acrossthecourt wrote:But to play down some consistent patterns on a guy who won an MVP makes it seem like some people are just not open to the idea of Garnett as a top player or RAPM in general. It won't matter how accurate the stat is or how pertinent it is to rating players. And it won't matter what new information people will unearth about Garnett.
I haven't been one of the people trashing RAPM or KG. But my issue regarding KG and RAPM as it relates to other players already being discussed or soon to be(Dream, Mailman, Admiral, Chuck to name but a few) is this:
We have this RAPM data for KG that we simply don't have for a lot of the players from a prior generation. And that RAPM data for KG is outstanding and thus I think there is a real danger of starting with that knowledge(that we don't have for other great players) and then finding reasons to argue him above them. Note: Im not accusing anyone of doing that intentionally, but we would be naive to not at least ask the question of what influence this information is having on how we see him in comparison with those we don't have the data on.
No, I worry about that a lot and I'm checking myself for it. Though we've been doing pretty well given Russell/Wilt. I do fear the lack of nifty stats will hurt some players but I trust this group more with the big picture. We can always go back to contemporary accounts and MVPs and work from there: should we trust those results? How good were his teams with/without him? Etc.
But I'm leaning toward Magic anyway. And I'm considering Bird anyway.
I think most guys are doing the same thing. I think guys are doing their best to be fair.
But how do you "forget" that you have seen the RAPM data? So, with KG(and others of that generation) we can look at the data and work backwards to decide what allows him to have those results. Whereas with the former players we don't have that option so their flaws cost them "points" where KG's don't because we can see the end results and they look great so its much easier to rationalize away his weaknesses. This is the reason I asked about Admiral because of how similar so much of their narrative is. And the statistical data we have for both seems to show an edge to Robinson and while KG is a great defender so is David.
ThunderBolt wrote:I’m going to let some of you in on a little secret I learned on realgm. If you don’t like a thread, not only do you not have to comment but you don’t even have to open it and read it. You’re welcome.
magicmerl wrote:If you go and look at the Boston roster the the previous season, and ordered the players by DRtg, the Celtics literally kept every good defender, and traded every bad defender away. Again, this is analysis that you can do before their championship season even starts.
That's not even true. Al Jefferson was second on the team in DRtg (with at least 1000 minuets) and was traded. He was the highest rated big man besides Olowokandi, who played like 300 minutes and retired. Plus DRtg is a pretty awful evaluator of individual defense.
Chuck Texas wrote:I think most guys are doing the same thing. I think guys are doing their best to be fair.
But how do you "forget" that you have seen the RAPM data? So, with KG(and others of that generation) we can look at the data and work backwards to decide what allows him to have those results. Whereas with the former players we don't have that option so their flaws cost them "points" where KG's don't because we can see the end results and they look great so its much easier to rationalize away his weaknesses. This is the reason I asked about Admiral because of how similar so much of their narrative is. And the statistical data we have for both seems to show an edge to Robinson and while KG is a great defender so is David.
Eh I think of it like this: use your available tools to estimate how far a player separated himself from his peers. You can try to think of every player in a +8 or whatever sense. You don't need RAPM for that. I look at Garnett and see his +8 or more seasons. Then I look at Magic, his team, the stats, etc. and also see that he appears to be another +8 player. I go from there and try to estimate career value (like championship odds or wins added.)
Robinson has poor longevity and doesn't appear to play well in the playoffs, Going against guys who were legends at all levels like Bird/Magic he has no chance and against guys with long careers like Olajuwon/Garnett he has that disadvantage. Garnett/Olajuwon/Robinson are all amazing defenders and some of the best ever though. I don't think Robinson is better, and then there's that short prime/peak.
Robinson's decline in efficiency isn't small like Garnett's it's huge, especially during his best seasons. It's not the same.
I'm still open to Robinson. But someone will have to give a great argument. (I'm pretty sure he'd have monstrous RAPM numbers too because he did well from '98 on and those Spurs teams pre-Duncan didn't have a lot of help and won more than you'd think.)
acrossthecourt wrote:Okay, I'm pretty tired of this "RAPM isn't very useful because it's only finding rotation trends."
First of all, what does that even mean? It's adjusted for strength of opposition and who you're playing with. If you play with a defensive unit, those defensive players will have good defensive values and it won't assume you're a good defender either.
The rest of your post is pretty spot on, but this portion is only mostly true. This only holds in situations where you believe there to be no tangible interdependence that exists outside of an individual player's value. For example, NBA teams are deeper on the perimeter. This is and has been a fact for some time. You are structurally altering your team more when you take a starting big off the floor (not all the time obviously, just on average) than any starting wing.
Purch wrote:1. The Boston bench was significantly better than the heat bench, so even though the big three didnt compare by heatlers standard, their depth was significantly better.
2. The Celtics had a significantly better coaching staff between Rivers, and the greatest defensive Arcitect of the past 7 years, Thibs.
3. In 2008 their defense took the league by storm, teams like the Spurs still hadn't developed the kind of offense to counteract the trapping and zone principles in that style of defense.
4. Most Importantly the 08 celtics were NOT nearly as dominant as people make them out to be. Just by comparing them to the spurs recently. I made this post like a week ago
Compared to something like the 2014 Spurs I feel the 08 celtics run looks underwelming.
Even when people talk about their great regular season run, I actually think when you look at the two teams it's obvious that the spurs were better.
