RealGM 2017 Top 100 List: #15

[andrewww]

I have the Logo as not only the most impactful individual player remaining with his offense, defense, shooting and as a post season performer.

The real debate is after him imo, and it comes down to Dirk, Admiral, Chuck, Moses, Dr. J. Admiral is the center version of KG, Dirk has the go to scoring, Dr. J with his peak years in the ABA but was most impressive.

Vote: Jerry West
Alternate: Dirk Nowitzki

[Doctor MJ]

I think if Karl Malone got somewhat of a pass, you can give something to Robinson.

Speaking as someone considerably more impressed with Robinson than Malone, I don't think you should assume that those who will vote for Robinson gave Malone a pass.

One of the hardest things to weigh in a project like this. Our eyes and our instincts tell us guys like Robinson and even Barkley in my case, are superior players to Karl Malone, but when you put Malone's career on paper you see how significant it is, how utterly unrivaled it is by guys like Admiral and Chuck, KG, Dirk, Pettit etc. And begrudgingly you almost have to give the Mailman the nod, or at least be willing to accept it if he gets it.

I think this years list is the best one Real GM has done so far.

Here's how mine compares
1) Jordan
2) Russell
3) Jabbar
4) James
5) Magic
6) Wilt
7) Duncan
8) Bird
9) Shaq
10) Mikan
11) Kobe
12) Hakeem
13) Doctor J
14) Oscar
15) West
16) Moses
17) Pettit
18) Mailman
19) KG
20) Barkley
21) Dirk
22) Durant
23) Baylor
24) Robinson
25) Barry

Malone certainly accumulated big, big stats. I think they inflate perception of his impact though.

Also, just so you know it's coming:

I think Baylor is really, really overrated.

[drza]

Some of the net on/off/- data for players under consideration (or that just got in)

For those that weren't around when this came out before, the 76ers statistician started keeping +/- data on the 76ers in the 70s, and for the whole league starting in 1994. fpliii discovered this, reached out to the guy, and he sent him that data a couple of years back. That led to a flurry of activity, with guys like Lorak and coltsfan18 (among others) taking the raw data and making on/off (and even some rudimentary RAPM-type estimates) based on it. A lot of the guys that are currently under consideration (or that recently got in) had peak/prime years in that time range, so I'll post some here.

With on/off +/- data, I look for a few things:

1) The magnitude of the overall net number
2) Whether that value led the team or not
3) If leading the team, the difference between them and the 2nd place guys

I've found that leading a team in on/off +/-, with a big number, and a big difference between 1st and 2nd, is what translates most faithfully to huge RAPM numbers.

If a guy has a big number, but has teammate(s) with similar scores, it could mean that the unit (or player combo) is strong and not necessarily that the individual player is carrying things.

If a guy isn't leading the team, it's even less likely that he's carrying the team on a solo mission. So, without further ado:

Robinson
1994:+19.9 (1st in NBA; 2nd on team = +10.3)
1995:+19.8 (1st in NBA; 2nd on team = +12)
1996:+16.6 (1st in NBA; 2nd on team = +9.2)
1999 PLAYOFFS: +36.7 (2nd on team +10.6)
2001 PLAYOFFS: +24.9, 2nd on team (Duncan 1st, +38.8)

Olajuwon
1994: +14.5 (2nd on team = +4.8)
1995: +11.9 (2nd on team = +6.1)
1996: +10.3 (2nd on team = 9.9)
1997 PLAYOFFS: +7.8, 2nd on team (Drexler 1st, +17.9)

Karl Malone:
1994:+17.6 (2nd on team = 7.3)
1995:+9.6 (2nd on team = 7.9)
1996:+13.5, 2nd on team (Stockton 1st at +14.5)
1997: +16.1, 2nd on team (Hornacek 1st at 17.6)
1997 PLAYOFFS: +20.3 (2nd on team +16.5)
1998: +13.2, 2nd on team (Stockton 1st, 13.4)
1998 PLAYOFFS: +5.4, 2nd on team (Hornacek +6.3)
1999: 13.3 (tied for 1st on team w Hornacek)

Charles Barkley:
1985: +1.9
1986: +10.6
1987: +7.9
1988: +2.7, 6th on team (1st on team +7.9)
1989: +11 (2nd on team = +6.5)
1990: +8.3 (2nd on team = +7.6)
1991: +8.8 (2nd on team = +5.3)
1992: +6.0 (2nd on team +5.9)
1994: +6.8, 2nd on team (1st KJ, +7.5)
1995: +6.8, 2nd on team (1st +8.4)
1996: +7.8, 2nd on team (1st KJ, +7.9)

Julius Erving:
1977: +6.0
1978: +0.6
1979: -0.1
1980: +3.0
1981: -6.7
1982: +10.0
1983: +10.3
1984: +4.4
1985: -3.2
1986: +4.6
1987: -3.8

Moses Malone:
1983: +15.6 (1st on team)
1984: +3.3
1985: +21.7 (Mo Cheeks at 14.1, Bobby Jones +10.4)
1986: +7.2 (Mo Cheeks led team at +20.3)

Some thoughts:
Robinson looks the most impressive, with league-leading scores and big differences between he and 2nd on team. Robinson also had a heck of a postseason +/- run for the 1999 championship team.

Olajuwon was right there with him, with slightly lower magnitudes in 94 and 95 but similar gaps between he and 2nd on team. In 96 he led the team, but the margin was closer. By 1997, when NBA.com has +/- data, Drexler had the better postseason score.

