RealGM 2023 Top 100 Project - #35 (Walt Frazier)

[iggymcfrack]

Playoff career scoring averages for some notable players:

Anthony Davis: 25.9 PPG on .618 TS%
Hakeem Olajuwon: 25.9 PPG on .553 TS%
Damian Lillard: 25.2 PPG on .588 TS%
Kobe Bryant: 25.0 PPG on .541 TS%
Kareem Abdul Jabbar: 24.6 PPG on .571 TS%
Shaquille O'Neal: 24.3 PPG on .565 TS%
Larry Bird: 24.3 PPG on .551 TS%
James Harden: 22.7 PPG on .585 TS%
Wilt Chamberlain: 22.5 PPG on .524 TS%

[penbeast0]

Frazier has been docked for a short prime but just for fun, let's look at WS to estimate career value as injuries have been the big issue with AD.

RS Winshares
Frazier 113.5
Davis 100.3
But wait, Frazier played 2 more years. So, let's eliminate the Cleveland years for Frazier as they are about as much a positive factor as Jordan with the Wizards

Frazier 10 years 108.8 WS
Davis 11 years 100.3 WS

But Davis is a playoff monster!
Frazier 8 seasons in 10 years @ 15.9 (best in league twice in playoff WS in 72 3.3, 73 3.2)
Davis 5 seasons in 11 years @ 9.6 (best in league once in 2020 4.5)

I think that's closer to how most people see things. You can still make a case for Davis over Frazier as he has flashier stats/48 and a higher peak playoff WS but Frazier took his team to the playoffs more often and has superior WS because he was reliably out there year in and year out plus two strong finals v. one.

[Doctor MJ]

I know people have an irrational hatred of PER and it does have its flaws, but Havlicek peaked at a PER of 20.8 and Frazier peaked at a PER of 21.6. Davis beats that in 11 of 11 seasons including his rookie year! As a league adjusted stat, it’s actually harder for Davis to get a good PER in his tougher era than it would be for the old heads. And of course he’s a much better defender than either. Let’s be real here, Hondo and Clyde may have racked up accomplishments in weak leagues, but they’re nowhere near AD’s level in terms of talent. If you’re drafting between them for a team, Davis is a no brainer.

Consider Andre Drummond. He's spent most of his career with a PER north of that and never learned how to play basketball well enough to be a starter on a serious team. Folks overrating metrics like this is what led some to conclude he was the "real rookie of the year" over Lillard and Davis and argue that he would be a perennial MVP candidate, and less dramatically, led the Pistons to treat him as a franchise player which resulted in him making all-star a couple times early in his career.

Davis is awesome and I don't mean to imply otherwise, but it's not an irrational hatred of PER that makes us so skeptical of its usage as proxy for player goodness/value/etc.

Andre Drummond led the league in rebounding 4 times and in his rookie year, he had 2.8 blocks per 36 minutes. He didn't need PER to overrate him. He's a poor defender though which kills a lot of his value. AD was voted DPOY twice on this board and is 3rd in DPOY shares over the 9 years the project's been active. And it's not like he's just a finisher either. He's 16th all-time in playoff PPG. I think that when PER's the best tool we have to compare across eras, it's helpful to just look at the context and say "is there any reason to think that PER would vastly overrate or underrate one of these players?" In Drummond's case, it's obvious. With AD, there's not really much reason to think so. Now might it overrate him a bit since he's not as good of a passer as some of the players above him? Sure. I think it's still fair to say that he has a solid 4 seasons better than anyone else in the project though. He's just a different class of talent.

Okay, if you're point is that it's more than just PER that overrated Drummond, you are of course correct.

My point is that it's problematic to say Davis > X because PER when we know that PER would say Drummond > X, when we know Drummond << X.

I'm not saying it's ridiculous to look at the stat and mention it in passing, but fundamentally, the existence of the Drummonds of the world makes it hard for some of us to be influenced much when we see anything favoring one guy over another.

Re: but there's no reason to think Davis is overrated by it. Hmm. That logic seems to imply that if you know a guy is great, and he looks better than someone else by a certain stat, then the stat must have some authority, and I would say that's false.

In the end, PER is just clumsily adding up box score stuff and can't really tell the difference between Davis & Drummond, which means it's just not that effective of a tool.

[AEnigma]

What makes PER innately more valuable than Bill Simmons’ “just add points, rebounds, and assists”. If you do that then you will probably get decent enough results — last year’s per game leaders were Luka, Giannis, Embiid, and Jokic — so why bother with all this extra noise?

[MrLurker]

Consider Andre Drummond. He's spent most of his career with a PER north of that and never learned how to play basketball well enough to be a starter on a serious team. Folks overrating metrics like this is what led some to conclude he was the "real rookie of the year" over Lillard and Davis and argue that he would be a perennial MVP candidate, and less dramatically, led the Pistons to treat him as a franchise player which resulted in him making all-star a couple times early in his career.

Davis is awesome and I don't mean to imply otherwise, but it's not an irrational hatred of PER that makes us so skeptical of its usage as proxy for player goodness/value/etc.

Andre Drummond led the league in rebounding 4 times and in his rookie year, he had 2.8 blocks per 36 minutes. He didn't need PER to overrate him. He's a poor defender though which kills a lot of his value. AD was voted DPOY twice on this board and is 3rd in DPOY shares over the 9 years the project's been active. And it's not like he's just a finisher either. He's 16th all-time in playoff PPG. I think that when PER's the best tool we have to compare across eras, it's helpful to just look at the context and say "is there any reason to think that PER would vastly overrate or underrate one of these players?" In Drummond's case, it's obvious. With AD, there's not really much reason to think so. Now might it overrate him a bit since he's not as good of a passer as some of the players above him? Sure. I think it's still fair to say that he has a solid 4 seasons better than anyone else in the project though. He's just a different class of talent.

