2001-2014 RAPM (Updated now to 1997-2014 RAPM)

[Doctor MJ]

Why multiyear approach is superior, Doc? Because of bigger dataset? Didn't mysticbb said it doeasn't matter in RAPM if 25% (or so) season is played?

Well to be clear, I didn't say it was superior, I said it was superior if you could make a particular assumption. That assumption isn't really at all true over something like 10 years.

Why in theory is it better? It's just more sample size factored in together. Using separate years you're essentially adding in the noise from each year all together, which isn't good.

To try to read mystic's mind ahead of time: He's not a believer in multi-year data sets like this for the reasons I've given above - but they bother him more than they bother me. So while I'm essentially saying I like the separate approach here bet than the compound approach with some caveats, I'd imagine he'd state the same opinion in a more resounding fashion while adding in some additional insights along the way.

[lorak]

Why in theory is it better? It's just more sample size factored in together. Using separate years you're essentially adding in the noise from each year all together, which isn't good.

Are you sure that's the case with RAPM? I mean, it's the problem with APM, where standard errors are big, but RAPM doesn't have the same issue, right?

[Doctor MJ]

Having too much fun with this so I'll have to stop, but breaking it down by draft year (and grouping undrafted with the year they were eligible), here are the top player from drafts 2007 back through 1996, which I think are the years represented here that it gives some pretty confident information for, listed with the top player first (if less than 5 shown, it's because less than 5 were in the Top 100):

2007: Durant, Conley, Young, Gasol (Less than 5. Misses: Noah & Horford)
2006: Aldridge, Millsap, Rondo, Lowry (Less than 5. Miss: Roy)
2005: Paul, Deron, Amir Johnson, Felton, Bogut
2004: Howard, Deng, Iguodala, Smith, Varejao
2003: LeBron, Wade, Bosh, Howard, Collison (Melo is 6th)
2002: Nene, Prince, Yao, Dunleavy, Amare
2001: Parker, Gasol, Battier, Johnson, Chandler
2000: Turkoglu, Martin (Seriously, only 2 made the cut)
1999: Ginobili, Artest, Baron, Marion, Odom (Brand is 8th!)
1998: Nowitzki, Pierce, Carter, Lewis, Brad Miller
1997: Duncan, Billups, McGrady, Anthony Parker (Less than 5)
1996: Kobe, Nash, Allen, Wallace, Iverson

Number of players represented for each of these drafts in Top 100:
2007: 4
2006: 4
2005: 6
2004: 9
2003: 8
2002: 5
2001: 11
2000: 2
1999: 13
1998: 6
1997: 4
1996: 10

[Doctor MJ]

Why in theory is it better? It's just more sample size factored in together. Using separate years you're essentially adding in the noise from each year all together, which isn't good.

Are you sure that's the case with RAPM? I mean, it's the problem with APM, where standard errors are big, but RAPM doesn't have the same issue, right?

RAPM doesn't eliminate the issue, it just reduces it. One can argue it reduces it enough that it's not a concern, but I don't believe that. One can argue it reduces it enough that it's better than any realistic alternatives, and I would agree there's truth to that.

[SideshowBob]

mysticbb,
does it make sense what he did, considering we shouldn't compare directly RAPM from different years (I mean, for example +4 in 2001 isn't the same as +4 in 2012)?

colts18,
could you explain how you did that? Use Durant as example and show step by step how you get his total value, please.

It doesn't make sense to me. The whole sample is including multiple studies. Half of it is NPI RAPM, other parts of it are RAPM informed RAPM and then the 2013 RAPM is prior informed but not done by J.E. but v-zero instead. 2014 is using talkingpractice's NPI RAPM. Looking at all of that as a whole makes no sense to me at all. Different parameters were used for each of the different models. They are then, by definition, different models that use the same regression method and have just been given the same label by us (and whoever ran the regression).

[O_6]

I love the idea of this. In theory this could be basketball's answer to baseball's Wins Above Replacement. Basketball will never be able to be measured by metrics to the degree baseball can because of nature of the game, but we can definitely learn things through RAPM and measurements like that.

Seems like there are questions about the execution of this, I don't know enough about RAPM to know how reliable this data is. But I love the idea of RAPM Over Average as a career measurement metric, and I thought the results passed the eye test and at the very least is pretty accurate even if you should take it with a grain of salt.

It's easier to measure and understand offensive value than defensive value, so I was interested in seeing who the top defenders were according to this data.