Karl Malone had an outstanding score in 94 with huge separation. After that, his magnitude was really good but there was little/no separation between he and either Stockton or Hornacek most years. He had a nice playoff +/- run in 97, but again not huge separation.

Moses was really impressive in 2 of his 4 Philly seasons (including 1983), leading the team in both years with seemingly solid separation. In the other two years, not so much. But his scores in the odd years were more impressive than anything I would have expected for him, based on his style of play.

For Barkley, we don't have his 1993 MVP year. But his other scores were just solid, a clear step behind Malone and way behind Robinson and Olajuwon. He led the 76ers, but with only a moderate score and usually without much separation, from 1989 - 1992. In Phoenix, his scores were moderate and slightly behind teammates (usually KJ).

As for Dr. J...this was the most stunningly negative outcome of the data to me. I don't believe that he was ever a team leader in this particular stat, from 1977 - 1987. He had two moderately solid years in 82 and 83, but at least in 83 he was clearly behind Moses. And he had some years where he came up out-right negative (which, I should note, does NOT mean that he was a negative impact player...often, it means that another teammate/unit is much more successful and thus leads to negative results for a player...but still).

Just some +/- numbers from the pre-databall era (or at least not stuff found on basketball-reference), for some players under consideration.

Vote: David Robinson
2nd: Jerry West

[mikejames23]

To be honest, Karl Malone shouldn't have gotten that pass to begin with. No one really responded to his playoff drops compared to Dirk or West.

Anyway, even though I generally believe D-Rob's offensive metrics to drastically overrate his actual offensive ability (borne out in the playoffs) and I've been low on players whose offense can't stand up in the playoffs, I could actually take him over Malone because of all the other stuff he brings to the table. It's kind of like KG - his scoring tails off, more so than Malone in fact, but D-Rob also provides far superior defense as a legitimate defensive anchor.

The thing is, someone (trex?) posted that the Spurs defenses underperformed a little bit in the playoffs, and even in 1995 they from -2.9 to +0.9 vs Houston - they were -5.3 over the entire playoffs because they destroyed the Lakers and Nuggets. I've brought this up before, but is it possible that D-Rob's defense falls off along with his offense? I could definitely be talked into thinking that.

Yeah, my thoughts on Dirk/West are similar. I don't see anyone left that is as resilient in the playoffs as either. It's why I haven't really bothered with posting a case for D-Rob.

Defense - I tried searching for this in old realGM posts but couldn't find. Can someone repost Robinson's playoff defense drop off? If it's still around somewhere. Frankly if the Spurs defense did get worse, I am not certain how much of that really is on Robinson. There appears to be no real weakness in his defensive game that enables a collapse to happen? He's got the movement speed, his paint defense is better than anything I've seen, has high IQ and incredible reaction timing for that kind of size. It'd be more than anything on external factors or fatigue issues based on bad Spurs offense in the playoffs - Robinson was sometimes the only real scoring option they had.

[Outside]

Some of the net on/off/- data for players under consideration (or that just got in)

...

Vote: David Robinson
2nd: Jerry West

I'm curious -- for someone like yourself who relies heavily on these metrics that aren't available for players in earlier eras, how do you evaluate a player like West?

[drza]

Some of the net on/off/- data for players under consideration (or that just got in)

...

Vote: David Robinson
2nd: Jerry West

I'm curious -- for someone like yourself who relies heavily on these metrics that aren't available for players in earlier eras, how do you evaluate a player like West?

These metrics are meant to help us, to the best of our ability, estimate how much impact a player is having on their team's fortunes. Where +/- doesn't exist, I look at everything else. I appreciate the WOWY work that ElGee has done, and I've made use of it in this project, like in my Shaq vs Wilt post http://hoopslab.rotowire.com/post/163011568256/impact-comp-of-the-giants-wilt-vs-shaq or my Kareem vs Duncan post http://hoopslab.rotowire.com/post/162535870741/kareem-vs-duncan-peak-impact-and-functional.

Sometimes we don't have even any consistent WOWY runs for players. In instances like that I try other methods to put some type of quantification to the other information that I can gather. Like, for instance, when I looked at team offensive/defensive ratings for the Celtics (Russell) and the Bucks/Lakers (Kareem) in the years before the superstar bigs, to after they came, and with confounds like other players (e.g. Oscar and Magic, on Kareem's offenses) like I did in the Russell vs Kareem post http://hoopslab.rotowire.com/post/162660433401/abdul-jabbar-vs-russell-observations-and-rough.

So, I've been pretty consistent in trying to quantify impact in the best way that I can for players across eras. I don't look at it like I have to do a direct comp of RAPM score vs RAPM score to be able to evaluate. If player A has +/- data, it helps me to have more confidence in the level of their impact. But, failing that, I can still do the best I can to get an idea of the level of impact a player is having. And, since this is an approach that yields itself to noise, I have to be satisfied with that and broaden my comparisons. For example, in the Peaks project, I pointed out that late-70s Walton maximized the WOWY scale while mid-2000s Garnett maximized the RAPM scale...it's not a direct comp, but generally speaking I'd say that puts their relative impacts on a comparable level.

For West specifically, as I mentioned last thread, his WOWY scores are a) very strong and b) due to his injuries, he has a lot of WOWY runs that are extended...which helps build confidence that the estimated impacts are reasonably robust. So, this makes another example where a past player (West) came close to maximizing the WOWY impact scale while a more modern player (Robinson) came close to maximizing the +/- scale. In rough terms, I take that to mean that they had impacts of at least comparable degree. I incorporate that into the other aspects of my analysis, and go from there.