The celtics finished 66-16, whiles the spurs finished 62-20
What makes the spurs regular season even more impressive, is the fact that none of their players played an average of 30 minutes or more (first time in nba history) Not only that, but Parker sat out for a full two weeks just because Pop wanted him to rest. This is in the midst of one of the best western conferences of all time, yet the spurs were resting their players and still finished with the best record by 3 games, and had the best road record by 5 games (30 road wins). Not only that but the spurs played a total of 28'different lineups ( the second most behind the knicks). Not only that but at the same time they amounted a 19 game win game win streak(most in franchise history).
When I look at those factors it's hard for me to put Boston's regular season over them.
Then you talk about the Post season... Boston went 16-10 and had two series go to game 7 whiles they had home court. And their finals run as impressive as it was doesn't compare to what the spurs did at all.
The Spurs went 16-7 in the post season. 12 out of 16 of the Spurs wins came by 15 or more. In the finals they won in 5 games against the 2 time defending champs, and set the record for highest point differential even in a finals series.
Then when you consider the 2014 western conference> 2008 eastern conference (not by nearly as much as today)
People talk about the 08 celtics dominance, but it really doesn't compare to the 2014 Spurs team, where Duncan had a very similar role to Garnett on the defensive end, and had the exact same Usage %.
I'm willing to listen, but I don't see how the 08 celtics dominance compares to the 14 spurs when you factor in the minutes played across the spurs roster, and the games their stars would literally sit out.
No, The 28 different starting lineups they played due to injury and rest, the win streak, and one of the best road records in nba history makes it more impressive to me. It wasn't simply due to rest, nearly all their players missed significant time at some point during the season, and yet they were still destroying the league with ease. It didnt matter when Manu was out, Kawhi, splitter, Duncan.. And they did it the whole season ( Especilly on the road). Its hard for me to put Boston over them simply because of 4 more wins.
The Spurs were still a top 6 defensive team in the league ( Even better in the post season) but Offensivly they were galaxies ahead of the celtics. I think that's why you saw them blowing out many games then the celtics did
Not to mention, you guys talk about Boston stumbling out the gate against Atlanta.... But the celtics let the Cavs take them to 7, whiles the spurs dismantled a better Heat team with a better version of Lebron in 5 games, in the most dominant finals performance in nba history
5. The reason that's important is because Duncan played a nearly identical role (as a defensive anchor, with the same usage%) on a 2014 that was significantly more dominant against tougher competion.
6. Just in general when people use the 08 run as proof of KG's Abilty to be the best player on a dominant championship team, it bothers me because even with home court, the celtics barley squeezed by inferior teams. And were pretty awful on the road in the playoffs.
Well first off, your point by point break down is a bit more of a direct comparison debate than I'm trying to start. LeBron is already in, LeBron's peak impact I've already said I think is bigger than Garnett's. So a true Garnett vs LeBron comparison here isn't needed.
That said to your points:
1. Celtic bench better. Well the Heat's depth has been pretty variant as its gone alone, and part of that is raw ability, but part of that also comes from the top. People were lamenting the Heat bench...until they started praising it...and then they bashed it again. It was honestly funny seeing the whiplash Battier's name was getting along the way.
2. Celtics better coaching. Yes and no, and how much does it really matter?
Yes, in the sense I have more respect for what Thibs did. No, in the sense that there's no reason to think Thibs could work the exact same magic everywhere, and Spo actually did a great job building a new scheme around his team after the first year. Really matter? I mean it does some, but while Thibs is great, he was never so great that the Celtics D looked great when Garnett went out. In the end the allocation of credit between coach and player just shouldn't be seen as zero sum.
3. 2008 D took the league by surprise. It did, but it also continued to look great basically as long as Garnett was healthy. I understand why you might think the Spurs have 'solved' that paradigm now, but I don't think that's really how it works. Garnett & co got old, or they'd still be together with a dominant defense now. The Spurs are awesome and would do about as well against that defense as anyone, but it would still be a tougher for them than other defenses.
4. You neglect to recognize that the Celtics actually coasted (and were injured) to that 66-16 record. That year they started 29-3, and the next year they started 28-4. The team looked every bit as dominant as the record suggested and then some.
As for the playoff struggles, they took a while to get started - as did the Spurs. As I said to people this year when they doubted the Spurs: Don't judge a contenders' ceiling based on the first round.
All that said I agree with you that the '14 Spurs are the more impressive team based on what I saw at the end. The '08 Celtics were the most impressive team since the Shaq-Kobe Lakers, and now the '14 Spurs are.
5. "Duncan had a near identical role in '14." Hmm, I'll go with you there. Duncan played the Celtic Garnett role on a team that didn't win because of anything like the reasons why the Celtics won. That's the difference.
If you look at any of the +/- data from the year, you'll see a stark contrast. Here's the pure raw stuff:
mtron929 wrote:What needs to be accounted for in the statistical argument is the fact that the NBA is a long season with virtually half of the teams making the playoffs. Accordingly, top players realize that they can slack off in the regular season (e.g. Shaq in the 3 peat times), save energy, and exert more effort in the playoffs. Now, if the goal of the NBA was to win as many regular season games as possible with the largest scoring margins, then I would have to think that guys like Shaq and Lebron during the Heat era would have played a lot differently.
That is, they would have played like how KG plays.
KG is getting overrated in the statistical arguments because he always plays really hard. But that is not necessarily prudent. And I don't necessarily think that this is a positive attribute in the context of winning a championship. But it sure did wonders for his +/- stats.