Defensive RAPM Above Average: Since 2001
1. Tim Duncan ---- 2502.4
2. Kevin Garnett ---- 2198.6

3. Ben Wallace ---- 1437.3
4. Metta World Peace ---- 1277.9
5. Andre Iguodala ---- 1202.6
6. Shane Battier ---- 1087.3
7. Jermaine O'Neal ---- 1042.8
8. Luol Deng ---- 1026.3
9. Jason Collins ---- 1024.5
10. Dwight Howard ---- 1001.4
11. Rasheed Wallace ---- 996.8
12. Nene ---- 995.8
13. Bruce Bowen ---- 948.4
14. Lamar Odom ---- 941.8

15. Josh Smith ---- 887.2
16. Gerald Wallace ---- 880.9
17. Jason Kidd ---- 879.9
18. Dirk Nowitzki ---- 837
19. Dikembe Mutombo ---- 828.8 (despite missing most of his prime, probably at 2500+ career value)
20. Brendan Haywood ---- 786.7
21. Marcus Camby ---- 754.6
22. Tyson Chandler ---- 746.2
23. Kenyon Martin ---- 725.9
24. LeBron James ---- 713.7
25. Shawn Marion ---- 702.5

I think that's a pretty damn interesting list. Jason Collins, Odom, Dirk, and Haywood are the only guys who really surprise me on that list. Not because they are bad defenders but I didn't think they would rank quite as high as they did. According to this, Jason Collins (6.5 Rebs + 0.9 Blks per 36) was a more valuable defensive player than Dwight Howard (12.9 Rebs + 2.2 Blks per 36). Shows you how impressive Collins' man defense must've been during his prime. Haywood is another big who was probably an underappreciated defender. It's not like people thought Haywood was a bad defender but I don't think anyone thought of him as one of the better defenders of the past 14 years.

And maybe Odom and Dirk were underrated defenders at the PF position. Both were strong guys with good feet who were strong defensive rebounders that could hold their own when it came to low post D. Both guys also put up respectable block totals in their prime so they weren't very bad rim protectors either.

I think a lot of people appreciate Iggy's defense and understand that he is an elite wing defender, but he still comes out looking really good in this. Ahead of Battier and Bowen, right below Metta. Simply put, Iggy is a spectacular defensive player who is still in his prime and playing great defense.

LeBron at 24th is also interesting. He is in the same tier as Josh Smith, Gerald Wallace, and Shawn Marion which feels acceptable. But some people think LeBron is a way better defender than those 3 guys were. 8th best wing defender of the past 14 years does feel a little disappointing for LeBron, but all the guys ahead of him were excellent on defense so it's hard to argue with it. But at the end of the day, this is still a very impressive ranking for LeBron and further confirms the fact that he is a great defensive player.

Jason Kidd is also someone who looks really good here. He is the only PG that is even close to the top of this list in terms of defensive value. Kidd is 17th on this list, the next PG is Rajon Rondo at 50th. People say PG defense is overrated and that's true to a degree. But Kidd was the exception to that rule, he was an elite perimeter defender who could even guard SFs like LeBron at times. One of the most unique players in history and this is just more proof of that.

I'm not taking this as gospel and I definitely disagree with some of it, but this is really interesting and I appreciate the effort to try and create something like RAPM Over Average/Replacement as a tool to measure career value. Good stuff even if the specifics of it can be cleaned up.

[RSCD3_]

I love the idea of this. In theory this could be basketball's answer to baseball's Wins Above Replacement. Basketball will never be able to be measured by metrics to the degree baseball can because of nature of the game, but we can definitely learn things through RAPM and measurements like that.

Seems like there are questions about the execution of this, I don't know enough about RAPM to know how reliable this data is. But I love the idea of RAPM Over Average as a career measurement metric, and I thought the results passed the eye test and at the very least is pretty accurate even if you should take it with a grain of salt.

It's easier to measure and understand offensive value than defensive value, so I was interested in seeing who the top defenders were according to this data.