Okay, if you're point is that it's more than just PER that overrated Drummond, you are of course correct.

My point is that it's problematic to say Davis > X because PER when we know that PER would say Drummond > X, when we know Drummond << X.

I'm not saying it's ridiculous to look at the stat and mention it in passing, but fundamentally, the existence of the Drummonds of the world makes it hard for some of us to be influenced much when we see anything favoring one guy over another.

Re: but there's no reason to think Davis is overrated by it. Hmm. That logic seems to imply that if you know a guy is great, and he looks better than someone else by a certain stat, then the stat must have some authority, and I would say that's false.

In the end, PER is just clumsily adding up box score stuff and can't really tell the difference between Davis & Drummond, which means it's just not that effective of a tool.

Not to champion this stat - I think I share your concerns - but wouldn't a great defensive force like Davis be underrated by a stat like PER?

[penbeast0]

What makes PER innately more valuable than Bill Simmons’ “just add points, rebounds, and assists”. If you do that then you will probably get decent enough results — last year’s per game leaders were Luka, Giannis, Embiid, and Jokic — so why bother with all this extra noise?

First time I thought about and tried to create my own versions of an all in one stat was when I read an interview with Larry Bird where he talked about something called Tendex which basically added all the statistical categories of his day together to produce a rating.

I liked PER when it came out in that Hollinger tried to back engineer it for values by playing with the values of each box score category until it could more accurately predict the previous decades team records. Thought that was about the best you could do with just box score stats. Now there is more information available and PER seems creaky and problematic but when it came out, it was interesting (okay, except for always having Jose Calderon as a superstar).

[Doctor MJ]

Andre Drummond led the league in rebounding 4 times and in his rookie year, he had 2.8 blocks per 36 minutes. He didn't need PER to overrate him. He's a poor defender though which kills a lot of his value. AD was voted DPOY twice on this board and is 3rd in DPOY shares over the 9 years the project's been active. And it's not like he's just a finisher either. He's 16th all-time in playoff PPG. I think that when PER's the best tool we have to compare across eras, it's helpful to just look at the context and say "is there any reason to think that PER would vastly overrate or underrate one of these players?" In Drummond's case, it's obvious. With AD, there's not really much reason to think so. Now might it overrate him a bit since he's not as good of a passer as some of the players above him? Sure. I think it's still fair to say that he has a solid 4 seasons better than anyone else in the project though. He's just a different class of talent.

Okay, if you're point is that it's more than just PER that overrated Drummond, you are of course correct.

My point is that it's problematic to say Davis > X because PER when we know that PER would say Drummond > X, when we know Drummond << X.

I'm not saying it's ridiculous to look at the stat and mention it in passing, but fundamentally, the existence of the Drummonds of the world makes it hard for some of us to be influenced much when we see anything favoring one guy over another.

Re: but there's no reason to think Davis is overrated by it. Hmm. That logic seems to imply that if you know a guy is great, and he looks better than someone else by a certain stat, then the stat must have some authority, and I would say that's false.

In the end, PER is just clumsily adding up box score stuff and can't really tell the difference between Davis & Drummond, which means it's just not that effective of a tool.

Not to champion this stat - I think I share your concerns - but wouldn't a great defensive force like Davis be underrated by a stat like PER?

Compared to another big who is bad at defense? Absolutely.

But bigs in general are not underrated by the stat.

[Doctor MJ]

Just as a thought exercise, who is the best unnominated player left from each decade (adding those mentioned so far):

50s -- Schayes
60s -- Baylor
70s -- Gilmore, Barry, Reed
80s --
90s --
00s -- Kidd, Ginobili
10s -- Howard
20s --

Good thing to do. I'm going to go by NBA/NBL/ABA debut decade just because I was getting tangled in knots bit trying to decide where to place certain guys.

40s - Schayes, Davies
50s - Arizin, Baylor
60s - Barry, Reed
70s - Gilmore, Cowens
80s - McHale, Isiah
90s - Kidd, Allen
00s - Ginobili, Gasol
10s - Green, Butler

Am I taking crazy pills, or isn't Kidd already nominated? (I hope so: I've been voting for him for two threads).

You're right lol!

I've updated with Allen & Pierce as my next to '90s debut guys.

[Doctor MJ]

Good response, Doc, and gets to more of what I wanted. The logic of “Pierce was never top five and hovered more around top ten/fifteen” is fine by itself. I disagree on the extent to which Manu was meaningfully separating himself outside of 2005, and I think most years I would rather have Pierce’s minutes than Manu’s minutes, but if someone were less critical of the minutes gap, then there is not much else to be said having conceded that a minute of Manu is probably more valuable than a minute of Pierce.

For the sake of this Malone analogy, let us momentarily cast aside the longevity question. Instead, the new hypothetical is whether Malone would be more valuable than… Shaq, say, if Shaq averaged thirty minutes a game throughout his prime. And there I might be inclined toward yes, because a 30 mpg Shaq may well have no titles. A 30 mpg Shaq would be playing a lot less than his perimetre partners, and while his possessional “impact” would be a lot higher than theirs, his per game impact would go down, as would his per season impact and his career-wide impact.

I am not totally divorcing success from my analysis, so if Malone had fifty years as a top five player but was never able to break through, yeah, it would be tricky for me to crown him… but in another sense, that would make him a squished version of Kareem who (seemingly in this hypothetical) never found a partner as good as even 1980 Magic or 1971 Oscar (both of whom could have won titles with Malone in Stockton’s place during those Hornacek years). Does Jordan have any titles without Pippen? Does Lebron have any titles if we just keep making his teams progressively worse (or maybe making the 2012 Celtics one shot better in Game 2)? Intellectually, sure, the idea of a guy with an extended peak outside of my personal top 20 being the greatest player seems off, but I also do not automatically push people to the top just for having an all-time extended peak either.