Oddly enough, this is one of the arguments that used to be made in the Hakeem v. DRob threads . . . Hakeem wasn't that motivated in the regular season where Robinson was always playing at full speed . . . then the playoffs come and Hakeem has an extra gear that Robinson doesn't.
However, then you look at how even the greats have a huge difference in winning playoff serieses at home v. one the road (see . . . JordansBulls HCA thing is of some use, at least when combined with non-HCA) and you wonder if slacking in the regular season is good for team success or only for individual statistical advantage.
Oh, and comparing Duncan with Garnett through 2010 and 2011 and calling those Duncan's prime when they were clearly the worst years of his long career seems a bit disingenuous.
“Most people use statistics like a drunk man uses a lamppost; more for support than illumination,” Andrew Lang.
#1 all time in winning % GOAT peak with 3 straight MVPs Essentialy top 3 player from the min he begins playing for the 1st 8yrs of his career. Once he is injured and loses mobility he then posts his greatest scoring season.
Maybe the most clutch player to ever play.
HomoSapien wrote:Warspite, the greatest poster in the history of realgm.
So for everyone calling Boston a non-dominant team, Hollinger rated them as the tenth best finals team ever (updated to the 2011 season) based on their season. They won 66 games, outscored teams by 10 points, outscored teams by 5.2 points in the playoffs (very strong: their losses were by small margins usually and they blew out a lot of teams including the clinching title game), and their defense was arguably the best ever since Bill Russell.
Chuck Texas wrote:I think most guys are doing the same thing. I think guys are doing their best to be fair.
But how do you "forget" that you have seen the RAPM data? So, with KG(and others of that generation) we can look at the data and work backwards to decide what allows him to have those results. Whereas with the former players we don't have that option so their flaws cost them "points" where KG's don't because we can see the end results and they look great so its much easier to rationalize away his weaknesses. This is the reason I asked about Admiral because of how similar so much of their narrative is. And the statistical data we have for both seems to show an edge to Robinson and while KG is a great defender so is David.
This is an interesting line of thought. But I'm not sure, when I follow it to its endpoint, that I come to the same conclusion as you. To me, this is how it plays out:
*Without +/- data, we are lacking some player evaluaiton tools that could help us
*We don't have +/- data prior to 1998, so for any player whose prime is before that, we have a higher degree of uncertainty as to how to quantify their contributions
*We DO have +/- data for 1998 and after. This helps us to better quantify the contributions for these players
To me, I don't think that the +/- data universally helps the current generation. It helps with KG and Dirk because, in part, their styles are so historically unique that without the +/- data they likely would have been UNDER rated. I think that's a good thing.
On the flip side, it's pretty likely that the +/- data does not help players like Kobe. His stylistic play, accolades and team achievements can (and often do) paint him as the closest thing we've seen to Jordan. But the +/- stats give quite a bit of evidence that Kobe's overall and defensive impacts aren't as large as his reputation. That he's awesome, but not 2nd-of-all-time-behind-Jordan type awesome.
When it comes to comparisons with players from previous generations, I haven't really seen any evidence that the lack of +/- data has harmed those players. For one thing, with the WOWY data and the historical team trends there's a big push to get as close to a +/- type profile as possible with older players. Then, when that's not possible, I know I have tried to estimate what several player's +/- footprint might look like. I did it with Russell (when he got my vote) and then again with Magic (to their favor), I also did it with Kareem and Wilt (not to their favor). But again, I don't think that the existence of present day +/- data was in any way adverse to the ranking of the older players.
Now, with Robinson in particular, it's interesting. But NOT because the presence of present day +/- hurts him. Instead, it's more that if we would have had access to +/- data of that time it may have helped him because I agree that his +/- scores would likely have been outstanding. But for me, I feel like I account for that. The same way I did with Magic and Russell, I expect that he would have looked great with +/- so I don't know that the actual numbers would change my opinion that much. In fact, I'd say that Robinson is pretty universally regarded as a Megatron impact player in the regular season. And that's the kicker. Like KG, Robinson is primarily criticized for perceived weaknesses in the postseason. And as we're seeing with KG, no amount of RAPM scores is going to change people's minds when they have questions about your postseason ability.
So again, at the end of the day, the conclusion that I reach is that having more +/- data in recent years is outstanding for helping to evaluate players in recent years. But I don't think the presence of that data universally helps the more modern players, nor does it necessarily hurt older players. And the negatives to lacking +/- data in previous years wouldn't be helped at all if we didn't have current +/- data...it would just mean that we were working more in the dark for everyone...which I don't think would be better.
acrossthecourt wrote:Okay, I'm pretty tired of this "RAPM isn't very useful because it's only finding rotation trends."
First of all, what does that even mean? It's adjusted for strength of opposition and who you're playing with. If you play with a defensive unit, those defensive players will have good defensive values and it won't assume you're a good defender either.
It means exactly what is says. The datasets being used are based on rotations. Seems pretty straightforward.
And if you play with a defensive unit, then one would expect better defensive trends. Which is what happens.
Spoiler:
If you're talking about collinearity, then yes, that's a problem, but that's why we use priors and RAPM.
"6th men like Manu/Odom will rate high. "
Manu started every game in 2005 and was first in prior-informed and NPI RAPM in 2005. He wasn't a sixth man then. He was a part of their crunchtime lineups (he has been for a long time) and was often the primary ballhandler in close games.
Manu played 29.6 mpg, so while he did start, his role was still more of a 6th man. I think you're overstepping the point which is that Manu had utility to SA's rotations, and that his numbers are more reflective of how Pop used him in SA's rotations, and nothing else.
