RealGM Greatest Peaks Project (2022): #6 - 2002-03 Tim Duncan

[trex_8063]

Just realized I forgot to link the video I was referring to above. Have edited that in. Sorry....

[Doctor MJ]

Fascinating to see the way Duncan's stature has grown over time.

[falcolombardi]

Fascinating to see the way Duncan's stature has grown over time.

He has gone 7-6-6 in the last projects

[DraymondGold]

Well, depending on how seriously you take wowy with hakeem you could credibly argue what he did after 86 was as or more impressive. Ala, 10 win pace rockets in 88 win at 45 win pace with him. 2-10 rockets(without hakeem) in 92 go 40-20 with him and then win 53 games the following season with hakeem playing the whole year. Taken at face value those are argubale _more_ impressive than the 86 regular season though we don't really get a fair playoff comparison.

What happens if you compare bird to hakeem's best scoring years rs and ps instead of specfically 94.

For me I'm not sure it would be "credible" to be that selective with the WoWY samples.

Overall his WoWY stuff is very good I believe but the 86, 88, 91 and 91-92 composite numbers from Ben's old spreadsheet are far more pedestrian. Latterly '95 looks very good, '96 good, '97 pedestrian, I would suggest.

Sidenote: The nature of the '92 absence too, may, depending on your reading, harm Olajuwon from an intangibles perspective.

there's a spreadsheet? I thought wowy was just you take the record with thep layer and the record without the player

Hi OhayoKD! haha, I noticed you were using a different definition for WOWY a few threads back and was wondering if it would ever come up All good though

What is WOWY? Why does it not just look at record? There's definitely information to be learned from looking at a team's record, so your self-calculated WOWY numbers are definitely valuable. But usually analysts think there's more information from judging team results by Margin of Victory over record (especially if you can adjust margin of victory by the opponent). For example, a team's margin of victory in the regular season is usually a better predictor of playoff success than record (especially if it's adjusted for opponent difficulty, etc.). So that's what WOWY looks at. In other words, WOWY is a "game level" plus minus, compared to the normal "possession level" plus minus. It looks at the change in the (possibly opponent-adjusted) Margin of Victory when player plays games vs when they're out.

Can you Regularize WOWY just like you regularize Plus Minus Data? Yep! Just like Adjusted Plus Minus (APM) has Regularized Adjusted Plus Minus (RAPM), which limits the impact of outliers, we can regularize/regress WOWY to get WOWYR. Why would we want to limit the impact of outliers? Well, people think there's usually not much difference between a 25 point blowout and a 35 point blowout... a blowout's a blowout and we don't really care if the 10th-15th players on a roster scored 10 extra points in 7 minutes of garbage time. So we curve down blowouts so there's not much difference between them.

Sample size? Plus minus data requires a lot of possessions to be stable and trustworthy. Similarly, WOWY requires a lot of games to be stable and trustworthy. If we want to have confidence in WOWY, people usually recommend looking at a many-season sample (e.g. a player's full prime). This loses a bit of granularity (e.g. if a player was much better in 1 year vs the next), but it gives us more confidence in the data.

Where is the WOWY spreadsheet? the regularized WOWY data can be found here: https://backpicks.com/metrics/wowyr/, along with links to a larger explanation. More granular game-specific WOWY data (e.g. tracking exactly when players missed games in different seasons) can be found here: https://backpicks.com/metrics/wowy-data/. Patreon subscription required unfortunately.

Let me know if you have any questions!

[Owly]

Well, depending on how seriously you take wowy with hakeem you could credibly argue what he did after 86 was as or more impressive. Ala, 10 win pace rockets in 88 win at 45 win pace with him. 2-10 rockets(without hakeem) in 92 go 40-20 with him and then win 53 games the following season with hakeem playing the whole year. Taken at face value those are argubale _more_ impressive than the 86 regular season though we don't really get a fair playoff comparison.

What happens if you compare bird to hakeem's best scoring years rs and ps instead of specfically 94.

For me I'm not sure it would be "credible" to be that selective with the WoWY samples.

Overall his WoWY stuff is very good I believe but the 86, 88, 91 and 91-92 composite numbers from Ben's old spreadsheet are far more pedestrian. Latterly '95 looks very good, '96 good, '97 pedestrian, I would suggest.

Sidenote: The nature of the '92 absence too, may, depending on your reading, harm Olajuwon from an intangibles perspective.

there's a spreadsheet? I thought wowy was just you take the record with thep layer and the record without the player

There are different versions. Record is a pretty crude one. Binary w/l versions lose a lot of information unnecessarily where points differential versions don't. Surface level versions also fail to account for home/road, schedule strength, other injuries unbalancing the samples etc.

Ben Taylor (ElGee) put out a spreadsheet, I think before most of his [written/numerical] stuff went behind a paywall (not that I don't appreciate the free stuff - tbh that some antivirus software hates his site is more of a concern). It would have been 2015 or earlier. It attempted to control for much of the above, give an SRS change and various other numbers.

Anyway
1) Someone has now already given a detailed answer, who is surely more up to date and likely better acquainted with Ben's numbers.
2) I'm by no means an expert.

I think there may be some touching on WoWY in Ben's impact episode in the metrics-y series he did (can't check right now).

Watch on YouTube

[Doctor MJ]

Fascinating to see the way Duncan's stature has grown over time.