I mean, the question is kind-of absurd because the idea of a player just staying top 5 for multiple decades is functionally incomprehensible. Not even Lebron managed that. And the hypothetical where someone did manage that would say a lot about how we assess era quality — imagine what that would do for our perceptions if we could definitively show that a top 5 player in 1969 would be a top 5 player in 2019 as well, despite all the development and international talent expansion.

So I guess my answer there is he definitely could be, but I also do not see it as too relevant to this question of Manu versus Pierce because I think the average Pierce season was more valuable in totality than the average Manu season anyway. Credit due to Manu for performing as he did in the 2005 postseason (although Pierce had some gaudy production of his own that year ), but he shot 44% from the perimetre over that stretch and then never crossed 39% in any other prime postseason; we have the ability to parse that for what it is while also acknowledging how it won them the title.

So I think we've mostly got our respective perspectives clarified, but your push back on the multiple decade thing seems like something I should address.

As I've said, it's about process.

If one were to go full linear analytical and let each year keep mounting on top of the next indefinitely, then there should be a specific point where the inferior player's longevity trumps the superior player's prime, basically in all circumstances, so long s the player with the longevity is still achieving positive success along the way.

If we're talking about something that takes an unrealistic amount of time it's certainly okay not to know how many years that would take...but the key thing is that there is a particular moment - apparently far removed from a player's peak - where your opinion over the greater career switches from one guy to the other.

And to me this doesn't sit right for evaluating world's best type athletes. Not saying others can't go for it, but I just don't. To me these guys are to be judged based on how you'd build with them with time beyond that operating largely as a tiebreaker.

For this reason, I'm more invested in the question of how we should hold MPG against a player than total minutes. I've defended Ginobili on this front too, but I don't mean to give the impression it isn't hurting him in my book. It is, even if not as much as it is for others.

[penbeast0]

So I think we've mostly got our respective perspectives clarified, but your push back on the multiple decade thing seems like something I should address.

As I've said, it's about process.

If one were to go full linear analytical and let each year keep mounting on top of the next indefinitely, then there should be a specific point where the inferior player's longevity trumps the superior player's prime, basically in all circumstances, so long s the player with the longevity is still achieving positive success along the way.

If we're talking about something that takes an unrealistic amount of time it's certainly okay not to know how many years that would take...but the key thing is that there is a particular moment - apparently far removed from a player's peak - where your opinion over the greater career switches from one guy to the other.

And to me this doesn't sit right for evaluating world's best type athletes. Not saying others can't go for it, but I just don't. To me these guys are to be judged based on how you'd build with them with time beyond that operating largely as a tiebreaker.

For this reason, I'm more invested in the question of how we should hold MPG against a player than total minutes. I've defended Ginobili on this front too, but I don't mean to give the impression it isn't hurting him in my book. It is, even if not as much as it is for others.

If it's just a tiebreaker, how come you haven't voted for Walton yet? Clearly there's some level at which the prime is too short or too injury prone. For me it's around 8 years of prime play to get full value of an average prime year as their baseline (I don't care as much about subprime play as many here). I will vote for people with less but I do dock them for it. I will give a bonus for more but I don't value longevity as highly as some. I seem to value reliability (not being injured multiple times, playing full seasons rather than partial ones) more than some as well.

[AEnigma]

^ Agreed, there is obviously a cutoff somewhere. Are you ignoring Karl Malone in favour of a player you see as exactly one spot better ten years in a row but who drops off precipitously after? What about nine years in a row? What if instead of doing so every year it is more of an average finish? What if the average is just half a spot? How high can we go if the one shortcoming in this hypothetical hyper-extended career is that Malone never finishes #1 in any given season, but is otherwise second or third in perpetuity?

My understanding is this is more of a career project rather than a primes project — although how primes are weighed against careers is open to personal interpretation. Alright, yes, Manu will and should go well above Kevin Willis despite a substantial gap in minutes, but that is not the scale of comparison. If Manu is a regular “weak MVP” player despite his minutes, then sure, prop him up against Frazier or Pippen, but for me, he is more typically in that lower all-NBA group with Pierce (2005 postseason aside).

[falcolombardi]

I am unfortunalte short on time to spend talking basketball so i will be brief

Vote-reggie miller
All time level playoff scorer with great (offense) portability and strong longevity, and unlike other "old" players his style of play fits perfectly in any era old or new

Unlike some other great scorers who were not great passers or on-ball playmakers for their teams say adrian dantley or george gervin. Reggie actually has the big impact signals leading a team offense run on par with many guys way more awarded than him (such as barkley)

His teams were not weak as he generally had some combination of high level players as mark jackson, rik smits or detlef schremp , but were not stacked with stars either with reggie being by far the epicenter of a top offensive team for a really long time

Compar3d to the other guys here who have big flaws in either longevity, durability or have a era handicap applied to them to some degree. Reggie fits the more boxes for me

Alternate: i cannot in good faith vote davis here with a durability/longevity that makes kawhi look like AC Green. And i am unfamiliar with many of the guys proposed here like hondo

I will avoid voting am alternate for now

If alts are a requirement then i will go with frazier who while mayve a tad mythologized on defense for me from watching film ( i think his style was made possible by the era dribble rules and limitationa) he seems like a tier 1 offense star whose offensive brillancy went weirdly under the radar

My nomination is russel westbrook

Not the best aging curve being arguably one of the few ever players to get less skilled as he aged so longevity is so-so. But the average prime year was really good. In some ways the apex of the inneficient scorer with huge creation archetype

And unlike what a lot of people would expect,he was able to be half of the motor of some truly monstrous offenses.

He is also a player with serious fit limitations (mediocre defender, bad off ball, not a top end om ball decision maker so his ceiling will always be lower than true offense mega stars)

But on aggregate i think the value he provided to legit contending thunder teams fron 2013-2016 and then as a floor raiser in 2017 before he started a not so gentle decline is worth noticing.