Priors don't do anything but build rotational trends with larger sample sizes. Defensive players will tend to be in more defensive rotations no matter what team there on. All you find out is how effective those rotations were.
Spoiler:
Yet advanced plus/minus metrics perform very well at predicting future wins, particularly when there's a lot of roster turnover.
You see what that means? That even with a new context the player values attached to RAPM are useful? It's not rotation trends. It was rotation trends, then when we tested the metric out of sample with future season sit wouldn't perform well. But it does well.
Using lineup data in a predictive way is to be expected. There's nothing magical about that. The very nature of coaching is to build consistent lineup rotations throughout the season.
Since RAPM does quantify "rotational trends", it able to predict the effectiveness of a team vs another fairly accurately. It's a team profile essentially, and not far off from predictive methods used in political campaigns.
For the life of me, I don't' get why some think the predictive nature of RAPM is special, it isn't.
Spoiler:
It's not a relic of rotational weirdness. But you know what? Other metrics have problems like that. Jason Collins sets a screen, gets Vince Carter open, doesn't get credit for it. Jason Collins plays Dwight Howard well, holding him to 16 points on middling efficiency without much help. He does this without blocked shots, so he doesn't get credit for it in the box score. PER ignores him.
Again, I don't find RAPM weird, I can't clearly see what it's doing. That "weirdness" is supposed to be there. And i agree about PER, I've pointed out its shortcomings for years around here.
Spoiler:
JJ Hickson crashes the boards, picks up rebounds, sometimes steals them from teammates, hits a lot of shots inside, plays terrible help defense but still picks up steals, and looks great on PER. Yet when he's replaced with Robin Lopez Portland jumps up 20 wins (it wasn't just Lillard's improvement.)
Carmelo outscored Durant in 2013. Who was the better scorer?
Alex English led the 80's in points. Who was the best scorer in the 80's? And best player?
Adrian Dantley looks comparable by PER to Bird from like '80 to '86. Who was the better player?
Chris Andersen led the 2013 playoffs in WS/48 minutes. Was he the best player?
Zach Randolph had more Defensive Win Shares than Noah in 2014. Who was the better defender?
Boozer was 13th and Rudy Gay was 17th in Defensive Win Shares. Were they great defenders?
All stats have problems. But we shouldn't ignore every stat because they're not perfect.
The issue isn't that people should ignore stats, but instead realize what they're actually saying. RAPM is fine for lineup analytics, but in no way are valid for comparisons. The data its built from is based on groups, not individuals, and no model can extract that.
You can nitpick all you want, find the weirdest results and make fun of them, but that is no way to rate a metric or system. You do that systematically with EVERY result (if you can.) No one's saying RAPM is perfect. There are flaws. There's noise.
But to play down some consistent patterns on a guy who won an MVP makes it seem like some people are just not open to the idea of Garnett as a top player or RAPM in general. It won't matter how accurate the stat is or how pertinent it is to rating players. And it won't matter what new information people will unearth about Garnett.
Again...that word 'weirdness". Strange that backers of RAPM think these ratings are 'weird", when they should be expected. Go look at RAPM rankings my friend...then look at the rotations the player was involved in. Everything there is too be expected.
As for KG, well yes, if people are ranking KG Top #4 ever because of the rotations he was used in, I have a problem. KG's impact never correlated to some GOAT level in his playing days. He's a versatile defender, who struggled to be a 1st option, but his impact was countered quite often by other greats in the playoffs.
7-time RealGM MVPoster 2009-2016 Inducted into RealGM HOF 1st ballot in 2017
I figure since we're bringing up Karl Malone, I'd go ahead and bring up his opposite- Bill Walton.
Yes, I know he was injured a ton and that we only got to see a little of him, but really, take a look at what the guy did in his limited time on the court. MVP, FMVP, two time champ, 6th Man of the Year. All-NBA first team over Kareem in 78, 2nd team in 77.
Let's break it down
Rebounding I figured it was good to start with the parts of his game that aren't so talked about. Bill Walton was an excellent rebounder. In his peak year, he lead the league with 14.4 rebounds per game. He had six seasons with a TRB% over 20%, 5 seasons with a DRB% over 30%, lead the league in TRB% twice (plus a second, third and fourth place finish) and DRB% four times (plus was second once). Walton was all over the glass, particularly the defensive glass, and was still a force on the boards when he was playing under 20 minutes a game for the Celtics, collecting 6.8 rebounds a game in 86, or 9.5 per 36. He was a high effort guy who really cleaned up the glass spectacularly well.
As an example, Walton played Moses Malone 6 times during his short prime (77 and 78). Moses did most of his scoring off the glass, and since I don't have the full box score for their stats, I'm going to have to approximate the battle on the boards between them by means of their points. Moses score 3, 4, 14, 15, 23 and 24 points in his games against Prime Walton; to me that indicates that Walton was keeping him off of the glass most games.
Defense Another part of Walton's game that doesn't get as much praise as it deserves (all anyone wants to talk about was his passing), Walton was a huge defensive presence. While he wasn't the freakishly tall or long dude like Wilt or Kareem, nor was he a physical specimen like Robinson, Walton was a very mobile center with an extremely high IQ. The easiest way to illustrate his defense is with his shot-blocking. Walton was a stud shot-blocker, and this was due to his great motor and activity. He worked to contest every shot like Bill Russel did (someone in a previous thread had illustrated this point, but I can't find it yet, there's a lot of pages to go through). In 1977, he lead the league in blocks per game with 3.2. If you'll recall, he also lead the league in rebounds per game this season as well. This is really incredible. Prolific shot-blockers are regularly taking themselves out of position to get rebounds by contesting shots, but Walton managed a double-feat matched only by guys like Bill Russel (we assume, since they didn't record blocks) and Ben Wallace. This is a testament to his defensive instincts, intelligence, mobility and effort.