He has gone 7-6-6 in the last projects

Sure, but in 2012 we had Duncan at 9 (while Hakeem was at 5). We've seen the two of them flip well after either of them were playing peak basketball.

I'm not saying that's wrong, but it's interesting, and worth understanding how people are reaching new and different conclusions.

[Dutchball97]

Fascinating to see the way Duncan's stature has grown over time.

He has gone 7-6-6 in the last projects

Sure, but in 2012 we had Duncan at 9 (while Hakeem was at 5). We've seen the two of them flip well after either of them were playing peak basketball.

I'm not saying that's wrong, but it's interesting, and worth understanding how people are reaching new and different conclusions.

Looking at the 2012 project it doesn't seem too strange to see things change a bit. There was only one vote without extra ballot options, seemingly no explanation needed and different seasons for one player needing to be tactically switched or be discounted. Lots of close calls that could've been a different outcome with the current voting system. Maybe just a different view on the game as well. Russell was 3rd despite still not being in now, while 09 LeBron made it all the way to 10th despite being in plenty of run-offs. Since it was a new concept people were also still debating the criteria in the voting threads. Not to be disrespectful but the discourse also looks a bit more anecdotally driven than statistically with a lot of news articles quoted and the most in depth stats used being things like SRS and TS%. Besides that, the voting pool seemed to shrink pretty fast compared to a more consistent turnout this edition. Smaller voting pools can also lead to more variance.

All in all plenty of reasons for things to change over time.

[Djoker]

Fascinating to see the way Duncan's stature has grown over time.

In 2012, Lebron had the #10 highest peak.

But yes Duncan jumped a lot as well.

[Doctor MJ]

He has gone 7-6-6 in the last projects

Sure, but in 2012 we had Duncan at 9 (while Hakeem was at 5). We've seen the two of them flip well after either of them were playing peak basketball.

I'm not saying that's wrong, but it's interesting, and worth understanding how people are reaching new and different conclusions.

Looking at the 2012 project it doesn't seem too strange to see things change a bit. There was only one vote without extra ballot options, seemingly no explanation needed and different seasons for one player needing to be tactically switched or be discounted. Lots of close calls that could've been a different outcome with the current voting system. Maybe just a different view on the game as well. Russell was 3rd despite still not being in now, while 09 LeBron made it all the way to 10th despite being in plenty of run-offs. Since it was a new concept people were also still debating the criteria in the voting threads. Not to be disrespectful but the discourse also looks a bit more anecdotally driven than statistically with a lot of news articles quoted and the most in depth stats used being things like SRS and TS%. Besides that, the voting pool seemed to shrink pretty fast compared to a more consistent turnout this edition. Smaller voting pools can also lead to more variance.

All in all plenty of reasons for things to change over time.

Okay, but that's all explanations that don't deal with the actual basketball.

In general, people used to see Olajuwon as better than Duncan, now it's flipped. Assuming this is about something other than differences in the voting base, the question is:

What has caused Duncan to be elevated while Olajuwon has fallen?

This is particularly interesting because we're not talking about drastically different types of players. If someone bumps Duncan over Magic they're almost certainly saying, "Magic is certainly better on offense, but I think Duncan's defense gives him the edge." In the case of Duncan & Olajuwon, we seem to have a situation where people have shifted on both offense and defense in the same direction at a time after the players played their prime. What is causing this change?

[DraymondGold]

Sure, but in 2012 we had Duncan at 9 (while Hakeem was at 5). We've seen the two of them flip well after either of them were playing peak basketball.

I'm not saying that's wrong, but it's interesting, and worth understanding how people are reaching new and different conclusions.

Looking at the 2012 project it doesn't seem too strange to see things change a bit. There was only one vote without extra ballot options, seemingly no explanation needed and different seasons for one player needing to be tactically switched or be discounted. Lots of close calls that could've been a different outcome with the current voting system. Maybe just a different view on the game as well. Russell was 3rd despite still not being in now, while 09 LeBron made it all the way to 10th despite being in plenty of run-offs. Since it was a new concept people were also still debating the criteria in the voting threads. Not to be disrespectful but the discourse also looks a bit more anecdotally driven than statistically with a lot of news articles quoted and the most in depth stats used being things like SRS and TS%. Besides that, the voting pool seemed to shrink pretty fast compared to a more consistent turnout this edition. Smaller voting pools can also lead to more variance.

All in all plenty of reasons for things to change over time.

Okay, but that's all explanations that don't deal with the actual basketball.

In general, people used to see Olajuwon as better than Duncan, now it's flipped. Assuming this is about something other than differences in the voting base, the question is:

What has caused Duncan to be elevated while Olajuwon has fallen?

This is particularly interesting because we're not talking about drastically different types of players. If someone bumps Duncan over Magic they're almost certainly saying, "Magic is certainly better on offense, but I think Duncan's defense gives him the edge." In the case of Duncan & Olajuwon, we seem to have a situation where people have shifted on both offense and defense in the same direction at a time after the players played their prime. What is causing this change?

Great points all around!

It's cool to see that there's an increased statistical literacy and increased turnout. And like Doctor MJ says, even if Duncan stays ranked around the same number, that doesn't mean everyone else is constant (e.g. LeBron rises, Hakeem falls).

After the project, it would be cool to do a historical comparison. Doctor MJ, I think I remember hearing you had a Google Doc that kept track of every order going years back? Or was that of for the Greatest Player project, instead of the Greatest 1-year Peak project?