[OhayoKD]

Vote
1. Frazier

-> Kind of a proto-Kawhi and has led teams to titles
-> One of the best all-rounders at his position
-> Probably better longevity than who we just voted in so

Nomination

1. Westbrook

Honestly weird he hasn't gotten discussion yet(and now that I think about he probably should have already been inducted by now)

but whatever, let's get this going

-> All-time Creator with all-time playoff elevation and all-time playoff impact
-> Was the most valuable piece on a team that thumped a 67-win team and took a 73-win team to 7, probably the best playoff performer in 2014 on a team which pushed the tiki-taka spurs without their best defender
-> Track-record of elevating against better opponents
-> Excellent cultural figure/teammate by all accounts, something which he leveraged to help OKC sign Paul George to a long-term contract, something they are still benefitting from
-> Great RS floor-raiser, 45-wins(full-strength) without KD with OKC's shallowest cast in 2015, and 2017 was even better
-> Saw a +9 srs team in 2013 turn into something like a +3 one when he was hurt
-> Excellent clutch player
-> Underrated longetivity, has been an elite playoff creator as early as 2010(when he elevated vs the eventual champions as he tends to do), had a strong 2023

Alt-nomination

2. Draymond

Will get into this more but he has the leas empirical question marks than Manu, the better real-world profile, arguably better RAPM/plus-minus, is more proven without Steph, and I'd say has the best series performance in the 2016 finals.

An important point to consider I think when using finals +/- is that Draymond has generally ran into much better finals opponents. Have not done it with the celtics(though I imagine they'd look good), but every other finals opponent Draymond has run into entered with a higher rolling srs/psrs than any finals opponent Manu has run into. The weakest, the 2015 cavs, came off a series where they performed at +16 vs the hawks with kyrie barely playing and no kevin love.

[HeartBreakKid]

I am unfortunalte short on time to spend talking basketball so i will be brief

Vote-reggie miller
All time level playoff scorer with great (offense) portability and strong longevity, and unlike other "old" players his style of play fits perfectly in any era old or new

Unlike some other great scorers who were not great passers or on-ball playmakers for their teams say adrian dantley or george gervin. Reggie actually has the big impact signals leading a team offense run on par with many guys way more awarded than him (such as barkley)

His teams were not weak as he generally had some combination of high level players as mark jackson, rik smits or detlef schremp , but were not stacked with stars either with reggie being by far the epicenter of a top offensive team for a really long time

Compar3d to the other guys here who have big flaws in either longevity, durability or have a era handicap applied to them to some degree. Reggie fits the more boxes for me

Alternate: i cannot in good faith vote davis here with a durability/longevity that makes kawhi look like AC Green. And i am unfamiliar with many of the guys proposed here like hondo

I will avoid voting am alternate for now

If alts are a requirement then i will go with frazier who while mayve a tad mythologized on defense for me from watching film ( i think his style was made possible by the era dribble rules and limitationa) he seems like a tier 1 offense star whose offensive brillancy went weirdly under the radar

My nomination is russel westbrook

Not the best aging curve being arguably one of the few ever players to get less skilled as he aged so longevity is so-so. But the average prime year was really good. In some ways the apex of the inneficient scorer with huge creation archetype

And unlike what a lot of people would expect,he was able to be half of the motor of some truly monstrous offenses.

He is also a player with serious fit limitations (mediocre defender, bad off ball, not a top end om ball decision maker so his ceiling will always be lower than true offense mega stars)

But on aggregate i think the value he provided to legit contending thunder teams fron 2013-2016 and then as a floor raiser in 2017 before he started a not so gentle decline is worth noticing.

Would you put Westbrook over Davis?

[f4p]

Consider Andre Drummond. He's spent most of his career with a PER north of that and never learned how to play basketball well enough to be a starter on a serious team. Folks overrating metrics like this is what led some to conclude he was the "real rookie of the year" over Lillard and Davis and argue that he would be a perennial MVP candidate, and less dramatically, led the Pistons to treat him as a franchise player which resulted in him making all-star a couple times early in his career.

Davis is awesome and I don't mean to imply otherwise, but it's not an irrational hatred of PER that makes us so skeptical of its usage as proxy for player goodness/value/etc.

Andre Drummond led the league in rebounding 4 times and in his rookie year, he had 2.8 blocks per 36 minutes. He didn't need PER to overrate him. He's a poor defender though which kills a lot of his value. AD was voted DPOY twice on this board and is 3rd in DPOY shares over the 9 years the project's been active. And it's not like he's just a finisher either. He's 16th all-time in playoff PPG. I think that when PER's the best tool we have to compare across eras, it's helpful to just look at the context and say "is there any reason to think that PER would vastly overrate or underrate one of these players?" In Drummond's case, it's obvious. With AD, there's not really much reason to think so. Now might it overrate him a bit since he's not as good of a passer as some of the players above him? Sure. I think it's still fair to say that he has a solid 4 seasons better than anyone else in the project though. He's just a different class of talent.

Okay, if you're point is that it's more than just PER that overrated Drummond, you are of course correct.

My point is that it's problematic to say Davis > X because PER when we know that PER would say Drummond > X, when we know Drummond << X.

I'm not saying it's ridiculous to look at the stat and mention it in passing, but fundamentally, the existence of the Drummonds of the world makes it hard for some of us to be influenced much when we see anything favoring one guy over another.

the existence of the drummonds of the world? even ignoring that the other 2 of the big 3 box score stats don't like drummond as much (and downright say he sucks in the playoffs), so what if PER says he's good. are the other stats perfect or something? it's not the skepticism with PER, it's the weird "dismiss at all costs" aspect of it that's so strange in a lot of conversations on this board. like it can't even be referenced to point out vast differences without there just being a "well, who knows if that means anything" response. as if it's a randomly generated number.

