at a leisurely pace. Is that fair? Is it fair to penalize Iverson relative to Jerry West when Iverson's playing 44 MPG against opposition starters averaging in the mid-30s while West is up against starters playing the same minute load he is?


It's a fair question to ask. I'm not sure exactly how I'd want to structure a study that combines standard deviations of PER and WS/48 vs stnd dev of mpg; but it's not bad consideration at all (if memory serves, you may be the second person to suggest something similar).
fwiw (since you mentioned it), I do think the mpg thing of the times may have been part of what contributed to the greater parity (at least for the ABOVE avg players): stars were likely more often fatigued and/or pacing themselves than stars in other eras. But in the Iverson vs. West example, I can see where you have a point that West is perhaps at an unfair advantage in scaled PER/WS48 metrics.


iggymcfrack wrote:I also still think you're underestimating some of the factors for why modern players have it more difficult against better competition. For instance, you said that the only valid reason that older players would perform worse in the modern era if they played now is due to the smaller player pool. This isn't really true. The fact is that especially in the 60s, and to some extent all the way up until Magic and Bird, basketball just wasn't that popular of a sport. There may well have been a lot of American players with the potential to be all-NBA talent that just didn't bother because it didn't seem intriguing enough for them with a professional league that wasn't televised live with relatively low salaries.


I think you misunderstood what I meant, because this^^^ is what I was referring to. The much smaller player pool (and thus the [presumably] significantly lower caliber of avg player that results) is the truly valid means of critiquing the caliber of competition in earlier eras. To some degree, that was mitigated by a much more condensed league (e.g. was just 8 teams from '56-'61, and was 10 teams or fewer for 16 years: '52-'67). But noting the smaller player pool, and that it must----to a virtual certainty---produce fewer [note: fewer, but not NONE] "great" players (I'm talking great even by modern standards), and fewer "good" players, and fewer "average" players, etc, is a valid way to approach a declaration that "such and such era was not as competitive as today" or similar.

The approach of saying "these players look silly compared to today's players" is not, because the set of circumstances and influences thru which they learned the game were drastically different.


iggymcfrack wrote:Also, the players that did play it didn't take it anywhere near as seriously. I mean if Kevin Durant spends every waking moment trying to make himself the best basketball player possible and Larry Bird drinks and eats fast food and smokes cigarettes, should KD just get no credit for all that work he put in? Even though that work made him actually better at basketball should we just rank Larry Bird higher because if he played now, he might have chosen to work harder because he was in a different culture? How far do you take that?


Side note: I don't think Larry Bird smoked, but I could be wrong (he did drink, and I'm as this was an era before paid nutritionists/etc, I'm sure he had fast food). And fwiw, from everything I've read was a helluva hard worker (and certainly had a great motor).
But I get the gist of what you're saying. It's an interesting question to ask. In today's league, franchises have ample personnel to try and keep players on task, and ample resources to provide whatever is needed. At this stage they're like, "OK, we've got $69 million invested in you over the next three years; PLAY GOOD! and stay healthy; here's a half-dozen people to help you do those things....".
That stuff just wasn't around for players until relatively recently. Even as recent as your Larry Bird example: players basically had to be self-motivated to improve and take care of themselves, etc.
Much of it is micro-managed by the organization now.

So I can turn this question right back at you. You ask "should KD just get no credit for all that work he put in?" when the organization is on him to do so and providing all the resources to make it happen; I can ask "should 'Player X' of 'Past Era Y' get no credit for being self-motivated to stay fit and improve his game during an era where the organization provided almost nothing and simply said 'have a good summer; be ready to play in the fall'?"



iggymcfrack wrote:Also, the lack of skills from older players aren't all just the limitations of the rules and equipment and the knowledge of those around them. There are plenty of skills that older players had every chance to work on that were just nowhere near as developed. For instance, I watched half of a Warriors/Celtics Finals game from the 60s the other day, and I must have seen at least 5 or 6 times where someone had a relatively easy shot and missed the rim completely, often not even coming close to hitting it.....


I've acknowledged skills have improved. But again, you're faulting players for not better mastering skills that NO ONE ON THE ENTIRE PLANET EARTH had mastered to your apparent satisfaction. That very literally NO ONE HAD should give you pause as to whether that's a valid criticism. Again, this is the "criticizing/taunting Edison for not understanding a smart phone" premise.