Walton had 6 seasons with over 2 blocks per game, include one year with the Clippers when he got 3.6 blocks per game. Despite playing only 65 games in '77 he was 3rd in total blocks, and in 78 he was 10th despite playing only 58 games. In addition to his lead-league in '77, Walton also finished 2nd in BPG in 83, 4th in 78, and 8th in 85. He is 17th in career BPG. He had 4 seasons in the top 10 in block% as well.
Walton also looks great by the advanced stats. He had 5 seasons with a top 10 DRTG, including 2 first place finishes (89.5 in the 78 season, and 97.5 in the 86 season), and 2 seasons in the top 10 of DWS.
He also was a 1st All-defensive team member twice, in 77 and 78.
Scoring Walton doesn't always get a ton of credit for his scoring, but it's sort of unfair because he played on teams that spread the wealth and was very unselfish himself.
Walton had plenty of offensive game himself. He could work guys in the post, had a great bank shot and of course could finish well around the rim. In his two prime seasons with Portland, he scored on 18.8 PPG 55.9TS%. He was not a stud at drawing fouls until his bench roles with Boston, when he broke 60TS% in 86. Sure, he isn't Shaq or Dirk, but he had a solid scoring game that translated to the playoffs, like Tim Duncan's level.
Passing Ah, what everyone really loved about Walton was his excellence as a passer. He never spent much time playing as the sole offensive focus on his team, sharing offensive responsibilities with everyone on Portland and Boston, but he really made an impression. He is often mentioned as the best-passing center of all time, one of a handful of centers to ever average 5 assist per game (including Wilt, Russel, Kareem, Unsled, Noah, Divac, Sam Lacey and Alvin Adams).
However, I admit that this is a pretty small case here. I'd greatly appreciate it if guys who actually have the data could show how he did head-to-head with guys like Moses and Kareem (especially in the playoffs), and data that indicates how the team performed with and without Walton (should be easy since he missed so many games), particularly in his two all-star seasons with the Blazers. SRS, record, team ORTg and DRTg, that sort of thing.
Notanoob wrote:This is unfair. By RAPM, Dirk has almost an identical overall impact on the game, and I recall RAPM being cited in Dirks favor before.
I'd also like to ask, what separates him from David Robinson?
At a somewhat superficial level, they're pretty similar.
Both were defensive savants who were forced to become their team's primary offensive anchor due to a total lack of talent.
Both saw their offense (scoring) fall off in the post-season and lack post games.
Both finally got support at the tail-end of their primes and were part of dominant defenses.
I'm not entirely clear on it, but the RAPM data for pre-2000 guys is estimated due to a lack of play-by-play data, correct? http://stats-for-the-nba.appspot.com/ has some RAPM data for Robinson has him as a +9.3 guy and KG as a +10.3 guy.
I bring it up because I'm not sure if I'm supposed to trust the data for the 90's stuff. I believe that Robinson could have a better impact on defense because he's about as mobile as KG and even longer, but stronger and better able to fight with centers down low. Then on offense, while neither were ideal #1 options, Robinson had more success as a volume scorer, winning a scoring title even, doing a lot better than KG ever did. Is KG's edge as a passer large enough to give him a better overall impact?
So this confusion just keeps coming up.
I'll state this acknowledging that there is difference in opinion about the stat itself, but speaking as someone who feels comfortable calling himself an expert in the use of +/- stats:
By both pure APM and by prior-informed RAPM, Garnett is very clearly ahead of Nowitzki throughout the bulk of his career, and therefore any reference saying that "RAPM" says otherwise is wrong.
The non-prior-informed RAPM rounds down Garnett considerably which is why in it Duncan appears better, and Dirk appears his equal, but there's a very clear flaw to using that stat and that flaw would seem to directly explain the disagreement. Simply put, non-prior-informed RAPM will directly chip away at the weight of outlier data under the assumption that it comes due to luck. You can tell it's happening if the prior-informed RAPM keeps seeming to undo this assumption year after year. To put it in the form of a story:
Pure APM: Wow, Garnett's amazing. NPI RAPM: Too good to be true, bet you he can't do it again. PI RAPM: He did it last year too.
Repeat year and after year.
Okay, but could you address the parts about Robinson and the 90's RAPM data? I want to know not only for this debate, but for general use, like comparing Marbury to Kyrie Irving (pretty similar RAPM numbers from that site through their first three seasons).
Ah, sorry about that.
Short of new information, I would say be very careful using anything from that site. I hate to say that because for a good while that was a site I used a lot. After Joe Sill got his NBA gig, Engelmann was the first to provide new RAPM data and I loved it. But then he started messing with things without explaining what he was doing on his site:
1. I would assume all data before '01-02 on there is not any kind of RAPM but just his variant of statistical +/-, because at least originally that's all it was despite him not labeling it to the contrary.
2. I would assume that more recent data is whatever he considers at the moment to be the best single metric he has, unless it says otherwise specifically. To my knowledge he still tends to call this 'xRAPM'.