AD had a 28 PER from ages 20-26. has anyone done that in PER that wasn't great? like really great? it's one thing to be skeptical, it's another to say that we can't know if it means something. i looked at 7 year stretches for everybody else in nba history (i skipped years and did 20-26, 22-28, 24-30, 26-32, and 28-34) and even set a really low limit on minutes of 5000 MP just to not miss anyone. here are all the people who have ever gotten even a 27 PER over 7 years:

Wilt
MJ
Lebron
AD
Shaq
Jokic
Robinson
Pettit
Kareem
Luka (hasn't actually played 7 years)
Giannis
Embiid
Mikan
Curry
Durant
Harden
Wade

So literally AD with a bunch of MVP's, a guy (Luka) who will probably get one, and an alpha champion in Wade. In other words, you don't get AD's numbers without it meaning something. Drummond picking up some low 20's PER's is a lot easier to not mean anything.

of course, we can look at stats like RAPM and find lots of weird results. JE 97-22 RAPM:

We've got Jayson Tatum in 4th above Jokic, Jordan and Duncan.

Paul George over Giannis.

We've got Jrue Holiday and Karl Anthony Towns above Kevin Durant.

Jakob Poeltl and Immanuel Quickley above Steve Nash (and Nash at 40th overall).

Chuck Hayes and Josh Howard edging out Kobe in 73rd place.

Marcus Smart just barely edging out Brian Cardinal.

Trent Forrest and Omer Asik above all of Anthony Davis, Russell Westbrook, and Dwyane Wade, all of whom are in the 160's.

do all the KG defenders have to stop mentioning plus/minus metrics?

Re: but there's no reason to think Davis is overrated by it. Hmm. That logic seems to imply that if you know a guy is great, and he looks better than someone else by a certain stat, then the stat must have some authority, and I would say that's false.

In the end, PER is just clumsily adding up box score stuff and can't really tell the difference between Davis & Drummond, which means it's just not that effective of a tool.

see above for who is in Davis's company. also, how does it not tell the difference between Davis and Drummond? there is a vast gulf between their PER's.


i posted this in thread #32.

top players by playoff PER in history (age 21-35):
1. MJ
2. Lebron
3. Shaq
4. Hakeem
5. Kareem
6. Duncan
7. Kawhi
8. Barkley
9. Dirk
10. Durant
11. Robinson
12. CP3
13. Wilt (extended age range sunk him)
14. Magic
15. Jerry
16. Steph
17. Kobe
18. Pettit
19. Wade
20. Karl
21. Harden

all of the top 21 except the massively injured kawhi already in. we even have
31. Reggie
35. Ewing
36. Stockton
38. Frazier

all just or about to be voted in right in order

how many other stats are tracking the top 100 this well (with the acknowledgement that there's nothing that say the Top 100 is perfect either)? and that's just a raw list. not factoring in longevity.

[Doctor MJ]

So I think we've mostly got our respective perspectives clarified, but your push back on the multiple decade thing seems like something I should address.

As I've said, it's about process.

If one were to go full linear analytical and let each year keep mounting on top of the next indefinitely, then there should be a specific point where the inferior player's longevity trumps the superior player's prime, basically in all circumstances, so long s the player with the longevity is still achieving positive success along the way.

If we're talking about something that takes an unrealistic amount of time it's certainly okay not to know how many years that would take...but the key thing is that there is a particular moment - apparently far removed from a player's peak - where your opinion over the greater career switches from one guy to the other.

And to me this doesn't sit right for evaluating world's best type athletes. Not saying others can't go for it, but I just don't. To me these guys are to be judged based on how you'd build with them with time beyond that operating largely as a tiebreaker.

For this reason, I'm more invested in the question of how we should hold MPG against a player than total minutes. I've defended Ginobili on this front too, but I don't mean to give the impression it isn't hurting him in my book. It is, even if not as much as it is for others.

If it's just a tiebreaker, how come you haven't voted for Walton yet? Clearly there's some level at which the prime is too short or too injury prone. For me it's around 8 years of prime play to get full value of an average prime year as their baseline (I don't care as much about subprime play as many here). I will vote for people with less but I do dock them for it. I will give a bonus for more but I don't value longevity as highly as some. I seem to value reliability (not being injured multiple times, playing full seasons rather than partial ones) more than some as well.

I suppose I'd say big picture that I'm more dynasty-oriented than championship-oriented. The goal isn't to win the chip once, it's to have a chance to win as many as possible with a given core.

You say 8 years, and I think that's pretty reasonable. Guys like Walton - and my fave Connie Hawkins - get dinged because of this. Of course, last time I voted for Hawk a number of times before the project reached 100 with no Hawk on the list, so I think I do ding guys less for this than many do.

[Doctor MJ]

^ Agreed, there is obviously a cutoff somewhere. Are you ignoring Karl Malone in favour of a player you see as exactly one spot better ten years in a row but who drops off precipitously after? What about nine years in a row? What if instead of doing so every year it is more of an average finish? What if the average is just half a spot? How high can we go if the one shortcoming in this hypothetical hyper-extended career is that Malone never finishes #1 in any given season, but is otherwise second or third in perpetuity?

My understanding is this is more of a career project rather than a primes project — although how primes are weighed against careers is open to personal interpretation. Alright, yes, Manu will and should go well above Kevin Willis despite a substantial gap in minutes, but that is not the scale of comparison. If Manu is a regular “weak MVP” player despite his minutes, then sure, prop him up against Frazier or Pippen, but for me, he is more typically in that lower all-NBA group with Pierce (2005 postseason aside).

It's a career project, but how we weight prime vs longevity is up to personal philosophy. The way I tend to see things nowadays effectively waits longevity less than most with an analytical bent, but that's just me personally, not a prescription for how others should work within this project.

Re: prop him up against Frazier or Pippen. One thing I'd note is that I've never made the case for Ginobili over Frazier in this project. I was in debate between Ginobili, Pippen, Stockton, Miller, Havlicek, and others, but I had Frazier distinctly above all of them.