Some of these players had "mastered" the skills that were being taught at the time; the better mechanics (on shooting, for example) simply hadn't been pioneered yet. And pioneering of new skills takes time......

.....It takes someone (or multiple someone's) having an epiphany (or perhaps stumbling on a potential improvement by accident--->did you know the discover of penicillin was accidental, btw?), or otherwise questioning if "conventional wisdom" actually has it right; and then tentatively trying out something new (which invariably encounters resistance from others----because we're somewhat resistant to change as a species----probably persuading many to give up). For those who persevere, they then must take time [usually lots of it] to do it over and over and thus "master" this new thing.
Then it takes more time for this new thing to catch on with others (and in eras before televised games, espn, youtube, or NBA league pass; or before---as you alluded to----many people were going to basketball games live and/or playing it themselves.....this process took zomg so long!!). Then it takes still more time for those handful who took notice to master this new skill (again, unless they're pushed off it by criticism/resistance from "conventional wisdom").......then this whole sequence of events needs to lather/rinse/repeat several times before it's mainstream.

Pre-Information Age, this kind of thing almost literally took generations to occur.
"The fact that a proposition is absurd has never hindered those who wish to believe it." -Edward Rutherfurd
"Those who can make you believe absurdities, can make you commit atrocities." - Voltaire
dhsilv2
RealGM
Posts: 50,627
And1: 27,310
Joined: Oct 04, 2015

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#55 » by dhsilv2 » Mon Oct 30, 2017 5:55 am

iggymcfrack

I will admit I didn't read all the above, so by all means if answered just point me. But PER should be lower for the top players today not higher if the league is better, right? That would imply the talent pool is larger and the bias of big men who by rule and genetics are nerfed a bit. Where am I wrong on this idea? Do you think that in the past the league missed THAT wildly on recruiting on the most extreme of extremes of the bell curve?

I'm a big believer we have a better league today and I think scaled PER is a great metric for this discussion. But I'll let you answer first.
dhsilv2
RealGM
Posts: 50,627
And1: 27,310
Joined: Oct 04, 2015

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#56 » by dhsilv2 » Mon Oct 30, 2017 5:57 am

BTW don't mods normally due blue font or something? I hate to have a mod fight on these votes. Would be nice if for historical reasons that mess gets moved (this post included ), to the trash bin of possible.
iggymcfrack
RealGM
Posts: 11,999
And1: 9,454
Joined: Sep 26, 2017

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#57 » by iggymcfrack » Mon Oct 30, 2017 6:41 am

dhsilv2 wrote:iggymcfrack

I will admit I didn't read all the above, so by all means if answered just point me. But PER should be lower for the top players today not higher if the league is better, right? That would imply the talent pool is larger and the bias of big men who by rule and genetics are nerfed a bit. Where am I wrong on this idea? Do you think that in the past the league missed THAT wildly on recruiting on the most extreme of extremes of the bell curve?

I'm a big believer we have a better league today and I think scaled PER is a great metric for this discussion. But I'll let you answer first.


Not necessarily. There are really myriad factors that could affect this. Players getting better could affect strategy in such a way that the value of the top players relative to those in the middle goes up or down. For instance, in a league where most players, even the elites are poor shooters from outside, it may be difficult for any one player to dominate since it's easier to defend someone who's one-dimensional and make them pass whereas a player who can score from outside or at the rim is unguardable enough to take over a game singlehandedly in a way that no one from previous eras is able to which could explain the bump in peak PERs lately.

On the other hand, the effect can just as easily work the other way. For instance, going from the 60s to the 70s, peak PERs went down. I could speculate that this is because the teams were playing the same style of basketball, but were getting more integrated meaning fewer unathletic white stiffs were playing at the bottoms of rotation while the tops of the rotations were largely staying intact. This is all very speculative and there are likely a number of significant factors in play at any given point, but I think this is a fair example that players improving could cause a shift in either direction depending on how and why.
dhsilv2
RealGM
Posts: 50,627
And1: 27,310
Joined: Oct 04, 2015

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#58 » by dhsilv2 » Mon Oct 30, 2017 1:00 pm

iggymcfrack wrote:
dhsilv2 wrote:iggymcfrack

I will admit I didn't read all the above, so by all means if answered just point me. But PER should be lower for the top players today not higher if the league is better, right? That would imply the talent pool is larger and the bias of big men who by rule and genetics are nerfed a bit. Where am I wrong on this idea? Do you think that in the past the league missed THAT wildly on recruiting on the most extreme of extremes of the bell curve?