The pure number of RAPM or on/off +/- or WOWY or whatever is never the important factor...the ability to quantify (to at least a reasonable extent) the caliber of a player's impact, and to be able to put that in perspective with context and other information, is much more vital to my evaluation process.

[Outside]

Some of the net on/off/- data for players under consideration (or that just got in)

...

Vote: David Robinson
2nd: Jerry West

I'm curious -- for someone like yourself who relies heavily on these metrics that aren't available for players in earlier eras, how do you evaluate a player like West?

These metrics are meant to help us, to the best of our ability, estimate how much impact a player is having on their team's fortunes. Where +/- doesn't exist, I look at everything else. I appreciate the WOWY work that ElGee has done, and I've made use of it in this project, like in my Shaq vs Wilt post http://hoopslab.rotowire.com/post/163011568256/impact-comp-of-the-giants-wilt-vs-shaq or my Kareem vs Duncan post http://hoopslab.rotowire.com/post/162535870741/kareem-vs-duncan-peak-impact-and-functional.

Sometimes we don't have even any consistent WOWY runs for players. In instances like that I try other methods to put some type of quantification to the other information that I can gather. Like, for instance, when I looked at team offensive/defensive ratings for the Celtics (Russell) and the Bucks/Lakers (Kareem) in the years before the superstar bigs, to after they came, and with confounds like other players (e.g. Oscar and Magic, on Kareem's offenses) like I did in the Russell vs Kareem post http://hoopslab.rotowire.com/post/162660433401/abdul-jabbar-vs-russell-observations-and-rough.

So, I've been pretty consistent in trying to quantify impact in the best way that I can for players across eras. I don't look at it like I have to do a direct comp of RAPM score vs RAPM score to be able to evaluate. If player A has +/- data, it helps me to have more confidence in the level of their impact. But, failing that, I can still do the best I can to get an idea of the level of impact a player is having. And, since this is an approach that yields itself to noise, I have to be satisfied with that and broaden my comparisons. For example, in the Peaks project, I pointed out that late-70s Walton maximized the WOWY scale while mid-2000s Garnett maximized the RAPM scale...it's not a direct comp, but generally speaking I'd say that puts their relative impacts on a comparable level.

For West specifically, as I mentioned last thread, his WOWY scores are a) very strong and b) due to his injuries, he has a lot of WOWY runs that are extended...which helps build confidence that the estimated impacts are reasonably robust. So, this makes another example where a past player (West) came close to maximizing the WOWY impact scale while a more modern player (Robinson) came close to maximizing the +/- scale. In rough terms, I take that to mean that they had impacts of at least comparable degree. I incorporate that into the other aspects of my analysis, and go from there.

The pure number of RAPM or on/off +/- or WOWY or whatever is never the important factor...the ability to quantify (to at least a reasonable extent) the caliber of a player's impact, and to be able to put that in perspective with context and other information, is much more vital to my evaluation process.

So would it be fair to say that you do your best to create or estimate equivalent data for players from earlier eras?

If so, I guess that's an opposite approach to the one I take, where I rely more on basic stats and observation to develop a subjective analysis (what could be called informed opinion) and then use additional stats like RAPM to confirm or adjust my thinking. But where you use RAPM, WOWY, a lot of other stats to create your baseline, I use basic stats and the eye test to create my baseline.

While you voted KG, Dirk, and David Robinson higher than me, which is to be expected since they do well in the metrics that you use, and I have guys like Bird and Magic higher on my list, but we're generally voting for the same guys, give or take a slot or two. I was skeptical at the start that people who relied on advanced stats would give a fair shake to players from early eras, but you've certainly done that.

The amount of work you put in to your posts is impressive, but to me, that's only the tip of the iceberg. Digging a little deeper, it's apparent that you don't rely solely on the stats but use them as the basis for thoughtful, extensive analysis, which I particularly like. I've also seen that with each new thread, you often begin your process over again and may make different selections. I don't begrudge those who set up a formula and use the results to create their ranking -- creating a basketball version of a unified field theory formula is an admirable goal -- but the time required for your process is a whole 'nother deal. I'm confident enough in my reasonably informed subjective assessments, but I'm basically spitballing compared to what you're doing. I'm not in a position to use the same process you do, but that doesn't mean I can't appreciate your artful blend of data and analysis. It's been an education.

[Joao Saraiva]

1st vote - Jerry West

Great shooter and floor spacer, way ahead of his time. Good slasher and defender.

Was also a tremendous playoff performer. In several of his losses he played really well, so I've got nothing against him for winning only one finals series and losing several of them.

Really consistent.

Very efficient - lead the league twice in ts% without 3P shots - that's something great for a guy who shot as much from the perimeter as he did. And he did it while putting up great volume.

He has 11 RS of great production and playing almost all the games (well missing 10 takes some value from it but it's not something that would bother me).

He also has 10 post seasons where he did play very well.

2nd vote - Dirk Nowtizki

[trex_8063]

Some of the net on/off/- data for players under consideration (or that just got in)

.......

Vote: David Robinson
2nd: Jerry West

With impact and impact data being such massive considerations in your criteria, I'm sort of surprised Dirk isn't your alternate pick or at least one of the guys you're considering (not listed in above post, anyway). Beyond Shaq, Lebron, KG, and Duncan, he's pretty much been the biggest impact player of the last 20-some years (as measured by RAPM, anyway).