[Doctor MJ]

Andre Drummond led the league in rebounding 4 times and in his rookie year, he had 2.8 blocks per 36 minutes. He didn't need PER to overrate him. He's a poor defender though which kills a lot of his value. AD was voted DPOY twice on this board and is 3rd in DPOY shares over the 9 years the project's been active. And it's not like he's just a finisher either. He's 16th all-time in playoff PPG. I think that when PER's the best tool we have to compare across eras, it's helpful to just look at the context and say "is there any reason to think that PER would vastly overrate or underrate one of these players?" In Drummond's case, it's obvious. With AD, there's not really much reason to think so. Now might it overrate him a bit since he's not as good of a passer as some of the players above him? Sure. I think it's still fair to say that he has a solid 4 seasons better than anyone else in the project though. He's just a different class of talent.

Okay, if you're point is that it's more than just PER that overrated Drummond, you are of course correct.

My point is that it's problematic to say Davis > X because PER when we know that PER would say Drummond > X, when we know Drummond << X.

I'm not saying it's ridiculous to look at the stat and mention it in passing, but fundamentally, the existence of the Drummonds of the world makes it hard for some of us to be influenced much when we see anything favoring one guy over another.

the existence of the drummonds of the world? even ignoring that the other 2 of the big 3 box score stats don't like drummond as much (and downright say he sucks in the playoffs), so what if PER says he's good. are the other stats perfect or something? it's not the skepticism with PER, it's the weird "dismiss at all costs" aspect of it that's so strange in a lot of conversations on this board. like it can't even be referenced to point out vast differences without there just being a "well, who knows if that means anything" response. as if it's a randomly generated number.

I didn't say to dismiss PER at all costs. What I pointed out was that you could use the same argument iggy used with PER to advocate for Davis over other players to also argue for Drummond. With iggy presenting the argument like, "How is this even a debate, Davis is way ahead, isn't he?", the fact that you could present the same argument for someone where it would obviously be wrong makes clear that the argument by itself is problematic.

Could you do something similar with other stats? Quite possibly, and to the extent you can, it speaks to the limitations of those stats.

But look, I think I know what you mean with the "dismiss at all costs" thing. There were responses to the use of PER that seemed to laugh it off. I'm not saying to laugh it off though, I'm just pointing out politely a flaw in an argument to help someone understand why I don't draw the same conclusions he does.

AD had a 28 PER from ages 20-26. has anyone done that in PER that wasn't great? like really great? it's one thing to be skeptical, it's another to say that we can't know if it means something. i looked at 7 year stretches for everybody else in nba history (i skipped years and did 20-26, 22-28, 24-30, 26-32, and 28-34) and even set a really low limit on minutes of 5000 MP just to not miss anyone. here are all the people who have ever gotten even a 27 PER over 7 years:

Wilt
MJ
Lebron
AD
Shaq
Jokic
Robinson
Pettit
Kareem
Luka (hasn't actually played 7 years)
Giannis
Embiid
Mikan
Curry
Durant
Harden
Wade

So literally AD with a bunch of MVP's, a guy (Luka) who will probably get one, and an alpha champion in Wade. In other words, you don't get AD's numbers without it meaning something. Drummond picking up some low 20's PER's is a lot easier to not mean anything.

So, the issue here for me isn't that I don't think Davis is great - as I've already said, I think he is - frankly I was on this bandwagon before virtually everyone else - but whether you can specifically damn someone else because he's not on this list.

Probably the obvious thing here is that Wilt's #1 on the list while Russell isn't on it, but I personally rank Russell's career much higher than Wilt. Whether I'm right or wrong on that, this is something I've been very clear on for a long time.

Perhaps more subtle is the absence of great floor generals who aren't mega-scorers. That 28 PER threshold, guys like Magic & Nash never hit that, yet I think on average in their primes they were considerably more valuable than most of the guys on this list.

of course, we can look at stats like RAPM and find lots of weird results. JE 97-22 RAPM:

We've got Jayson Tatum in 4th above Jokic, Jordan and Duncan.

Paul George over Giannis.

We've got Jrue Holiday and Karl Anthony Towns above Kevin Durant.

Jakob Poeltl and Immanuel Quickley above Steve Nash (and Nash at 40th overall).

Chuck Hayes and Josh Howard edging out Kobe in 73rd place.

Marcus Smart just barely edging out Brian Cardinal.

Trent Forrest and Omer Asik above all of Anthony Davis, Russell Westbrook, and Dwyane Wade, all of whom are in the 160's.

do all the KG defenders have to stop mentioning plus/minus metrics?

Weird results mean we should ask ourselves what the data means. I'm a big user of +/- stats and welcome such back & forth. Since you mentioned various examples, I'll hit some things I've already thought about:

- I think Tatum is more valuable than any of us (myself included) really seem to understand. He doesn't seem like an outlier BBIQ guy to me, but basically his whole career he's looked fantastic by +/- stats. I expect to advocate for Tatum because I take this data seriously, even if I don't treat it as gospel, and thus will end up seeing him more accomplished than most.

- I do think the fact that Tatum has been in one place his whole career, and it's largely been a place where things are working very well, gives him a boost he would not have if he'd been drafted other places.

- I don't think Tatum's game is as resilient against the toughest competition as the very best, and that matters a lot to me, but won't show up on long-term RAPM studies.

- George over Giannis. Giannis has a tremendous RAPM peak surrounding by many, many years of not actually seeming to be that impactful. I think this is something that folks with a foot in analytics in general have avoided dealing with and it's caused him to be overrated even by groups such as this one. I'd still put Giannis well ahead of George for career for what he's demonstrated at his best, along with the fact that I knock George for some of his choices along the way.