My issues with this stat originally was that it used a box score-based prior, which given that I already use the box score in my analysis meant that he was adding in a kind of double-count for me. This wouldn't be so bad if it literally just counted things double because then I would know how the wind blew and just change my swing, so to speak, but instead with its use as a prior it becomes very difficult for me to know just how much extra bias is being thrown in there.
To be clear, that doesn't make it a bad metric. It's a good metric if it's the only metric you plan on using. You just shouldn't ever use only one metric, so philosophically his goal to make a Holy Grail stat runs counter to the actual purpose of using stats, and he doesn't understand that.
There are additional issues that I stopped trying to keep track of. In his Holy Grail quest he's started factoring age and I believe even height into his priors. All of it makes sense from his perspective, but I would imagine I wouldn't have to convince people here that factoring in non-basketball stuff into a metric meant to estimate what impact a player actually had causes problems. I can imagine a day where ESPN gets into hot water finding out the RealPlusMinus metric they use from was using race and religion to assess players and Engelmann scoffs in annoyed condescension at their inability to understand the math as he gets blacklisted from the world.
For the record, the spreadsheet that I put on Google docs has the following data:
Prior-informed RAPM from Engelmann (from before his XRAPM phase) going back to 2002 from 2012 which is when he switched. Across the court's prior-informed RAPM from 2000 back to 1998.
To my knowledge that's basically the data we have going back into the '90s.
Though 2 things: 1) I know there are people doing RAPM from the last couple seasons. I've been busy and a bit reluctant to add them to my spreadsheet, but at a certain point I'll have to really look into them and see if I can find a basis for which to include.
2) shutupandjam with the stuff he's been post makes me think that he might be doing a lot of things I need to understand better.
Getting back to Robinson: acrossthecourt's data makes him look quite good given that we haven't actually seen his real prime. The data we have taken on face value clearly puts him as weaker than Garnett, but it's very possible that when we see Robinson's actual prime he'll surpass Garnett.
ronnymac2 wrote:I'm going to try and post about Hakeem Olajuwon, specifically his seasons from 1986-1990. Everybody pretty much agrees his peak is at an extremely high level, but he doesn't seem to be getting much mention, and I figure perhaps it's because people are tentative about how to rank his non-peak prime seasons.
I'm going to start with 1990 because I feel it is the most important one to get to and I might not get to all of them.
1990 Hakeem Olajuwon is quite possibly the greatest defensive season in NBA history by a player other than Bill Russell. Dream played 82 games for 38.1 minutes per game. He averaged 10.4 defensive rebounds, 2.1 steals, and a league-leading 4.6 blocks per game. He led the league in total defensive rebounds and total blocks.
He led the league in defensive rebound rate (28.3%) and block percentage (7%). For those who like these stats, he led the league in defensive wins shares (8.7) and individual defensive rating (93).
Houston was a —4.7 defense. This ranked as the best in the NBA in 1990. For Defense Four Factors, Houston was 3rd in eFG%, 9th in Turnover%, 6th in defensive rebound%, and 7th in FG/FTA. All marks were above the league average.
Olajuwon was All-Defense first team, but somehow, he finished second in DPOY voting to Dennis Rodman. This is perhaps the most egregious award error in NBA history. Olajuwon played 3,124 minutes that season (again, 82 games for 38.1 minutes per game). Dennis Rodman played 2,377 minutes (82 games for 29 minutes per game). Rodman's Pistons were an inferior defense. None of his defensive stats are really all that close to Olajuwon's. He averaged 3.4 fouls per game. Olajuwon averaged 3.8 fouls in significantly more minutes (Olajuwon was actually 9th in minutes per game that season in the NBA).
Olajuwon's Rockets faced the best offensive team in the NBA that season, Magic Johnson's LA Lakers in the first round, and lost 3-1, with Magic's Lakers performing well offensively. Olajuwon averaged 5.8 blocks, 2.5 steals, and maintained roughly the same defensive rebound rate as during the season. LA was a legitimate title contender that year.
Olajuwon in 1990 also gave you 24 points on 50 percent shooting and over 70 percent free throw shooting as a big man. He averaged 2.9 assists and 3.9 turnovers.
My Opinion: If you meld him onto a team with more offensive talent, I do not doubt he could average 21-22 points, 3 assists, 3 turnovers and replicate his defensive dominance.
I'll keep traveling backwards in time: 1989
Defensively, Olajuwon not winning DPOY isn't as egregious as it was in 1990. Mark Eaton was playing about 35 minutes per game and anchoring the best defense in the league, so it's at least an understandable selection. But Olajuwon was most certainly playing at a somewhat similar level that he did in 1990.
Hakeem played 82 games and played 36.9 minutes per game. He averaged 9.4 defensive rebounds, 3.4 blocks, and 2.6 steals for Houston's 4th-ranked defense (Houston was a —3 defense). He led the league in total defensive rebounds and was 3rd in both total blocks and total steals. He led the league in defensive rebound rate (27.2), defensive win shares (7.8), and individual defensive rating (94.9).
Defensive Four Factors: Houston ranked 8th in eFG%, 10th in turnover%, 5th in defensive rebound percentage, and 12th in ft/FGA. All of these marks were above league average that year, which makes sense since Olajuwon is affecting offenses in every way imaginable (he ends possessions with defensive rebounds, causes a crazy amount of turnovers, and lowers team's efficiencies by blowing up plays on the perimeter and contesting shots in the paint).
Offensively, Olajuwon is scoring 24.8 points with 1.8 assists, 4.1 offensive rebounds, and 3.4 turnovers per game. He shot 50.8 percent and 69.6 percent on FTs. 55.2 True Shooting percentage.