Defensive RAPM Above Average: Since 2001
1. Tim Duncan ---- 2502.4
2. Kevin Garnett ---- 2198.6

3. Ben Wallace ---- 1437.3
4. Metta World Peace ---- 1277.9
5. Andre Iguodala ---- 1202.6
6. Shane Battier ---- 1087.3
7. Jermaine O'Neal ---- 1042.8
8. Luol Deng ---- 1026.3
9. Jason Collins ---- 1024.5
10. Dwight Howard ---- 1001.4
11. Rasheed Wallace ---- 996.8
12. Nene ---- 995.8
13. Bruce Bowen ---- 948.4
14. Lamar Odom ---- 941.8

15. Josh Smith ---- 887.2
16. Gerald Wallace ---- 880.9
17. Jason Kidd ---- 879.9
18. Dirk Nowitzki ---- 837
19. Dikembe Mutombo ---- 828.8 (despite missing most of his prime, probably at 2500+ career value)
20. Brendan Haywood ---- 786.7
21. Marcus Camby ---- 754.6
22. Tyson Chandler ---- 746.2
23. Kenyon Martin ---- 725.9
24. LeBron James ---- 713.7
25. Shawn Marion ---- 702.5

I think that's a pretty damn interesting list. Jason Collins, Odom, Dirk, and Haywood are the only guys who really surprise me on that list. Not because they are bad defenders but I didn't think they would rank quite as high as they did. According to this, Jason Collins (6.5 Rebs + 0.9 Blks per 36) was a more valuable defensive player than Dwight Howard (12.9 Rebs + 2.2 Blks per 36). Shows you how impressive Collins' man defense must've been during his prime. Haywood is another big who was probably an underappreciated defender. It's not like people thought Haywood was a bad defender but I don't think anyone thought of him as one of the better defenders of the past 14 years.

And maybe Odom and Dirk were underrated defenders at the PF position. Both were strong guys with good feet who were strong defensive rebounders that could hold their own when it came to low post D. Both guys also put up respectable block totals in their prime so they weren't very bad rim protectors either.

I think a lot of people appreciate Iggy's defense and understand that he is an elite wing defender, but he still comes out looking really good in this. Ahead of Battier and Bowen, right below Metta. Simply put, Iggy is a spectacular defensive player who is still in his prime and playing great defense.

LeBron at 24th is also interesting. He is in the same tier as Josh Smith, Gerald Wallace, and Shawn Marion which feels acceptable. But some people think LeBron is a way better defender than those 3 guys were. 8th best wing defender of the past 14 years does feel a little disappointing for LeBron, but all the guys ahead of him were excellent on defense so it's hard to argue with it. But at the end of the day, this is still a very impressive ranking for LeBron and further confirms the fact that he is a great defensive player.

Jason Kidd is also someone who looks really good here. He is the only PG that is even close to the top of this list in terms of defensive value. Kidd is 17th on this list, the next PG is Rajon Rondo at 50th. People say PG defense is overrated and that's true to a degree. But Kidd was the exception to that rule, he was an elite perimeter defender who could even guard SFs like LeBron at times. One of the most unique players in history and this is just more proof of that.

I'm not taking this as gospel and I definitely disagree with some of it, but this is really interesting and I appreciate the effort to try and create something like RAPM Over Average/Replacement as a tool to measure career value. Good stuff even if the specifics of it can be cleaned up.

well LeBron wasn't part of the NBA during 01-03 and wasn't a plus defender until around the 07 season so it makes sense why he is only 24 on here

[colts18]

colts18,
could you explain how you did that? Use Durant as example and show step by step how you get his total value, please.

Here is the Kevin Durant data. Above average is RAPM / 100 * Possessions played

Code: Select all

year   poss   Off per 100   Def per 100   Ovr per 100  off   def   above avg   above rep
2008   5581.5   -1.5   -3   -4.5   -83.7   -167.5   -251.2   -195.4
2009   5656   -1.8   -2.8   -4.6   -101.9   -158.2   -260.2   -203.6
2010   6777.5   2.9   1.2   4.1   196.4   81.4   277.9   345.7
2011   7315.5   3.9   0.6   4.5   286.0   43.8   329.2   402.4
2012   6585   3.5   -0.1   3.4   230.9   -6.6   223.9   289.7
2013   5475   3.0   0.0   3.0   161.9   1.8   163.7   218.5
2014   3925.5   3   0.8   3.8   117.8   31.4   149.2   188.4


[colts18]

It doesn't make sense to me. The whole sample is including multiple studies. Half of it is NPI RAPM, other parts of it are RAPM informed RAPM and then the 2013 RAPM is prior informed but not done by J.E. but v-zero instead. 2014 is using talkingpractice's NPI RAPM. Looking at all of that as a whole makes no sense to me at all. Different parameters were used for each of the different models. They are then, by definition, different models that use the same regression method and have just been given the same label by us (and whoever ran the regression).

I should correct your post because my whole data set is from J.E. except for the 2013 season (I don't believe he posted one from that year). So I assume J.E. used the same methodology for his RAPM's.