- X over Durant. We have to start here by recognizing that +/- stat always said that young Durant wasn't just meh, but outright just about the least valuable player in the entire league - far below replacement level. It's a great example of how teams groom proto-superstars by placing them in high primacy roles not because it helps the present team, but because it will hopefully help going forward. Beyond that with Durant you've got a classic individualist who thinks that the part of the game he's really, really great at is all that matters. So the general rule with Durant is that he's generally less valuable than most think he is...but that doesn't mean his skills can't be extremely valuable in the right context, and there's a resiliency to his ability to apply those skills against elite competition that matters a great deal.

I'll note that I was voting Durant over Paul in this project despite the fact that Paul is much, much better at being valuable in any given moment of play. Paul is a much smarter basketball player...but he's also a shrimp whose size and body limitations really get exposed in elite play, in my assessment. I think Durant was instrumental in creating the best team that's ever existed. I'm skeptical Paul could be a starter on such a team.

- Jrue. Is super-underrated in general, and really should have been a perennial all-star. He's got major offensive limitations that keep him from reaching higher heights. I would not advocate for him over Durant, or Paul for that matter, and frankly I don't expect to champion him in this project...but I'm honestly tempted to, and I think that if his health is solid, Boston was super-smart to go get him.

- KAT. An excellent case study in the limitations of regression stats. If you're on a bad enough team, and your presence is enough to make your team decent, a RAPM style stat isn't going to be able to distinguish between you and someone contributing similar lift on a better team. My take on KAT in general is that good teams can exploit his weaknesses when they need to, but since KAT played on the Timberwolves, they really didn't need to.

- Various guys who play limited minutes scoring well. So, the low minute thing is a known thing. It's not a given that someone playing low minutes with high impact indicators couldn't do the same thing playing big minutes, but it's always something to be wary about.

Incidentally, the reason why I didn't give Jokic my hypothetical ROY vote in '15-16 (I gave it to KAT) was because he played much of his time in limited minutes on the bench. That wasn't me betting against Jokic's future, but it was me being cautious because of the broader phenomenon here. Once Jokic proved he could keep it up as a star, I honestly wished I could go back in time and vote Jokic for ROY, but alas, I left my DeLorean in Hill Valley.

I will say that I'm particularly eager to see Quickley in a larger role. I think it's possible he could end up being a Jrue-like figure. Not a superstar, but someone quite valuable because of his two-way impact.

- AD, Westbrook, Wade. I'd say these are all guys that have something in common with Durant in the sense that they don't have top tier BBIQ and thus can be really impactful but are not always really impactful. With AD I'm actually a big fan of his defensive BBIQ, but offensively I think his impact is very much context dependent. With Wade I think he's got some great instincts, but he's neither a great shooter nor a great passer, and so when he lacked the ability - by body or context - to exploit his perhaps unmatched driving capacity, his impact was merely human. With Westbrook you've got a guy with massive BBIQ/stubbornness concerns - once he really got going in OKC, he was generally the team's MVP...but outside of that period, he's mostly a guy you don't want on a contender in practice. I've long argued that if he had continued to play the style he did at UCLA he'd be a more valuable player to contenders, but also concede he'd never win an MVP that way.

Re: but there's no reason to think Davis is overrated by it. Hmm. That logic seems to imply that if you know a guy is great, and he looks better than someone else by a certain stat, then the stat must have some authority, and I would say that's false.

In the end, PER is just clumsily adding up box score stuff and can't really tell the difference between Davis & Drummond, which means it's just not that effective of a tool.

see above for who is in Davis's company. also, how does it not tell the difference between Davis and Drummond? there is a vast gulf between their PER's.

Again, what I said was clear in the context I wrote it. I was responding to iggy, and you're confusing matters by taking the words as absolute statements.

i posted this in thread #32.

top players by playoff PER in history (age 21-35):
1. MJ
2. Lebron
3. Shaq
4. Hakeem
5. Kareem
6. Duncan
7. Kawhi
8. Barkley
9. Dirk
10. Durant
11. Robinson
12. CP3
13. Wilt (extended age range sunk him)
14. Magic
15. Jerry
16. Steph
17. Kobe
18. Pettit
19. Wade
20. Karl
21. Harden

all of the top 21 except the massively injured kawhi already in. we even have
31. Reggie
35. Ewing
36. Stockton
38. Frazier

all just or about to be voted in right in order

how many other stats are tracking the top 100 this well (with the acknowledgement that there's nothing that say the Top 100 is perfect either)? and that's just a raw list. not factoring in longevity.

My point is not to argue for any specific stat but to advocate caution when trying to reduce the game down to such stats.

I personally have no issues with the usage of any all-in-one stat as a starting point (with the exception of Wins Produced, for those who remember that toxic cult whose Jim Jones figure send his community in absolutely the wrong direction), but there's more to the game than can be captured by any stat we have.

I could go through that list above and speak to each player about how I think the stat is overrating or underrating the player in question, but it wouldn't be a PER-only thing. I could do the same for any single stat, not because I'm so smart, but just because I've considered all sorts of stuff along these lines, and have reasons for the conclusions I've drawn which do not lineup exactly with any algorithm.

[Doctor MJ]

My vote is the same as last time:

Induction Vote 1: Walt Frazier



I think my previous posts where I showed how massive the gap is between Frazier and his teammates really hammers in how I'm seeing things. I completely respect others who value longevity more, but I do see Frazier as a clear cut alpha superstar leading the top team of his era, and that's not something I can say about the other guys on the board.

Induction Vote 2: Reggie Miller

Arguing for Reggie a good bit over in the Reggie-Top-10-season thread.

Nomination Vote 1: Manu Ginobili



Bumping him up. I think what I said is still basically relevant, so I'll just un-spoiler for the event.

Yup, there I go. Manu's my next man.

Now as I've said, I'm less fixed on where exactly Ginobili is than I am feeling compelled to spread the gospel on the guy. I'm not purposefully doing that early - but it's possible I'll end up raising someone else above him before all is said and done here.