To answer your question, I think the thing that swayed me for Duncan over Hakeem was the metrics (I know, big shocker considering my posting history... ). Especially considering how similar they are and how similar their context is, I feel better thinking the metrics are making an apples to apples comparison.

Qualitatively, I have two concerns about Hakeem.
1) I have serious concerns about his lack of passing, and I see Duncan as the clearly better playmaker and facilitator. This lead expands if we consider screening, offensive rebounding, and other forms of offensive (/cultural) leadership. People tend to oversimplify two-way players into Defense and Scoring. When I oversimplify, I tend to divide things into 3 parts: scoring, creation, and defense (so I rank creation higher relative to other posters).
Put another way, I think Hakeem's lack of playmaking and reliance on on-ball scoring gives me major portability concerns vs other players. While Duncan is also not the pinnacle of portability, his portability weaknesses are less concerning, and he also shows a greater mental willingness to play a more auxiliary role and be the team's cultural pillar (compared to Hakeem who had a greater reputation of being a somewhat-stubborn ballhog)
2) I worry people are overrating Hakeem's defense (at least for the peak years that they're arguing). The metrics and my eye test are very favorable on slightly younger Hakeem (e.g. 89, 90, maybe 93). But I worry by the time he finally starts to make ground with Concern 1 (though he never closes the gap), he starts to lose ground to peak Duncan (who peaked on offense and defense at the same time much more effectively than Hakeem).


To be clear, they're both top tier players and there's arguments for both. But during this project, I've struggled to be high on the players who perform well by Eye test / reputation among the analytical crowd, but have clear flaws by the metrics/qualitative analysis (i.e. Hakeem, KG, etc.). I think you've provided some great analysis Doctor MJ, so I'm happy to discuss more if you see flaws in my reasoning.

[falcolombardi]

Sure, but in 2012 we had Duncan at 9 (while Hakeem was at 5). We've seen the two of them flip well after either of them were playing peak basketball.

I'm not saying that's wrong, but it's interesting, and worth understanding how people are reaching new and different conclusions.

Looking at the 2012 project it doesn't seem too strange to see things change a bit. There was only one vote without extra ballot options, seemingly no explanation needed and different seasons for one player needing to be tactically switched or be discounted. Lots of close calls that could've been a different outcome with the current voting system. Maybe just a different view on the game as well. Russell was 3rd despite still not being in now, while 09 LeBron made it all the way to 10th despite being in plenty of run-offs. Since it was a new concept people were also still debating the criteria in the voting threads. Not to be disrespectful but the discourse also looks a bit more anecdotally driven than statistically with a lot of news articles quoted and the most in depth stats used being things like SRS and TS%. Besides that, the voting pool seemed to shrink pretty fast compared to a more consistent turnout this edition. Smaller voting pools can also lead to more variance.

All in all plenty of reasons for things to change over time.

Okay, but that's all explanations that don't deal with the actual basketball.

In general, people used to see Olajuwon as better than Duncan, now it's flipped. Assuming this is about something other than differences in the voting base, the question is:

What has caused Duncan to be elevated while Olajuwon has fallen?

This is particularly interesting because we're not talking about drastically different types of players. If someone bumps Duncan over Magic they're almost certainly saying, "Magic is certainly better on offense, but I think Duncan's defense gives him the edge." In the case of Duncan & Olajuwon, we seem to have a situation where people have shifted on both offense and defense in the same direction at a time after the players played their prime. What is causing this change?

Well, we seem to look at plus-minus and rapm more now, and those metrics are incredibly kind to duncan

We also may have looked at numbers mpre closely and realize that for how much more easthetical and impressive hakeem moves look, his actual scoring, creation, efficiency are fairly close to duncan numbera

[LukaTheGOAT]

...

Definitely agree that my middle point was mislabeled, although it doesn't really affect my argument (the more meaningful part was the first point - him starting the season off hot with KD). I think your synopsis looks fair to me:

Part 1 (Game 1 to Game 15ish): Good start with KD
Part 2 (Game 15ish to game 40ish): Slump, with KD
Part 3 (Game 40ish to game 55ish): Improvement out of slump, with KD
Part 4 (Game 55ish to game 60ish): smaller slump, with KD
Part 5 (Game 60ish to game 82ish): further improvement, without KD
Part 6 (playoffs): further improvement with KD

So, what do I see here? Complete inconsistency until the postseason. I look at this and just see rockiness with no clear trend or pattern. Once again, I just don't really see the statistical evidence for this theory.

But I don't think there's much more discussion left to be had on that front. I think we've refined our summary of the statistics to a satisfactory degree, and now it's up to interpretation. I guess I've made my interpretation clear, but it's only right that I flip it:

I need to consider the possibility that all of the Warriors players and staff are right and that I'm wrong. Because while I don't think their word is gospel (no one does, of course), they do certainly have value to add in many cases. I did quote Draymond Green saying that the Warriors had begun to be figured out in 2016 and probably don't win another championship without KD in post #76 of the last thread (voting for #5) when I replied to you. And I did believe that someone like Draymond saying that had value and needed to be considered, so it's only reasonable that I do the same here.

So while I'm not actually convinced that was the case, I'm going to start looking at this from the other perspective of Curry's slumps being caused by him struggling to fit with KD.

To me, Curry took a step back to help KD get comfortable (which shows good leadership, unselfishness, and team chemistry). But at some point, around Game 15ish, he took too much of a step back, started slumping, and needed to be more aggressive to get out of the slump.