(Just a slight tangent...I hate when non-volume scoring, non-creative players average a ton of turnovers, which is exactly what Otis Thorpe does. The same thing pisses me off about early Charles Oakley. Thorpe is giving you 17 and 2.5 with 2.7 turnovers per game...smh).
Playoffs: Rockets face a decent Seattle team. Seattle clearly isn't a contender, but they do have a 5th-ranked offense, which slightly underperforms against Houston compared to the REG SEA (111.5 in REG SEA, 108.9 in 4-game sample against HOU). Olajuwon does 25/13 with 3 assists and 2.5 turnovers per game. 2.8 blocks, 2.5 steals. True Shooting percentage of 54.9 percent. If anything, he did a little better with his assist/turnovers ratio here.
Houston actually outscored Seattle over the course of the series. Seattle won Game One by 4, Game Two by 12, and Game Four by 2. Houston won Game Three by 19. After Olajuwon, Thorpe, and Sleepy Floyd, the rest of the Rocket offense just wasn't very good.
To me, this season provides more GOAT-level defense (not including Russell) with very strong offense that I think would look even better in terms of impact (but slightly worse in terms of counting stats) if paired with better teammates.
For those who worry about Hakeem not drawing fouls because he's all stepbacks and fadeaways, Dream got to the line a career-high 8 times per game this year. Went over 6 times per game in the 4-game sample we have in the playoffs. We aren't talking about an Al Jefferson case where he gets off clean shots at the expense of not getting to the free throw line for even more efficient, clean shots.
Pay no mind to the battles you've won It'll take a lot more than rage and muscle Open your heart and hands, my son Or you'll never make it over the river
Chuck Texas wrote:Can you really defend the idea that KG is better than Dirk at every other aspect of offense besides scoring? I'm not at all sure I buy that idea. He's a much better offensive rebounder. He's a better passer. I know people rave about his screen setting so I'll give him that. But does he really create more good shots for teammates than Dirk does? I just don't see that at all. Does he distort defenses the way Dirk does? I struggle to see that as well. Dirk rarely turns the ball over. Dirk is better at transitioning from offense to defense(tho admittedly part of this is due to system, but Dirk is tremendous at getting back on D)
I think we have to be careful to look at their assist totals and think that KG is a better offensive creator than Dirk. Dirk loses a lot of raw assists for a number of reasons--he's always been the best finisher on his teams is a simple one, but also the Mavs system has always been about ball movement. All this stuff guys are losing their minds over with SAS offensively, well the Mavs do it too, just not typically as well as the Spurs just did it. Tho we saw them peak in the 2011 playoffs in this regard. And if you look at the 5 man units of the Mavs in the Dirk era you can really see that guys like JET and JJ Barea to name a couple of title year Mavs are light years better in units with Dirk than when they played without him. And these were supposed to be the other Mavs on that title team not dependent on Dirk.
I get KG is a more versatile player than Dirk, but let's not be so quick to try and put Dirk in the "he's just a great scorer" box. Yes, he's a great scorer, and yes most of his value is derived both from that, and the distortion and opportunity his scoring creates, but Dirk brings more to the table than that.
Note: I appreciate the Admiral response and I will comment some on that probably later tonight or tomorrow as time avails.
I think that by going into more depth here you're double counting.
If Garnett is a better passer then the only way for Dirk to create better opportunities for his teammates is simply based on the gravity he tugs on the defense with...which is another way of saying his scoring ability. We're still just talking about one thing then. One big thing sure, but one thing.
Re: turnovers. Turning the ball is something you do while you try to score and try to pass. Yes you could divide it and talk about a specific ability to literally hold on to the ball rather than being a butterfingers, but for the most part you turn the ball over while trying to do something, and whatever that thing is, it has a name.
If you want to argue that Dirk's playmaking ability is better than Garnett's based on this particular subtlety okay, but truly trying to split every thing into independent details when they aren't actually independent doesn't seem helpful to me when evaluating a player's overall contribution.
I understand though if what you're really taking issue with is me saying something along the lines that Dirk is a one-trick pony. I don't like calling him that either. It's not how I see him, but if I were to some of the situation like this:
Dirk is the better overall scorer Garnett is the better offensive player outside of that Garnett is the better defensive player
Can you see why I find it frustrating when someone goes into great nuance analyzing the first part, and then waves their hands at the rest and says "but really, by how much?" essentially making the assumption that it can't be that big. It's the bulk of what is being done by the 10 players on the court at any given time, but how important is that really? Can't we get them to sell concessions instead?
In regards to having RAPM for modern players and not for older ones: I think it actually helps some older players. We see the trends with modern RAPM, like how dynamic offensive creators who can pass do really well, so we look to Bird and assume he's one of them. Or Magic. And highly active defensive big men who are unselfish and play well with their teammates on the offensive end look great. So we look at Russell and Walton and see the same thing.
An Unbiased Fan wrote:Manu played 29.6 mpg, so while he did start, his role was still more of a 6th man. I think you're overstepping the point which is that Manu had utility to SA's rotations, and that his numbers are more reflective of how Pop used him in SA's rotations, and nothing else.
Priors don't do anything but build rotational trends with larger sample sizes. Defensive players will tend to be in more defensive rotations no matter what team there on. All you find out is how effective those rotations were.
How exactly can Pop use him in lineups that makes the team much, much better that has nothing to do with Manu given that we also account for the competition and his teammates? People say similar things like this all the time, but I have no idea what they mean. What are these rotation relics that cause Manu to be rated him year after year, and why is it basically only Manu and I guess Odom for like three years (hm two players known for passing ... nah that's a coincidence)?