I quote my posts from the '04-05 thread before, and it's not just a coincidence they come from there. As I was going back through the years evaluating POY, I ended up siding with Ginobili at #1. This actually shocked me, and it's incredibly funny talking about it here, given that I was first compelled to post at RealGM during that same season to argue for Steve Nash's MVP worthiness and that Nash has since become my all-time favorite player. I like Ginobili, but Nash is the one who I truly fear having bias for. Perhaps I overcompensate, but in the past my '04-05 POY considerations were really about Nash vs Duncan.

To understand how I got there logically:

1. I think that Ginobili impacts with the best of 'em per minute and is typically held back in total impact by his limited minutes.

2. When a player's lack of minutes seems clearly to have held the team back meaningfully - like keeping team from chip - it's easy to justify knocking him harshly for the lack.

3. But when the team wins the title, and does so on the back of how he plays when he does play, I need to seriously ask myself where the minute threshold is that would have been "enough" to be the most valuable player.

4. And so, in my analysis, I would say that Ginobili would be my pick for both the WC MVP & Finals MVP.

5. This is happening in a season where there Ginobili leads the league in +/- by a significant margin:

Ginobili +844
Duncan +765
Nash +728

6. Speaking today, I now believe with confidence pace & space is a just plain superior way to play to win, and Ginobili was the guy driving the pace (+5.5 Pace On/Off in the playoffs) and the space (made more 3's than any other Spur), which I think was likely critical to their success against the Suns in particular.

Without elaborating on Ginobili vs Duncan & Ginobili vs Nash specifically at this time, I'll just say:

I see compelling arguments for Ginobili against each of them, and I struggle to use minutes to negate them.

Okay then zooming out, I see Ginobili as a guy who at his best was really capable of being the top basketball player in the league. He's held back some due to his limited minutes...but he also shows a remarkable tendency to level up his impact in the playoffs, and in particular deep in the playoffs.

And when the Spurs won titles in the 2000s, it always coincided with Ginobili seeming to go particularly nuts with his impact. All 4 of those chips, Ginobili had the best team +/- - and for perspective with the data we have, we don't have any other player more than twice. (Though we do have Michael Jordan twice, during his final two chips, which bodes exceptionally well for what we may find when we get access to earlier data.)

I'm honestly not sure if there's any other player remotely like Ginobili on this front - demonstrating this level of deep playoff impact dominance, while not being something like a GOAT candidate.

Okay, imma stop there. I hope my plea doesn't end up making folks recoil. I'm less concerned with convincing people right now that they should Nominate Ginobili, and more hopeful they'll just chew on their assessment of the guy. I think we have some significant things to learn about basketball, and basketball analysis, by understanding the the time of the Argentine.

Nomination Vote 2: Rick Barry

Next guy on my list. Definite legend. Have some criticisms of him that keep him from going higher, but I remain highly impressed by him.

[AEnigma]

Alright, yes, Manu will and should go well above Kevin Willis despite a substantial gap in minutes, but that is not the scale of comparison. If Manu is a regular “weak MVP” player despite his minutes, then sure, prop him up against Frazier or Pippen, but for me, he is more typically in that lower all-NBA group with Pierce (2005 postseason aside).

I've never made the case for Ginobili over Frazier in this project. I was in debate between Ginobili, Pippen, Stockton, Miller, Havlicek, and others, but I had Frazier distinctly above all of them.

Personal point of distinction with Frazier acknowledged, but my point here is more that if people are pushing Manu for his prime value, then I feel like that is roughly the level of star he needs to be given his minutes. Manu has eight regular seasons where he played more than 24 minutes a game (i.e. half the game — pretty paltry bar), and 2009 was basically a throwaway year. You have stated you see his case more tied to the playoffs, and there he does increase his minutes (as key rotation players do) so that now he has a twelve-year stretch (eleven postseasons) where he crosses that 24-minute mark. Of course, in a playoff setting I think 24-minutes is a pitifully marginal cutoff, and then I note he only has six postseasons above 28 minutes a game. He still has value in those surrounding years — two championships and another narrow Finals loss — but we are talking about a player who is typically fourth or fifth on his team in postseason minutes, and the one series where he was top two (led the team) produced an ignominious sweep and the worst playoff “impact” of Manu’s prime.

Now, I am sympathetic to the idea that the Spurs were typically deepest on the wings between Bowen, SJax, Brent, Finley, Kawhi, Danny, etc., and that the 2010 exception was tied to Parker and George Hill splitting time for similar reasons. I have a hard time saying it is to Manu’s credit that for the first ~five years of his career he was playing fewer minutes than Bruce Bowen, but hey, if it were someone like Pete Myers then that would not have happened. However, if we want to say Manu could have played more than he did on worse teams (leaving a question of how much more and to what extent that may have affected his career length or general availability), and that swapping Pierce and Manu would to some extent close the per minute gap, then we also need to admit it is easier to exert “impact” in more select rotations and that some hypothetical Manu matching Pierce’s minutes would almost certainly not be as productive on any possessional basis.

To reiterate, we are for the most part referring to a true prime stretch of seven or eight regular seasons and six postseasons, and the only way I think that justifies his placement here is if he truly was a regular top ~five force throughout that period rather than more of an all-NBA one. And while I can acknowledge the framing of him as a valid top five choice in 2005 specifically based on his outlier postseason stretch, realistically, I am not taking Manu over Duncan, Garnett, Nash, Shaq, or Dirk on any random team. Furthermore, if we replayed that postseason a hundred times, I am skeptical Manu’s average production would provide more than Wade, McGrady, or Kobe (placed on a better team), and to be really honest, I am similarly uncertain that he would be clearly surpassing lesser peak wings like Pierce or even Vince either in total per game contribution on otherwise identical teams. I am willing to mark Manu with a higher peak than those two based on what actually happened, but take away the shooting variance and I am not sure that distinction would be anywhere near as easy to make.