Through the first 20 games of the season, Curry averaged 26.9 PPG on 66.5% TS% and the Warriors were 17-3, a bonkers 70-win pace. In this same stretch stretch, Durant averaged 27.3 PPG on 68.0% TS%.

After this is when Curry took a bigger step back (the initial step back was the standard one that you'd expect when two MVP level scorers team up - both KD and Steph seemed to have a drop in FGA relative to 2016) and became more passive to help KD get comfortable, demonstrating his strong leadership and selflessness.

But... was KD really not comfortable? He was averaging 27 on nearly 70% TS% and the Warriors were on a 70 win pace, and Curry thought "Well gee, hold on a minute, I need to change this up!" So now, under the presumption that Curry did truly take a step back after 15-20 games into the season, it only leaves me wondering ... why? He was trying to fix a problem that wasn't there? I get why someone might look at that and want to reward him for his great leadership, but I'd do the opposite and be critical because unless I'm missing a key detail here, it was wholly unnecessary for him to suddenly turn passive.

Thanks for the reply! And I tend to agree, we’ve gone about as deep as we can go on this one. I also appreciate your willingness to consider the other side — that’s a really valuable skill in a discussion.

I think your point that there still was some inconsistency (even if it was caused by fitting with KD) is valid. Your comment that Steph probably didn’t need to take as much of a step back as he did (even if it does show good leadership and chemistry) is also true.

If Curry didn't want to be too aggressive, didn't want to be too selfish, and didn't want to be too much of a ballhog when playing with KD, it makes sense that getting some time without KD (and actually having the team's performance improve!) would help Curry feel more confident upping his aggression a little bit. Then, once KD returned for the playoffs, Curry realized he could maintain his 2016 level of aggression, that KD would figure it out, and that the team would be all the better for it.

If it's so difficult for him to fit in with KD and it takes the bulk of the regular season along with discussions with guys like Bob Myers for him to finally **consistently** figure it out... is that not something that he should be docked for?

I think some of the most similar cases to Curry being joined by KD during his peak is peak Jordan being joined by Phil Jackson and peak LeBron switching teams to join Wade and the Heat.

Jordan was fairly effective at combining his peak offensive and defensive value at the same (that’s one of the reasons I have peak Jordan over LeBron), but he also didn't necessarily have his peak regular season at the same time as his peak postseason (1991), largely due to adapting to fit mostly with Phil Jackson's triangle. FiveThirtyEight actually did a study and found that it took the Bulls (and Jordan) a full ~1.5 years to fully embrace and maximize the value of Phil Jackson’s scheme (link to the data here: https://projects.fivethirtyeight.com/complete-history-of-the-nba/#warriors). LeBron similarly took time a little over a season to maximize his value when fitting alongside Wade and the Heat.

For these cases, I think it takes an underrated amount of time to maximize value in a newer scheme, and I tend to be more lenient for the kinds of drops in value vs something like coasting, but you're right that it is still a drop in value, which people might want to dock him for.

Near the end of the season, Curry actually referenced how he was figuring out how to balance aggression: "With the roster that we have and adding a guy like KD, there's obviously going to be more of a balanced attack," Curry said. "It's pretty clear how that's kind of evolved over the course of the season. My job as the point guard of this team is to balance all the talent that we have, plus at the same time staying aggressive with my own game." [Source: https://www.goldenstateofmind.com/2017/4/5/15191910/nba-2017-warriors-at-suns-preview-draymond-green-andre-iguodala-out].
Journalists and analysts at the time interpreted it similarly to how I did: Curry improved his aggression to 2016 form when KD was out, then realized he could maintain this aggression with KD on the court in the playoffs. [Source: https://goldengatesports.com/2017/05/26/golden-state-warriors-curry-impact-durant/]. I also remember KD also having a quote near the start of the playoffs, where he said getting some time off to see the success of a more aggressive Curry also helped him learn how to better fit with Curry and not take away from Curry, but I haven't been able to track that down yet.

I certainly agree with his assessment of the role of a point guard. Which sorta vocalizes why I feel inclined to dock Curry for the sudden passiveness if it really was because he wanted to take a step back when a step back wasn't necessary at all. Yeah, it was his job to remain aggressive with his own game and ... he apparently didn't fully grasp that balance until KD came back for the playoffs.

There's certainly a school of thought that might argue "well he figured it out in the playoffs where it all came together for the Warriors and they had a nigh perfect postseason, so the regular season struggles shouldn't matter." Or one may argue "regular season struggles? the warriors won 67 games with an 11 SRS. who cares about steph's individual numbers?" Both certainly valid perspectives, so I'm not sure there's a right way to look at this. All of this just leaves me further wishing that Curry had a clear peak season where everything came together nicely.

Like you suggest, this brings us to a meta discussion about what each voter values and what their criteria are for the greatest peaks. Some of the major questions might be:
1. How much do we prefer “goodness” (how good a player is in general, or regardless of context/role) vs “value” (how much they helped their team in their specific context/role)
2. How do we evaluate players who are inconsistent in value over a season, and how much do we care if their specific context influenced their value?

I’ve made arguments that 2017 regular season curry was just as “good” as 2016 regular season curry, but he was definitely less valuable.