[drza]

Some of the net on/off/- data for players under consideration (or that just got in)

.......

Vote: David Robinson
2nd: Jerry West

With impact and impact data being such massive considerations in your criteria, I'm sort of surprised Dirk isn't your alternate pick or at least one of the guys you're considering (not listed in above post, anyway). Beyond Shaq, Lebron, KG, and Duncan, he's pretty much been the biggest impact player of the last 20-some years (as measured by RAPM, anyway).

He is. I didn't want to just copy-paste my vote considerations from last thread so as not to get stale, so I posted the +/- info instead. But, since the guys I voted for didn't get in last thread, I could copy-paste and you'd see Dirk was there. In fact...

My main methods of evaluation thus far have been impact-oriented, trying to identify which (among the sea of great NBA players) have done the most to help their teams succeed. Evaluating this is more difficult once we get before the databall era, but with the +/- data from the 76ers statistician that fpliii gathered and the WOWY work that ElGee has spearheaded we've got more tools on that front than we have for any previous project.

Of the players left on the board, it seems to me that West, Robinson and Dirk have the best impact cases left. Dr. J, Malone and Malone also have strong cases in general, but there are more questions there. I'll get into them more in future threads (any that don't go in here), and they have the chance to move up my list. But coming in late without much time, for this thread I'll focus most on West, Robinson and Dirk.

Spoiler:

West's WOWY results support that his abilities as a scoring team offense initiator made him one of the biggest impact players of his era, or any other. His game also seems very translatable across era, as there's little doubt in my mind that he would have been able to use the 3-pointer as a weapon to make him even more effective in the modern game than he was in his own. Injuries are a big concern for West. I could consider them as another area, though, in which West would have benefited from the more modern era with better medical and training techniques. That's less tangible, though. What is tangible is that West, along with Oscar, was one of the greatest offensive players of his era with a separation from anyone else of that time. That West had a lot of injuries is a negative, but those injuries also gave us plenty of chances to evaluate what his teams looked like without him...and they struggled. With him, they were great. And he did have a history of putting up big boxscore numbers in the postseason that, without any method of estimating postseason impact, defaults to looking really elite considering his measured/estimated regular season impact.

Robinson was electric, to my eye test. His tournament run at Navy was some of the most exciting individual play I remember in the NCAA, and when he burst onto the scene in the pros after his tour of duty was up, he immediately looked like one of the best players in the NBA. Before the RPoY project I always felt like Robinson was overly downgraded for the Hakeem series in 95, but in that RPoY project several posters (especially Kaima) did a great job of pointing out how Robinson relatively struggled in 94, 96 and 98 against Karl Malone and the Jazz and used that as a basis for arguing that Robinson's playoff issues weren't just a Hakeem 95 thing, but a systemic issue. Subsequent research, posts and project discussions about Robinson's mechanisms of impact have been convincing that Robinson's game really does have tangible difficulty to be the focal iso-scoring lead in the playoffs...and that his overall offensive game wasn't diverse enough to maintain his offensive impact in the postseason. There is even some evidence that in the postseason in his peak, while trying to carry the load on both ends against some tough competition, Robinson's defensive impact slid a bit as well. These are all issues.

However, we have more information than that to work with to try to peg Robinson's level. It shouldn't have come as a shock to anyone, but Robinson's regular season on/off +/- data did peg him as the highest regular season impact player of the mid-90s (94 - 96). That's expected, but it is good to be able to quantify that. However, we also have quantitative impact estimates for another time period that is often minimized/ignored for Robinson...the 98 - 2000 period that has historically been considered the "Duncan era".

While it is unarguable that Duncan's presence as the focal point of the Spurs was huge in bringing the Spurs to championship level, and probably made the game much easier for Robinson...and while one could also strongly argue that Duncan may have been the actual leader of those teams and the player that opponents game-planned for...it's ALSO clear from the RAPM results that Robinson was having just about as much impact on the scoring margins of those 98 - 00 Spurs as Duncan was. Robinson was the defensive anchor on those teams, and with Timmy there as another offensive focus Robinson's offense was also able to flourish. And even in the postseason, the available on/off +/- numbers suggest that Robinson was able to maintain his huge regular season impact into the postseason in this era. Again, when we compare Robinson's postseason impact in the Duncan era to his impact at his peak, I think we have to credit Duncan's presence with making the game easier for him to maintain his best impact. However...that doesn't disqualify the impact itself. And I think that it also suggests that, while a team wouldn't want to move forward with Robinson as their focal offensive player, that a team that allows him to play to his strengths would be getting very possibly the highest impact player in the NBA in both the regular and post season.

Dirk played his whole career in the databall era, so we have the ability to measure/estimate his impact at a granularity that isn't available for previous generations. In both the regular and the postseason. He was a unique beast as a player, in that there isn't really a template for how a 7-footer with the offensive/scoring abilities of an elite wing can affect the game. What we saw, in practice, was that Dirk was able to make a very high impact on the game for much of the 2000s...he was regularly among the top 10, if not top-5, in the league as far as individual regular season impacts while carrying very successful teams.