[SideshowBob]

It doesn't make sense to me. The whole sample is including multiple studies. Half of it is NPI RAPM, other parts of it are RAPM informed RAPM and then the 2013 RAPM is prior informed but not done by J.E. but v-zero instead. 2014 is using talkingpractice's NPI RAPM. Looking at all of that as a whole makes no sense to me at all. Different parameters were used for each of the different models. They are then, by definition, different models that use the same regression method and have just been given the same label by us (and whoever ran the regression).

I should correct your post because my whole data set is from J.E. except for the 2013 season (I don't believe he posted one from that year). So I assume J.E. used the same methodology for his RAPM's.

Cool. So to be clear, can you tell me specifically which data sets you used?

If its the 01-06 NPI and then 07-12 PI then that's going to cause problems, I would split those up. And then 2014 RAPM he's provided is completely different from his 01-06 NPI/01-06 PI/07-12 PI, as its actually just a 12-14 multi-year RAPM with more weight on 14. Also, having v-zero's 2013 RAPM also muddles things up.

As I said earlier, more than anything its a labeling issue. We're calling any APM model derived through ridge-regression RAPM, which is fine but it needs to be recognized that there are different ways of doing this, and comparison between is not straightforward.

[Doctor MJ]

colts18,
could you explain how you did that? Use Durant as example and show step by step how you get his total value, please.

Here is the Kevin Durant data. Above average is RAPM / 100 * Possessions played

Code: Select all

year   poss   Off per 100   Def per 100   Ovr per 100  off   def   above avg   above rep
2008   5581.5   -1.5   -3   -4.5   -83.7   -167.5   -251.2   -195.4
2009   5656   -1.8   -2.8   -4.6   -101.9   -158.2   -260.2   -203.6
2010   6777.5   2.9   1.2   4.1   196.4   81.4   277.9   345.7
2011   7315.5   3.9   0.6   4.5   286.0   43.8   329.2   402.4
2012   6585   3.5   -0.1   3.4   230.9   -6.6   223.9   289.7
2013   5475   3.0   0.0   3.0   161.9   1.8   163.7   218.5
2014   3925.5   3   0.8   3.8   117.8   31.4   149.2   188.4


I'll add that Durant's relationship to +/- stats was a seminal case study to me. TrueHoop actually wrote an article about how hideously bad his +/- was early on - to be clear they weren't saying "so Durant's a bust", they were just noting that in spite of the obvious potential he was showing, things weren't clicking yet.

When I talk about young players, I'll often talk about the strategy of handing the team off to the proto-star before he deserves it and how that can be a great thing or a problematic thing. Durant's the most clear example of "great" to me. The very idea that you would have pulled his card in his first couple years because it wasn't working makes us all ill to think about - clearly he was in the process of figure that role out, and having him spend time mastering a lesser role might have permanently stunted his growth.

As I say all this, it's understandable if one is skeptical about this. Clearly his skills didn't just emerge from nowhere his 3rd year - and I'm not saying they did, I'm just saying in terms of having a functional team machine based around Durant playing the star role, it was a night and day difference between his 2nd and 3rd year.

[Doctor MJ]

I love the idea of this. In theory this could be basketball's answer to baseball's Wins Above Replacement. Basketball will never be able to be measured by metrics to the degree baseball can because of nature of the game, but we can definitely learn things through RAPM and measurements like that.

Seems like there are questions about the execution of this, I don't know enough about RAPM to know how reliable this data is. But I love the idea of RAPM Over Average as a career measurement metric, and I thought the results passed the eye test and at the very least is pretty accurate even if you should take it with a grain of salt.

It's easier to measure and understand offensive value than defensive value, so I was interested in seeing who the top defenders were according to this data.

Defensive RAPM Above Average: Since 2001
1. Tim Duncan ---- 2502.4
2. Kevin Garnett ---- 2198.6

3. Ben Wallace ---- 1437.3
4. Metta World Peace ---- 1277.9
5. Andre Iguodala ---- 1202.6
6. Shane Battier ---- 1087.3
7. Jermaine O'Neal ---- 1042.8
8. Luol Deng ---- 1026.3
9. Jason Collins ---- 1024.5
10. Dwight Howard ---- 1001.4
11. Rasheed Wallace ---- 996.8
12. Nene ---- 995.8
13. Bruce Bowen ---- 948.4
14. Lamar Odom ---- 941.8

15. Josh Smith ---- 887.2
16. Gerald Wallace ---- 880.9
17. Jason Kidd ---- 879.9
18. Dirk Nowitzki ---- 837
19. Dikembe Mutombo ---- 828.8 (despite missing most of his prime, probably at 2500+ career value)
20. Brendan Haywood ---- 786.7
21. Marcus Camby ---- 754.6
22. Tyson Chandler ---- 746.2
23. Kenyon Martin ---- 725.9
24. LeBron James ---- 713.7
25. Shawn Marion ---- 702.5

I think that's a pretty damn interesting list.