I'm a big believer we have a better league today and I think scaled PER is a great metric for this discussion. But I'll let you answer first.


Not necessarily. There are really myriad factors that could affect this. Players getting better could affect strategy in such a way that the value of the top players relative to those in the middle goes up or down. For instance, in a league where most players, even the elites are poor shooters from outside, it may be difficult for any one player to dominate since it's easier to defend someone who's one-dimensional and make them pass whereas a player who can score from outside or at the rim is unguardable enough to take over a game singlehandedly in a way that no one from previous eras is able to which could explain the bump in peak PERs lately.

On the other hand, the effect can just as easily work the other way. For instance, going from the 60s to the 70s, peak PERs went down. I could speculate that this is because the teams were playing the same style of basketball, but were getting more integrated meaning fewer unathletic white stiffs were playing at the bottoms of rotation while the tops of the rotations were largely staying intact. This is all very speculative and there are likely a number of significant factors in play at any given point, but I think this is a fair example that players improving could cause a shift in either direction depending on how and why.


But the top players aren't 3 point shooters and I'd argue some of the top players are poor shooters by all time standards. Unless you can make a clear argument for why PER peaks are up right now and claim the league is better...the scaled PER numbers make imo far more sense. PER is already after all a somewhat iffy stat.
iggymcfrack
RealGM
Posts: 11,999
And1: 9,454
Joined: Sep 26, 2017

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#59 » by iggymcfrack » Mon Oct 30, 2017 5:27 pm

Last year's Top 5 by PER:
Russell Westbrook: .343 3P% on 7.2 3PA per game
Kevin Durant: .375 3P% on 5.0 3PA per game
Kawhi Leonard: .380 3P% on 5.2 3PA per game
Anthony Davis: .299 3P% on 1.8 3PA per game
James Harden: .347 3P% on 9.3 3PA per game

And the year before:
Stephen Curry: .454 3P% on 11.2 3PA per game
Kevin Durant: .387 3P% on 6.7 3PA per game
Russell Westbrook: .296 3P% on 4.3 3PA per game
LeBron James: .307 3P% on 3.9 3PA per game
Chris Paul: .371 3P% on 4.4 3PA per game

Not sure where you get the idea that any of those guys are not 3-point shooters or are poorer shooters compared to previous eras.

Here's the Top 5 10 years ago:
Dwyane Wade: .266 3P% on 1.5 3PA per game
Dirk Nowitzki: .416 3P% on 2.2 3PA per game
Yao Ming: .000 3P% on 0.0 3PA per game
Tim Duncan: .111 3P% on 0.1 3PA per game
Kobe Bryant: .344 3P% on 5.2 3PA per game

And 20 years ago with a shorter 3-point line:
Karl Malone: .000 3P% on 0.2 3PA per game
Michael Jordan: .374 3P% on 3.6 3PA per game
Shaquille O' Neal: .000 3P% on 0.1 3PA per game
Grant Hill: .303 3P% on 0.4 3PA per game
Charles Barkley: .283 3P% on 3.9 3PA per game

Hell, let's try 30 years ago for fun:
Michael Jordan: .182 3P% on 0.8 3PA per game
Magic Johnson: .205 3P% on 0.5 3PA per game
Larry Bird: .400 3P% on 3.0 3PA per game
Charles Barkley: .202 3P% on 1.5 3PA per game
Kevin McHale: .000 3P% on 0.1 3PA per game
dhsilv2
RealGM
Posts: 50,627
And1: 27,310
Joined: Oct 04, 2015

Re: RealGM 2017 Top 100 List #56 (Willis Reed) 

Post#60 » by dhsilv2 » Mon Oct 30, 2017 6:41 pm

I mean they shoot them. Other than maybe KD, none would be 3 point shooters as I see it. But yes, we have more guards moving up PER due to changes is rebound distribution.

Return to Player Comparisons