But, while he actually improved on many of his boxscore stats in the postseason, he didn't have a very strong postseason impact run for most of the 2000s. I've argued, in the past, that those lower playoffs +/- numbers may have come from teams not distorting their defenses to defend him in the postseason the way that they did in the regular season. That teams may have been more willing to play Dirk straight up and let him get his as he could in the postseason, which allowed him to improve his scoring volume/efficiency, but may have weakened his spacing impact. That some of the teams would go so far as to defend Dirk with a wing instead of a big man, which further may have allowed him to get his but weakened his spacing impact on his team's results. And (as I've been arguing more and more in recent times), I think that a player's impact on team results can be much more important than his own scoring numbers, when it comes to estimating his overall impact.

Up through most of the 2000s, then, I'd argue that Dirk's seasons were similar impact-wise (in both the regular season and the playoffs) to what we saw from 90s Karl Malone. Malone's postseason scoring numbers dipped, unlike Dirk's, but again I don't know that Dirk's boxscore numbers were indicative of his actual impact. And in the larger regular season samples, both Karl (as measured by the available +/- numbers from 1994 on) and Dirk measured out in that top 5 - 10 players in the NBA range in the regular season and had trouble maintaining that in the postseason.

However, in the late 2000s up through 2011, Dirk upped his game. He added a more effective iso post game on offense, that took him out of the range of 7-foot scoring wing and gave him a legit big-boy component to his game. Thereafter, teams that tried to defend him with Stephen Jackson types would get punished. Also, I'd argue that by diversifying his scoring mechanisms an operating more from the post, it changed Dirk's center of offensive gravity and caused defenses to have to react to him in a more dynamic way. His shooting still provided interior spacing by bringing folks out, but his post-game could pull defenses more into the paint and make life a breeze for guys like Jason Terry. And his ability to partner the pick and roll/pop did the same, most famously around here for guys like J.J. Barea. And it's this late-prime modification to his game that separates Dirk from a player like Karl Malone, who also developed and diversified his offensive game in his late-prime to become a bigger offensive impact player...but never was able to do so to the full extent/effectiveness demonstrated by late-prime Dirk.

Overall: each of West, Robinson an Dirk had both a demonstrated/measurable all-time level impact that neither Erving, the Malones, or anyone else up in this category were able to match. However, each also had warts/limitations that kept them from going even higher on this list. To me, that means that right here is the exact right time to be discussing all three. And I'm not sure which is the correct order for the three. However, at the moment, Robinson's all-time defense in addition to his ability to operate at high level impacts as a secondary offensive threat (which is actually ideal for almost all big men) seems like the most effective of these three in the widest array of possibilities. West's injuries bother me, especially compared to a relative iron man like Dirk, but he also seemed to maintain his impact better throughout the course of his career...perhaps because his impact didn't rely so heavily on how teams chose to defend him, the way that Dirk's did until his post-game came into focus. Again, could be argued in any number of directions here (including with players that I didn't focus on), but having come in at this late stage of the thread, for now I have to just pick an order to vote, and then hope to have better discussions moving forward.

Vote: David Robinson
2nd: Jerry West

[cpower]

If Malone has voted in already, why can't we have a case for Stockton who was generally a better PS performer than Malone who actually led the team in WS/48, on/off, offensive efficiency..etc? It seems like Stockton gets penalized for Malone's underwhelming performances which does not make sense to me.

[theonlyclutch]

If Malone has voted in already, why can't we have a case for Stockton who was generally a better PS performer than Malone who actually led the team in WS/48, on/off, offensive efficiency..etc? It seems like Stockton gets penalized for Malone's underwhelming performances which does not make sense to me.

If there's an argument for him being in the top 15, I would like a very solid reason for why 2 top 15 players in their statistical primes (88-95), with plenty of roster continuity and good coaching, were head of teams which aren't appreciably better than this year's Utah Jazz...

[trex_8063]

.

So I'd mentioned I was looking into the correlation of OREB% to ORtg (obvious relevance wrt Moses), as well as OREB% to DRtg. Am quoting Outside here, as during a discussion with him about Dirk [and his low OREB%] I'd theorized that a strategy of crashing the offensive glass is knowingly giving something up defensively (by way of reduced transition defense). The findings so far have been interesting.

I looked at rOREB% and rORTG/rDRTG for each and every team for the years '74-'85 (that's 256 data points over 12 seasons, covering most of Moses' prime/career), as well as each team for the years '04-'07 and '11-'12 (that's 179 data points over 6 seasons of Dirk's prime). That's as far as I've got so far; will try to go further, but there is other stuff I want to move on to, so I'm going to share what I've got so far. Anyway, I made plot-point graphs (with a trend line) of those data sets, and calculated the Correlation Coefficients for each set.

Of interest, the league avg OREB% of Dirk's career (at least the years I've investigated) run ~5% lower than that of the '74-'85 sample (~27.5% vs ~32.5%), and obviously there have been a lot of other game trend shifts. And the correlations seen between the two eras are very very different.

For the '74-'85 sample, there definitely appears to be a fair correlation between increasing rOREB% and increasing rORtg. The correlation coefficient is 0.3729.

I can explain correlation coefficients to the best of my limited ability if anyone needs (I've only somewhat recently had it explained to me and have begun using it in a few of my studies). But suffice to say this indicates there's likely a significant correlation. Not super-high, but it wouldn't be reasonable to expect a really high correlation coefficient (say 0.6-0.7 or higher) because we, after all, are ignoring all of the other offensive factors that influence offensive efficiency (namely: eFG%, FTr, and TOV%). tbh, 0.3729 was perhaps slightly higher than I was expecting.