Me too.

Jason Collins, Odom, Dirk, and Haywood are the only guys who really surprise me on that list. Not because they are bad defenders but I didn't think they would rank quite as high as they did. According to this, Jason Collins (6.5 Rebs + 0.9 Blks per 36) was a more valuable defensive player than Dwight Howard (12.9 Rebs + 2.2 Blks per 36). Shows you how impressive Collins' man defense must've been during his prime. Haywood is another big who was probably an underappreciated defender. It's not like people thought Haywood was a bad defender but I don't think anyone thought of him as one of the better defenders of the past 14 years.

Also consider things like the team coordination involved in defensive rebounding. A guy like Nene when you look at the regression correlation in defensive rebounding puts him right up there with guys who actually rack up the big rebounding numbers.

And maybe Odom and Dirk were underrated defenders at the PF position. Both were strong guys with good feet who were strong defensive rebounders that could hold their own when it came to low post D. Both guys also put up respectable block totals in their prime so they weren't very bad rim protectors either.

Jason Kidd is also someone who looks really good here. He is the only PG that is even close to the top of this list in terms of defensive value. Kidd is 17th on this list, the next PG is Rajon Rondo at 50th. People say PG defense is overrated and that's true to a degree. But Kidd was the exception to that rule, he was an elite perimeter defender who could even guard SFs like LeBron at times. One of the most unique players in history and this is just more proof of that.

It is very interesting to see how Kidd stacks up on offense and on defense. His offense is only slightly ahead of his defense relative to average. I would assume that becomes a bit bigger offensive edge when you look at above replacement, but still, as someone prone to saying Kidd's a star because of his offense like all point guards are, it's fodder to look at Kidd primarily as an all-around guy.

I'll note that I think it's clear that Kidd's career arc on offense is much steeper than on defense. When Kidd was a star, it was dominated by offense, but his defense has largely remained quite solid with longevity.

Also worth noting that if you sort by defensive rating normalized for minutes, Kidd doesn't stand out so clearly. Here's a list of noteworthy point guards with their overall rank (note, some of the guys in the ranking don't have the minutes to really warrant inclusion, but I'm not going to hand edit the file for something like this)

35. Ricky Rubio +2.02
81. John Stockton +1.40 (obviously small sample size, but still, dude was super old, so WOW!)
85. Kyle Lowry +1.33
91. Rajon Rondo +1.29
103. Jason Kidd +1.16

Particularly if you grant the flat defensive career arc I mention before with Kidd, this very much indicates to me that he's not an exception to the rule about point guards & defense. He's good, but he's not a defensive star. The lack of falloff though meant he frankly could still be a starting point guard in the league in a way that a guy like Nash maybe couldn't be (even if he were healthy).

I'm not taking this as gospel and I definitely disagree with some of it, but this is really interesting and I appreciate the effort to try and create something like RAPM Over Average/Replacement as a tool to measure career value. Good stuff even if the specifics of it can be cleaned up.

Agreed, and I like your thoughts on this.

[Doctor MJ]

Playing around with the data a little more I realize you really need to adjust for pace - or perhaps just ignore possessions altogether and use minutes. A guy like Chris Paul seems like he's getting hammered by it.

[MisterWestside]

I love the idea of this. In theory this could be basketball's answer to baseball's Wins Above Replacement.

Wouldn't be much of an answer. As far as all-in-one metrics go, these stats aren't seeing WAR, and pure RAPM isn't even the best single metric. And they're only useful for impact, which makes up a fraction of basketball analysis.

Basketball will never be able to be measured by metrics to the degree baseball can because of nature of the game, but we can definitely learn things through RAPM and measurements like that.

Agreed here. However, the analytical future of the sport lies in video tracking, and that's what I'm amped for.

[Doctor MJ]

I love the idea of this. In theory this could be basketball's answer to baseball's Wins Above Replacement.