I’ve argued that his changes in regular season value were caused by the context of fitting next to KD (which I’m willing to be more lenient about compared to if he was dramatically coasting), but others might be less lenient here. I think your criteria is definitely valid — we just happen to have different criteria

3. The last question is: if we’ve established that a player has inconsistent value while still being just as “good” (e.g. if Curry was just as good in 2016/2017, but wasn’t able to combine the consistent “value” of the 2016 regular season with the 2017 postseason in one season), how does this compare to other peaks?

Conversations like this (more often focused on evaluating regular season coasting) have been one of the themes for me in this Greatest Peaks debate. Players often maximize their regular season value and their postseason value (or their offensive and defensive value) in different seasons.

For example, LeBron never combined the value of his 2013 regular season with the value of his 2012 postseason into a single season. Similarly, 2016 LeBron showed peak value in the postseason, but didn’t quite reach that level in the regular season.
Shaq peaked in regular season value in 2000, but his postseason peak value was 2001. Kareem maximized his regular season value earlier on, and he maximized his postseason value in 1977. Hakeem and Bird also struggled to maximize their defensive value at the same time as their offensive value.

To me, I’m more willing to be lenient with players whose value is inconsistent due to fit vs due to coasting, but the competition at the top is pretty close - so if you choose to dock Curry for not maximizing regular season and postseason value in a single year, that would be perfectly valid. Anyway, thanks for a great discussion on the topic!

There has been a lot of discussion about 2009 Lebron having an outlier playoffs due to small sample size. I think there has to be discussion about 2009 Lebron's season as a whole being outlier, because it would seem it trumps the best that Jordan ever had to offer.

If we look at their RS and or full-season metrics, 09 Lebron looks better on a rate basis than MJ's best seasons in the following respective metrics:

By RAPTOR (Since 77)

09 Lebron: 12.6

91 MJ: 12.3

Estimated Impact for RS(1952-2013)

09 Lebron: 10.6

88 MJ: 8.9

PIPM (Since 77)

09 Lebron: 9.83 (#1 All-time)

88 MJ: 8.58

TWPR for RS (Since 78)

09 Lebron: 89.78 (#1 All-Time)

88 MJ: 89.30

BPM (Since 74)

09 Lebron: 13.2

88 MJ: 13

I didn't bother to put the PS, as the gap between Lebron and MJ grows in the PS in favor of Lebron, and people are saying they are weary of 2009 Lebron's play being outlier. By including the RS and/or full season data here, I am showing that Lebron was historic unlike anyone we have ever seen, even outside of his PS play.

Your first question might be why is 1988 MJ's campaign coming out the best in some of these metrics, instead of 91 and it is very possible that 88 MJ's RS was the best of his career...it was his defensive peak after all. But furthermore, it reinforces the idea that MJ's regular season and PS peaks didn't necessarily happen in the same year (91 had a lower defensive motor).

This is counter to Lebron who put it altogether in 2009 and authored his most valuable RS and PS in the very same year, giving him a persuasive argument for having a better season than MJ ever did because of how well everything came together for him. Lebron wasn't just hot for a "couple weeks," he was insane for the whole year to the point that these metrics believe they have not seen anything like him.

Btw, I see people mentioning 2010 Lebron as a disappointment, and while I think 2010 is a worse player than 2009 Lebron, I also feel as if his elbow injury hampered his performance (along with the Delonte West situation). Someone actually did a thread, where when Lebron had more days of rest, he seemed to perform better. In 2011, he clearly was notably heavier and didn't beat guys off the dribble. To me 2009 Lebron is a different player from any other version of Lebron (further backed by these metrics that think he was more valuable). But to each their own.

The only metric MJ rates higher in is Backpicks BPM, which isn't going so much for direct value, as it is "attempting to measure the goodness of a player in several situations." By the numbers we have, it seems as if 09 Lebron might be the most valuable player ever.

Overall, 2009 Lebron's motor, focus on offensive rebounding, and scoring aggression were higher than any other points in his career.

[Doctor MJ]

Well, we seem to look at plus-minus and rapm more now, and those metrics are incredibly kind to duncan

We also may have looked at numbers mpre closely and realize that for how much more easthetical and impressive hakeem moves look, his actual scoring, creation, efficiency are fairly close to duncan numbera

I think where I struggle with this is that it's not like RAPM has told us that Duncan looks clearly better than his rivals. Shaq & KG look great by these metrics after all. So while it's fine to say that the RAPM is another feather in Duncan's cap in general, the point where it's being used against rival bigs from other eras for whom we don't have all the same data makes me feel squicky.

I should also say that the fact that teammate Ginobili arguably looks better than Duncan does by these metrics seems like it should be carefully considered. Now, I've been championing Curry, and you can argue something similar about Draymond relative to Curry. So Duncan isn't the only one this applies to, and I'm not saying I'd rank Ginobili's peak ahead of Duncan's (MPG does matter), but given that so much much of the '02-03 argument for Duncan is the "carry job" phenomenon people are fond of talking about nowadays, it's weird that it's so easy to find metrics where it's not exactly clear who is doing the "carrying".