The other interesting thing in this '74-'85 sample is that there appeared to be basically zero correlation between an increasing rOREB% and an increasing (worsening) rDRtg; the trend line is almost exactly flat along the x-axis. The correlation coefficient was 0.0093.
Obviously, there are a number of confounding factors not included in this study, so this is not "proof" that a strategy of banging the offensive glass didn't hurt transition defense; but for whatever it's worth, there was no correlation between the two for that sample of years.


With the 6-year sample from Dirk's prime, the correlations are VASTLY different. It would seem that during Dirk's career (based on these six years, at least) there is almost negligible correlation between OREB% and ORtg. The correlation coefficient was just 0.0879 (which is likely not statistically significant, even in light of confounding factors previously alluded to).

But interestingly, there DID appear to be correlation between a rising rOREB% and a rising (worsening) rDRTG in this set of years. The CC there is 0.2021, which is not high, but is high enough to suggest that there probably is a relationship between the two (and is likely why there has been a shift away from offensive rebounding to focus on getting back on D).

Dunno if this sort of exonerates Dirk wrt criticisms of his low OREB% or not; I leave that to each of you to ponder. But it was an interesting finding.

[THKNKG]

I've been convinced/convinced myself that the value of DRob/West trumps the longevity of Dirk.

Vote: David Robinson
Alt: Jerry West
HM: Dirk

[trex_8063]

Posting some of my arguments from last thread:

Spoiler:

Wow. That’s…….well that’s a tough sell (Dirk at #33). I read some of your argumentation for this (rebounding, lack of shot content coming <10 feet, deflated appreciation of the importance of big-man spacing, etc) and wish to reply to a few things.

Re: Rebounding (and spacing and shot selection criticisms, because they’re kinda related in Dirk’s case)
therealbig3 I think already responded to some of the rebounding, but I’ll just re-iterate that his DREB% and DRebs/100 are right in line with that of Karl or Charles or Moses (especially if we include his playoff numbers in our comparison). It’s his OREB rate that lags behind.

Some of this may be by design (get back on defense), and some is because he’s often operating on the perimeter (and thus not in position to bang the offensive boards). As has been stated, however, there’s a tangible offensive benefit to the latter factor of having a big man who is a threat from the outside (stretches the floor for the rest of the team; theoretically could raise the OREB% of some of his teammates, too, as he’s drawing a big defender out).

One could try to argue that the lack of individual offensive rebounding cancels out any benefit from spacing, but frankly, the burden of proof would be on you, given the offensive results for the Mavs during Dirk’s prime:

‘01 Mavs: 4th/29 ORtg, 29th/29 OREB%
‘02 Mavs: 1st/29 ORtg, 25th/29 OREB%
‘03 Mavs: 1st/29 ORtg, 27th/29 OREB%
‘04 Mavs: 1st/29 ORtg, 3rd/29 OREB% (addition of Danny Fortson likely primary reason for jump in OREB%)
‘05 Mavs: 4th/30 ORtg, 17th/30 OREB% (Fortson and Nash gone, added Dampier, but he misses 23 games)
‘06 Mavs: 1st/30 ORtg, 2nd/30 OREB% (Dampier healthy, added Diop)
‘07 Mavs: 2nd/30 ORtg, 8th/30 OREB%
‘08 Mavs: 8th/30 ORtg, 16th/30 OREB%
‘09 Mavs: 5th/30 ORtg, 16th/30 OREB%
‘10 Mavs: 10th/30 ORtg, 26th/30 OREB%
‘11 Mavs: 8th/30 ORtg, 26th/30 OREB%

Their offense was pretty consistently at or near the top of the league during Dirk’s prime, with or without Steve Nash and regardless of what their OREB% was. This is not to say that offensive rebounds don’t bring a lot of offensive value; but rather I’m suggesting that big man spacing does too. And again consider that his presence near the perimeter (drawing his defender out to him) could theoretically reduce the DREB rate of the opposing team (because one of their primary rebounders has been pulled out away from the basket); this could [again, theoretically] increase the OREB% of his teammates (while decreasing his own).

The other major factor to consider is that if you’re going to bang the offensive glass as a general strategy, you are (of necessity) going to give something up by way of transition defense. I don’t think it’s any coincidence that here in the modern----very scientific and analysis-based strategy/game philosophy----era that team offensive rebounding rates are the lowest they’ve ever been. As a general rule today, teams err more toward getting back on defense rather than banging the offensive glass.
However, it’s true that a lot of the lower OREB% are probably related to the increased 3PAr; but if so, that’s another fly in the ointment as far as comparing their rebounding rates straight up in the first place.

But suffice to say, it’s a difficult road to prove that Dirk as an individual averaging ~1.5 more offensive rebounds per game (some of them potentially cannibalized from his own teammates, if he’s no longer drawing an opposing big man out) would outweigh the loss of spacing and the sacrifices in transition defense.


Re: not enough shots from <10ft
I’ll be honest, this strikes me as a bizarrely arbitrary (read: personal preference) type of criticism. You’d noted how for his career (or since 2000, anyway) that only 22.8% of his attempts have come from <10 feet, iirc, making some derogatory implication (“allergic to the paint” or similar). The implication seems to be that playing in the mid-range makes him “soft”, and he should be banging away inside (because hey: you’re so tall!).

But I’m reminded of a quote from the Harry Potter series: “Play to your strengths, Harry.” How does it make sense to ask Dirk to play in a way that makes him LESS effective, just so he can better fit a “traditional” mold or expectation of a big man?