Wouldn't be much of an answer. As far as all-in-one metrics go, these stats aren't seeing WAR, and pure RAPM isn't even the best single metric. And they're only useful for impact, which makes up a fraction of basketball analysis.

To your last sentence: WAR is only useful for impact too, and there's certainly more to baseball than that, albeit not as much more as basketball has.

More interesting point: Would it be better to use XRAPM or something along those lines for this kind of analysis than RAPM which by definition is not the best indicator by itself?

It would certainly be good to see it done for XRAPM as well. On a simplistic level, what we are doing in this analysis is treating RAPM as an all-in-one stat, and XRAPM is actually the attempt to use RAPM to make an all-in-one stat. I'll be honest though: I don't know if I'd feel as comfortable analyzing those results as I do with a honed form of this analysis for the same reasons I'm reticent using XRAPM generally: I don't use anything as a pure all-in-one, and I think it's easier to mentally mod RAPM than XRAPM.

[lorak]

colts18,
could you explain how you did that? Use Durant as example and show step by step how you get his total value, please.

Here is the Kevin Durant data. Above average is RAPM / 100 * Possessions played

Code: Select all

year   poss   Off per 100   Def per 100   Ovr per 100  off   def   above avg   above rep
2008   5581.5   -1.5   -3   -4.5   -83.7   -167.5   -251.2   -195.4
2009   5656   -1.8   -2.8   -4.6   -101.9   -158.2   -260.2   -203.6
2010   6777.5   2.9   1.2   4.1   196.4   81.4   277.9   345.7
2011   7315.5   3.9   0.6   4.5   286.0   43.8   329.2   402.4
2012   6585   3.5   -0.1   3.4   230.9   -6.6   223.9   289.7
2013   5475   3.0   0.0   3.0   161.9   1.8   163.7   218.5
2014   3925.5   3   0.8   3.8   117.8   31.4   149.2   188.4


Thanx. What about replacement level? What value you assumed and why?

Also why PI RAPM over NPI RAPM? (however you used NPI for 2001 season)

[blessofcurse]

Wow. So Dirk is underrated defensively or it is bcoz he played a defensive center his almost entire career?

Sent from my V790 using RealGM Forums mobile app

[Texas Chuck]

Wow. So Dirk is underrated defensively or it is bcoz he played a defensive center his almost entire career?

Sent from my V790 using RealGM Forums mobile app

It obviously helps Dirk when he shares the court with even a marginal center who is competent defensively. Its not like Dirk played with great defensive centers excepting the one year with Tyson and even he was a sub 30mpg guy. But Bradley, Damp, Diop, Haywood, heck even this year with the mercurial Dalembert alongside Dirk worked really quite well.

But it wasnt until Avery and Rick were coaching that he wasnt spending a ton of time as the de facto center alongside Walker or Van Horn or Crosure (sp) or Walt Williams or whatever SF Nellie labeled the center.

But in addition to that Dirk is underrated defensively. Most people know by now what a solid defensive rebounder he was in his prime and he amped that up in the PS ot the point that for his career he, Duncan, and KG all average the same number of DRB per game. But he also is a legit 7 feet which helps. He has great hands and stole the Karl Malone move of stripping the ball as guys go up rather than going for the block. He holds position really well--underrated strength and of course he knows what scorers are looking to do. He closes on shooters really well too. He didnt crash the offensive boards and so was always back on D.

His flaws are terrible perimeter defense. He's great at showing on the PnR but dreadful at getting back to his assignment after the hedge. In his prime, he ole'd guys a lot because he was instructed to avoid early fouls so the offense didnt collapse(post-Nash era) with him on the bench. Slow on rotations as the help defender. Never takes charges and not really a shot-blocker as a help defender. Most of his blocks are as a on-man defender.

But yeah Irk was pretty much a myth by his 3rd year in the league. He's always been a decent enough defender.

[colts18]

Playing around with the data a little more I realize you really need to adjust for pace - or perhaps just ignore possessions altogether and use minutes. A guy like Chris Paul seems like he's getting hammered by it.

The stats I posted included postseason so a guy like CP3 will take a hit while Kobe and Duncan benefit by that.

[ceiling raiser]

Playing around with the data a little more I realize you really need to adjust for pace - or perhaps just ignore possessions altogether and use minutes. A guy like Chris Paul seems like he's getting hammered by it.

The stats I posted included postseason so a guy like CP3 will take a hit while Kobe and Duncan benefit by that.

Do you have offense/defense per 100 for notable players and other guys in the top 10 of each?