[Doctor MJ]

Looking at the 2012 project it doesn't seem too strange to see things change a bit. There was only one vote without extra ballot options, seemingly no explanation needed and different seasons for one player needing to be tactically switched or be discounted. Lots of close calls that could've been a different outcome with the current voting system. Maybe just a different view on the game as well. Russell was 3rd despite still not being in now, while 09 LeBron made it all the way to 10th despite being in plenty of run-offs. Since it was a new concept people were also still debating the criteria in the voting threads. Not to be disrespectful but the discourse also looks a bit more anecdotally driven than statistically with a lot of news articles quoted and the most in depth stats used being things like SRS and TS%. Besides that, the voting pool seemed to shrink pretty fast compared to a more consistent turnout this edition. Smaller voting pools can also lead to more variance.

All in all plenty of reasons for things to change over time.

Okay, but that's all explanations that don't deal with the actual basketball.

In general, people used to see Olajuwon as better than Duncan, now it's flipped. Assuming this is about something other than differences in the voting base, the question is:

What has caused Duncan to be elevated while Olajuwon has fallen?

This is particularly interesting because we're not talking about drastically different types of players. If someone bumps Duncan over Magic they're almost certainly saying, "Magic is certainly better on offense, but I think Duncan's defense gives him the edge." In the case of Duncan & Olajuwon, we seem to have a situation where people have shifted on both offense and defense in the same direction at a time after the players played their prime. What is causing this change?

Great points all around!

It's cool to see that there's an increased statistical literacy and increased turnout. And like Doctor MJ says, even if Duncan stays ranked around the same number, that doesn't mean everyone else is constant (e.g. LeBron rises, Hakeem falls).

After the project, it would be cool to do a historical comparison. Doctor MJ, I think I remember hearing you had a Google Doc that kept track of every order going years back? Or was that of for the Greatest Player project, instead of the Greatest 1-year Peak project?

To answer your question, I think the thing that swayed me for Duncan over Hakeem was the metrics (I know, big shocker considering my posting history... ). Especially considering how similar they are and how similar their context is, I feel better thinking the metrics are making an apples to apples comparison.

Qualitatively, I have two concerns about Hakeem.
1) I have serious concerns about his lack of passing, and I see Duncan as the clearly better playmaker and facilitator. This lead expands if we consider screening, offensive rebounding, and other forms of offensive (/cultural) leadership. People tend to oversimplify two-way players into Defense and Scoring. When I oversimplify, I tend to divide things into 3 parts: scoring, creation, and defense (so I rank creation higher relative to other posters).
Put another way, I think Hakeem's lack of playmaking and reliance on on-ball scoring gives me major portability concerns vs other players. While Duncan is also not the pinnacle of portability, his portability weaknesses are less concerning, and he also shows a greater mental willingness to play a more auxiliary role and be the team's cultural pillar (compared to Hakeem who had a greater reputation of being a somewhat-stubborn ballhog)
2) I worry people are overrating Hakeem's defense (at least for the peak years that they're arguing). The metrics and my eye test are very favorable on slightly younger Hakeem (e.g. 89, 90, maybe 93). But I worry by the time he finally starts to make ground with Concern 1 (though he never closes the gap), he starts to lose ground to peak Duncan (who peaked on offense and defense at the same time much more effectively than Hakeem).


To be clear, they're both top tier players and there's arguments for both. But during this project, I've struggled to be high on the players who perform well by Eye test / reputation among the analytical crowd, but have clear flaws by the metrics/qualitative analysis (i.e. Hakeem, KG, etc.). I think you've provided some great analysis Doctor MJ, so I'm happy to discuss more if you see flaws in my reasoning.

Ah, I've only ever made a spreadsheet for the Top 100 (career) list, though perhaps I should rectify that in the future if no one else makes one for Peaks. Frankly it's just never seemed as meaningful to me as the Career project. I'm glad we're still trying to do it, but it's considerably harder to get the same amount of thought expressed in Peaks projects compared to Career for reasons that I spoke to early on in the project this time around.

Re: playmaker/facilitator/screening. Fair enough. This has been discussed some and my response isn't much of a rebuttal - it doesn't bother me so much because I'm convinced you can run a great offense with Hakeem in the middle.

Re: offensive rebounding. Really? Why do you say that?

Re: more willing to take an auxiliary role. That's an important thing for career, but I'm expecting anyone voting for Duncan at this stage doesn't think he should have been doing that in his prime.

Re: cultural pillar. Duncan deserves a shout out for this generally, let's just not confuse it for classical leadership. Duncan's main leadership contribution was in letting himself be coached by Pop, and thus setting an example for others. And while Duncan deserves extra credit for how long he had that role, we shouldn't forget that it was David Robinson who did it first, and David Robinson who convinced him not to ditch the Spurs for the Orlando Magic.

Re: Olajuwon ballhog. One thing I'll point out: While Duncan was his college team's leading scorer once he got established, and Robinson allowed Duncan to be the main offensive engine from the jump, Olajuwon was not his team's leading scorer in college, and wasn't his team's leading scorer in the pros until his 2nd year when the team blasted through to get to the Finals.

None of that means I'd call Duncan a ballhog or that I'd insist Olajuwon never behaved selfishly, but it's certainly not the case that Olajuwon came to the NBA insisting "The ball goes through me!". Rather, he showed up on a team where a Duncan-like prospect who had been leading his team in scoring for several years already existed, and Olajuwon shockingly proceeded to quickly surpass him as a scoring threat .