San Antonio has changed its lineup many times as well as their style. Manu has gone from not starting to starting again and then coming off the bench. It's not how he's "used" in the rotation.
I said he was first in non-prior informed too. It wasn't just the prior.
Defensive players are not more likely to be in defensive lineups. That's not even true. Coaches often go with more balanced lineups and someone like Asik has to play with Harden and a small PF. Basketball isn't so specialized teams go with all defensive lineups for five minutes and then sub them all out and go with all offensive players. It's usually quite mixed.
And again it's adjusted for who your teammates are. If your lineup has great defenders in it, you're not simply getting credit because they defend well. The lineup has to perform better than expected, and we then give that player the credit ... which makes sense!
I honestly have no idea what this rotations stuff is. This was the old criticism with raw plus/minus, and it was valid. Ostertag has ridiculous raw plus/minus some seasons because he plays a lot with Malone and Stockton. So what you do is adjust for your teammates, opposition, and homecourt advantage to get rid of the pesky rotation problems.
Is it something about how Manu only plays with lineups that gel and none of the troubling ones? That he's saved for the best lineups? Even then, Manu played most of the minutes in his prime, was there in crunch-time, so I don't see how he can be saved for super specific rotations and possessions where he can trick RAPM.
Doctor MJ wrote:I understand though if what you're really taking issue with is me saying something along the lines that Dirk is a one-trick pony. I don't like calling him that either. It's not how I see him, but if I were to some of the situation like this:
Dirk is the better overall scorer Garnett is the better offensive player outside of that Garnett is the better defensive player
Can you see why I find it frustrating when someone goes into great nuance analyzing the first part, and then waves their hands at the rest and says "but really, by how much?" essentially making the assumption that it can't be that big. It's the bulk of what is being done by the 10 players on the court at any given time, but how important is that really? Can't we get them to sell concessions instead?
How is it not fair to parcel out the difference between being a better scorer and distorting defense when, materially, scoring here seems to be a measure of Volume x Efficiency. It's also worth differentiating because players who measure out similarly on the volume x efficiency paradigm do not actually cause comparable amounts of "defensive havoc," for lack of a better term, in many many instances.
What is everything else? Passing. dribbling, and setting picks? That deserves it's own classification of "everything else" absolutely no more than Dirk's defense-distortion does.
My personal opinion has always been KG is a historically underrated offensive player, and he actually does offer a lot of the same "distortive" qualities (just not to the same degree), but I can't agree with the way you presented that. Dirk is/was a MUCH better offensive player than KG. The defensive gap is just even bigger.
mtron929 wrote:What needs to be accounted for in the statistical argument is the fact that the NBA is a long season with virtually half of the teams making the playoffs. Accordingly, top players realize that they can slack off in the regular season (e.g. Shaq in the 3 peat times), save energy, and exert more effort in the playoffs. Now, if the goal of the NBA was to win as many regular season games as possible with the largest scoring margins, then I would have to think that guys like Shaq and Lebron during the Heat era would have played a lot differently.
That is, they would have played like how KG plays.
KG is getting overrated in the statistical arguments because he always plays really hard. But that is not necessarily prudent. And I don't necessarily think that this is a positive attribute in the context of winning a championship. But it sure did wonders for his +/- stats.
Two issues with this:
1) If this were true and agreed to be true then Garnett would have all sorts of regular season accolades based on this success, and he doesn't. If you'd like to stand up and say that most people are clueless as to how good Garnett is in the regular season before coming back to the playoffs and saying "but it turns out they are right in the end" I suppose you can, but it's quite the coincidence you end up agreeing with the status quo despite this zig zag and that you don't seem to think that others who miss the zig zag are doing anything wrong.
2) There really isn't a basis for saying that Garnett is a big disappointment in the playoffs. In Minny his team only got "upset" twice, and both time it was to a Laker team that everyone knew had played below its capabilities in the regular season, and in Boston there was no such track record for getting upset at all.
Now I recognize that that kind of slacking off in the regular season the Lakers did is exactly the thing you're talking about, but make sure you recognize that the Lakers are a completely different animal than everyone else. It's not that everyone else slacks off in the regular season except Garnett. Most teams don't. Most teams recognize that they are being judged quite a bit on the regular season. It's really only extreme cases like a repeating champion led by older and/or lazy talents where this becomes a big issue.
That said, of course I'm talking about Shaq here, and that can be used to make the following argument:
If Shaq's +/- data is that impressive given how lazy his ass was, imagine what it would be if he played 100% all the time.
BTW, I very much like the Karl Malone and Julius Erving mentions. Malone's prime lasted forever, and he gives you a lot of chances at winning a title. Erving's peak is insane, but his time as a super portable prime player might be even more impressive. As someone who is finally seeing how important longevity can be when it comes to total career value, these 2 players are entering my mind.
Pay no mind to the battles you've won It'll take a lot more than rage and muscle Open your heart and hands, my son Or you'll never make it over the river
Doctor MJ wrote:That said, of course I'm talking about Shaq here, and that can be used to make the following argument:
If Shaq's +/- data is that impressive given how lazy his ass was, imagine what it would be if he played 100% all the time.
It's pretty crazy to consider actually.
You've just turned into the biggest Shaq fan, haven't you.
Pay no mind to the battles you've won It'll take a lot more than rage and muscle Open your heart and hands, my son Or you'll never make it over the river