And further, why does this matter? You noted that (compared to Dirk’s 22.8% of attempts that come from <10 feet) that Tim Duncan had 30.9% of his attempts coming from <3 feet, and seemed to imply this makes Duncan intrinsically better as a big man scorer. How, exactly?

Single best scoring years (for rs).....
‘02 Duncan: 33.5 pts/100 possessions @ 57.6% TS (+5.6% rTS)
‘07 Dirk: 36.4 pts/100 possessions @ 60.5% TS (+6.4% rTS) *’06 is roughly equal to this, too

Duncan ‘98-’13: 30.6 pts/100 possessions @ 55.2% TS (+2.23% rTS)
Dirk ‘00-’14: 33.1 pts/100 possessions @ 58.4% TS (+5.22% rTS)

Despite any preferences on where the shot attempts come from, or how often that shot selection gets him to the FT line----(his prime FTr in rs is only slightly behind that of prime Duncan, btw, and his FTr in the playoffs in his prime is slightly HIGHER than prime Duncan’s)----it’s clear Dirk was a more effective scorer than Duncan.

And he had the spacing effect that Duncan generally does not. AND he generally scales UP in the playoffs (whereas most others do not).


And the the other MAJOR factor about Dirk’s offensive game is that he has a substantially lower turnover rate than just about everyone (this is in part related to his mid-range preference). I’ve railed against the mid-range shot in other places as a relatively low-efficiency shot (though ‘07 and ‘11 Dirk turned that generalization up on its head); however, I’ve also noted that a definitive benefit or advantage the mid-range game has going for it is that it carries a very low probability for turning the ball over.

Dirk has a ridiculously low turnover rate, turning the ball over just 1.8 times per 36 minutes played for his career. His career TOV% is just 8.5%.

One might try to counter by saying he doesn’t make a ton of assists or plays for others, though. And fwiw, I’ve always been annoyed that TOV% doesn’t factor in play-making at all. It’s formula is simply:

TO / [TO + TSA]

In response to my irritation, I’ve gone and generated a “Modified Turnover %”, which is as follows:

TO / [TO + TSA + (Ast * 2) + (Reb * 0.04)]

Note this includes assists (multiplied by 2, figuring for every assist made, there is another potential assist that doesn’t occur because the teammate missed the shot); and it also include total rebounds (multiplied by very small modifier) figuring a player is occasionally stripped or throws it away on the outlet pass, etc…...the modifier assumes this happens once every 25 boards.

So Dirk’s modest assist rate and his [in your opinion] mediocre rebounding rate works against him in this formula (relative to guys like Duncan, Barkley, or Karl Malone). And yet here are their respective career rs Modified TOV%:

Duncan - 9.26%
Karl - 9.50%
Charles - 10.64%
Moses - 12.90%
Dirk - 6.73%

I haven’t run EVERYONE thru this formula, but so far the ONLY player I’ve found whose career rs Mod TOV% is lower than Dirk’s is Chris Paul (6.30%). The only other guys I’ve found who are even close are Michael Jordan (6.82%), Horace Grant (6.91%), and Tracy McGrady (7.18%).
And guys like Duncan or Mailman obviously were NOT at all turnover-prone. And yet they still lag that far behind Dirk where turnovers are concerned.

Looping full-circle back to the lower offensive rebound rate criticism, I would say his lower turnover rate very nearly cancels out any deficit on the offensive boards (i.e. while he’s not retaining possessions as often via OREB, he’s also not turning the ball over as often, and by a similar number, too).


Anyway, coming back to the #33 thing……..
Objectively, I just don’t see how this is tenable. Dirk’s played nineteen seasons (the last few---years many players never got around to---are dragging his career avg’s down), and yet:
*He’s still 25th all-time in career rs PER.
**He’s 11th all-time in career playoff PER.
***He’s 22nd all-time in career rs WS/48 and is 8th all-time in career rs win shares.
****He’s 19th all-time in career playoff WS/48 and is 13th all-time in career playoff win shares.
*****He’s 18th all-time (or since ‘74) in career rs VORP.
******He’s 22nd all-time (or since ‘74) in career playoff VORP.

EDIT: I also have a formula to figure up cumulative value over a roughly replacement level player as measured by PER and WS/48 in both rs and playoffs (each playoff minute played weighted 3.25x heavier than each rs minute), assuming when they're not on the court it's a replacement level player subbing in for them......Dirk is 8th all-time by this formula.
I've noted there was greater parity in these metrics in certain earlier eras, and figured out some standard deviations for each (year-by-year); from this created a "scaled" PER and WS/48 for everyone, and again calculated cumulative career value above replacement level based on scaled PER and WS/48.......Dirk is 10th all-time by this version.
I have other much more complex formulations which include A LOT more data; without going into a ton of detail, I'll just state that those formulas rank Dirk 16th, 10th, and 15th all-time, respectively.

And one cannot say it’s empty stats (lacking impact). If we look at RAPM, best 10 years combined, only five players have a better 10-year combined (this is in the last 24 seasons, data back to ‘94 if we use colts18’s rs-only regressions; plus have RAPM for Barkley from ‘88 to ‘92, provided by Dipper, iirc): Shaq, Lebron, KG, Duncan, and *Manu Ginobili (*important to note Manu’s comes with a lot of minute-restriction, though).
Now given the data cuts off before ‘94 (except for Barkley), some players are missing huge chunks of their primes (DRob, Jordan, Hakeem, Mailman, Stockton, etc). But if we look at the best 7-years combined, it’s still just the same five players who had better than Dirk.
If we look at best 5-year combined, it’s again the “pantheon four” of Lebron, Shaq, KG, and Duncan, plus *David Robinson [*barely, and with ps numbers excluded from ‘94-’96].
If we look at best 3-year combined, only “the pantheon four” rated higher than Dirk.