Re: worry about overrating Olajuwon's defense. My guess you feel this way because the Rockets didn't have the regular #1 DRtgs that the Spurs did, combined with the lack of individual data for Olajuwon's era. That's understandable, but I will say this:

When I recently went back through the years and judged DPOY for each year (RS & PS, plus did it for POY, OPOY & COY), I was actually surprised at how often I ended up siding with Olajuwon as a top candidate. Olajuwon earned the #1 spot 4 times, and was on my 3-man ballot 9 times. By contrast for Duncan I had him 2 times in the #1 spot, and on the ballot 5 times.

Now, not only am I fine with others disagreeing, I'll flat out say that when I do this DPOY analysis in deeper history, I tend to end up deferring to the guys who are leading highly ranked defenses more than I think would be ideal. The deeper in history we go, the harder it is to avoid winning bias. But while that's a reason to think I might have been overrating Olajuwon, if I did, I was overrating him for reasons very similar to why Duncan is so revered defensively by many: Team success. Olajuwon led a Top 5 defense 8 times, and that's really quite a lot. Duncan of course was on far more Top 5 defensive teams, but the defense was elite before he got there, he wasn't the best defensive player in the early years, he wasn't the best defensive player in the later years, and the defense was elite after he was gone.

Anyway, I should probably let all this go, but I appreciate your thoughts and your kind words DG!

[falcolombardi]

Well, we seem to look at plus-minus and rapm more now, and those metrics are incredibly kind to duncan

We also may have looked at numbers mpre closely and realize that for how much more easthetical and impressive hakeem moves look, his actual scoring, creation, efficiency are fairly close to duncan numbera

I think where I struggle with this is that it's not like RAPM has told us that Duncan looks clearly better than his rivals. Shaq & KG look great by these metrics after all. So while it's fine to say that the RAPM is another feather in Duncan's cap in general, the point where it's being used against rival bigs from other eras for whom we don't have all the same data makes me feel squicky.

I should also say that the fact that teammate Ginobili arguably looks better than Duncan does by these metrics seems like it should be carefully considered. Now, I've been championing Curry, and you can argue something similar about Draymond relative to Curry. So Duncan isn't the only one this applies to, and I'm not saying I'd rank Ginobili's peak ahead of Duncan's (MPG does matter), but given that so much much of the '02-03 argument for Duncan is the "carry job" phenomenon people are fond of talking about nowadays, it's weird that it's so easy to find metrics where it's not exactly clear who is doing the "carrying".

You dont think is clear who was doing the carrying in san antonio...in 2003? Aka the year in question?

Why do you not apply similar criticisms to curry who actually has the same "issue" in his actual peak seasons discussed (15-17) of a teammate with equal or better impact metrics?

I mean you mention in in passing but also kinda gloss over ut regarding curry but not for duncan?

I hope this doesnt come across as wrong but i dont get your points here



Also literally npbody sans maybe lebron looks ahead of garnett in plus minus metrics

duncan, ginobili, garnett, curry and draymond plus minus metrics (RAPM mainly) hold up to everyone we knpw of othwr than maybe james (who is a bit ahead in his peak rapm seasons last i checked) curry peak rapm metrics are not ahead of ginobiili either for example.

And shaq values are actually a bit below duncan/garnett

[Doctor MJ]

Well, we seem to look at plus-minus and rapm more now, and those metrics are incredibly kind to duncan

We also may have looked at numbers mpre closely and realize that for how much more easthetical and impressive hakeem moves look, his actual scoring, creation, efficiency are fairly close to duncan numbera

I think where I struggle with this is that it's not like RAPM has told us that Duncan looks clearly better than his rivals. Shaq & KG look great by these metrics after all. So while it's fine to say that the RAPM is another feather in Duncan's cap in general, the point where it's being used against rival bigs from other eras for whom we don't have all the same data makes me feel squicky.

I should also say that the fact that teammate Ginobili arguably looks better than Duncan does by these metrics seems like it should be carefully considered. Now, I've been championing Curry, and you can argue something similar about Draymond relative to Curry. So Duncan isn't the only one this applies to, and I'm not saying I'd rank Ginobili's peak ahead of Duncan's (MPG does matter), but given that so much much of the '02-03 argument for Duncan is the "carry job" phenomenon people are fond of talking about nowadays, it's weird that it's so easy to find metrics where it's not exactly clear who is doing the "carrying".

You dont think is clear who was doing the carrying in san antonio...in 2003? Aka the year in question?

Why do you not apply similar criticisms to curry who actually has the same "issue" in his actual peak seasons discussed (15-17) of a teammate with equal or better impact metrics?

I mean you mention in in passing but also kinda gloss over ut regarding curry but not for duncan?

I hope this doesnt come across as wrong but i dont get your points here



Also literally npbody sans maybe lebron looks ahead of garnett in plus minus metrics

duncan, ginobili, garnett, curry and draymond plus minus metrics (RAPM mainly) hold up to everyone we knpw of othwr than maybe james (who is a bit ahead in his peak rapm seasons last i checked) curry peak rapm metrics are not ahead of ginobiili either for example.

And shaq values are actually a bit below duncan/garnett

I think in general the term "carry job" tends to imply that the team is falling apart without you and you specifically. Let's consider that the term is basically used as a way of saying "Yeah that team wasn't as impressive as some, but have you seen how little he had to work with?", and it's weird when you look at things and there's another guy on the roster that seems to be making even more of a stark difference.