In short: the impact is there (probably even goes at least marginally above what his box-based metrics indicate).


Additionally (narrative/accolade stuff)…..
*He’s been an MVP and is 26th all-time in NBA MVP Award shares.
**He’s won a title as “the man”, and won FMVP.
***He’s a 13-time NBA All-Star.
****He’s made 12 All-NBA teams (4 1st, 5 2nd, 3 3rd). Note this is also in one of the toughest eras in NBA history, and with a prime that exactly coincides with all or some of the primes of Lebron James (voted in #3), Tim Duncan (voted in #5) and Garnett (voted #12), as well as forwards like Pau Gasol, Chris Webber, Paul Pierce, Elton Brand, Chris Bosh, Carmelo Anthony, Shawn Marion (Tracy McGrady occasionally counted as a forward in award voting, too), and later Kevin Durant.


This is simply not the resume of a player who is appropriately ranked at #33 all-time. There are just NOT thirty-two careers better than what I’ve just outlined. Frankly, there are not even twenty-two careers better than that.

Also note I addressed some of the OREB concerns in my above post itt.

1st Vote: Dirk Nowitzki
2nd Vote: Julius Erving (though could be swayed to David Robinson)

Just by way of throwing a little reasoning for Dr. J.....

By those formulations I'd mentioned that measure total career value over approx replacement level player as measured by PER and WS/48 (playoff minutes weighted 3.25x heavier than rs): Erving ranked 9th all-time by the formula using raw PER and WS/48, 8th all-time by the one using scaled PER and WS/48 values (from my SD studies). However, some of that might be inflated from his ABA years, depending on how you feel about the strength of the ABA. I do have some year-by-year strength of era ratings of my own devising, which I applied to these studies. Doing so dropped Erving to 16th if using the raw PER and WS/48, only down to 9th with the scaled PER and WS/48.

[mikejames23]

If Malone has voted in already, why can't we have a case for Stockton who was generally a better PS performer than Malone who actually led the team in WS/48, on/off, offensive efficiency..etc? It seems like Stockton gets penalized for Malone's underwhelming performances which does not make sense to me.

Eh?

I do find Stockton interesting. His own era, however, never considered him to be this high. A whole bunch of MVP finishes outside top 10 and never cracked Top 5 MVP finish. Based on that pattern you'd have easily 10 others ahead of him.

Elgee doesn't appear to be high on him - if he posted more numbers on Stock that'd be awesome. Here's one gem -

https://elgee35.files.wordpress.com/2010/12/stockton-elimination1.jpg

More

viewtopic.php?f=344&t=1343246#start_here

What's interesting to me is how he comes at the top defensively in late career metrics - he even kills rising Nash from 01-03 by a significant margin. Anyway, I think this is a guy you discuss at #25 and not #15. Just not at the level he's being discussed.

[wojoaderge]

Not much more to say about George Mikan other than he's my first vote once again. To repeat, no one left on the board was as dominating in his particular time or served as the go-to guy on as championship teams as he did.

Now for my alternate. Moses - I just have no idea why he's so underrated. He's easily more dominating, more overpowering, and more impossible to stop than any of 5s or 4s remaining. 3 MVPs. He led a 65-17 team to the championship as its best player and led a sub-.500 team to the NBA Finals (only the 2nd one to date). If you want to talk longevity, he had at least 16 straight good to awesome seasons. So, I don't know the problem is here.

1-George Mikan
2-Moses Malone

[trex_8063]

.

.

When I get home, I'll dig up my materials that pertain to the popularity of basketball from the mid-50's thru mid-70's (as well as the limited amount I have on integration), as I think that bears direct relevance to the size of player pool (and thus quality/competitiveness of the league). imo, very relevant stuff wrt Mikan. Will try to get it posted this thread or next.

[Doctor MJ]

If Malone has voted in already, why can't we have a case for Stockton who was generally a better PS performer than Malone who actually led the team in WS/48, on/off, offensive efficiency..etc? It seems like Stockton gets penalized for Malone's underwhelming performances which does not make sense to me.

This is a reasonable question but as someone who thinks Malone was voted in too early in part because of Stockton, it chafes to see Stockton potentially lifted up based on the prior mistake.

Ftr, Stocktons case has grown for me over the years. I've had him behind Nash for a long time, but I'm hoping I reconsider this objectively.

The reality is that basically no one had such consistently impactful longevity like Stockton.


Sent from my iPhone using RealGM Forums

[trex_8063]

Thru post #59:

Jerry West - 5 (andrewww, Doctor MJ, Joao Saraiva, Outside, RCM88x)
Julius Erving - 4 (Winsome Gerbil, scabbarista, Pablo Novi, Dr Positivity)
David Robinson - 3 (drza, micahclay, Narigo)
George Mikan - 2 (penbeast0, wojoaderge)
Moses Malone - 1 (JordansBulls)
Dirk Nowitzki - 1 (trex_8063)

Thread will be open for at least 2-3 hours longer. So be heard before then, or forever (or at least until next thread) hold your peace.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.