Re: why not apply similar criticisms to Curry. Well, I mean, I kinda did in the post you just responded to, but clearly that's not something that feels satisfying, so:

In both cases, you've got a teammate (Ginobili, Green) who by some angles looks more impactful than the alpha (Duncan, Curry).
In both cases, after having thought seriously about it, I side with the alpha as generally the one who deserves the most credit.
In both cases I respect both the teammates in high esteem.
The Golden State teammates have the advantage in that they led more dominant teams.
The Golden State teammates have the advantage in that they played in a more sophisticated era.
The San Antonio teammates have the career advantage in that we've seen the entirety of their career.

But look, the reason why I hammer home the point about Ginobili when we talk about Duncan has everything to do with the way Duncan's volume scoring gets venerated - and probably has everything to do with why people like that term "carry job" - and I don't think it actually functioned all that well. It was good enough to win with a great defense, but it sucked energy away from defense while also removing opportunity for more perimeter-oriented attack.

This by contrast to a situation like Olajuwon's where I don't think anyone thinks there was someone else on the roster who was a better offensive player, and who won with a more dominant playoff run than Duncan's prime teams ever did.

Re: Curry's not ahead of Ginobili either. Aside from the fact that I've already talked about the MPG being a thing that holds Ginobili back some, what I want to emphasize here is that Ginobili was actually on Duncan's team and blows up the "carry job" narrative.

Curry, by contrast, isn't really in the "carry job" category. He's acknowledge to have good teammates, and it's emphasized that he's been the most valuable player on teams that were all-time great, and thus "What about Green?" doesn't really resonate the same way. If you personally think Green was more valuable than Curry, by all means, speak to that, but as far as Curry's teammates being great, well yeah, that's what you need in order to get a team to the absolute pinnacle of what we've seen able to be accomplished in basketball.

Re: Shaq below Duncan/Garnett on RAPM. Hmm, I'd appreciate if you share the single year metrics you're thinking of here. My thinking was shaped when we first got access to this data when it sure looked like Shaq & KG were the kings of the metric (to be joined by LeBron eventually).

RAPMs I've seen since tend to make the gap between Garnett & Duncan more murky, but I wasn't under the impression they put Shaq clearly below those two. If they do, that's something. Frustrating the way we aren't able to standardize the stat.

[falcolombardi]

I think where I struggle with this is that it's not like RAPM has told us that Duncan looks clearly better than his rivals. Shaq & KG look great by these metrics after all. So while it's fine to say that the RAPM is another feather in Duncan's cap in general, the point where it's being used against rival bigs from other eras for whom we don't have all the same data makes me feel squicky.

I should also say that the fact that teammate Ginobili arguably looks better than Duncan does by these metrics seems like it should be carefully considered. Now, I've been championing Curry, and you can argue something similar about Draymond relative to Curry. So Duncan isn't the only one this applies to, and I'm not saying I'd rank Ginobili's peak ahead of Duncan's (MPG does matter), but given that so much much of the '02-03 argument for Duncan is the "carry job" phenomenon people are fond of talking about nowadays, it's weird that it's so easy to find metrics where it's not exactly clear who is doing the "carrying".

You dont think is clear who was doing the carrying in san antonio...in 2003? Aka the year in question?

Why do you not apply similar criticisms to curry who actually has the same "issue" in his actual peak seasons discussed (15-17) of a teammate with equal or better impact metrics?

I mean you mention in in passing but also kinda gloss over ut regarding curry but not for duncan?

I hope this doesnt come across as wrong but i dont get your points here



Also literally npbody sans maybe lebron looks ahead of garnett in plus minus metrics

duncan, ginobili, garnett, curry and draymond plus minus metrics (RAPM mainly) hold up to everyone we knpw of othwr than maybe james (who is a bit ahead in his peak rapm seasons last i checked) curry peak rapm metrics are not ahead of ginobiili either for example.

And shaq values are actually a bit below duncan/garnett

I think in general the term "carry job" tends to imply that the team is falling apart without you and you specifically. Let's consider that the term is basically used as a way of saying "Yeah that team wasn't as impressive as some, but have you seen how little he had to work with?", and it's weird when you look at things and there's another guy on the roster that seems to be making even more of a stark difference.

Re: why not apply similar criticisms to Curry. Well, I mean, I kinda did in the post you just responded to, but clearly that's not something that feels satisfying, so:

In both cases, you've got a teammate (Ginobili, Green) who by some angles looks more impactful than the alpha (Duncan, Curry).
In both cases, after having thought seriously about it, I side with the alpha as generally the one who deserves the most credit.
In both cases I respect both the teammates in high esteem.
The Golden State teammates have the advantage in that they led more dominant teams.
The Golden State teammates have the advantage in that they played in a more sophisticated era.
The San Antonio teammates have the career advantage in that we've seen the entirety of their career.

But look, the reason why I hammer home the point about Ginobili when we talk about Duncan has everything to do with the way Duncan's volume scoring gets venerated - and probably has everything to do with why people like that term "carry job" - and I don't think it actually functioned all that well. It was good enough to win with a great defense, but it sucked energy away from defense while also removing opportunity for more perimeter-oriented attack.

This by contrast to a situation like Olajuwon's where I don't think anyone thinks there was someone else on the roster who was a better offensive player, and who won with a more dominant playoff run than Duncan's prime teams ever did.

Re: Curry's not ahead of Ginobili either. Aside from the fact that I've already talked about the MPG being a thing that holds Ginobili back some, what I want to emphasize here is that Ginobili was actually on Duncan's team and blows up the "carry job" narrative.

Curry, by contrast, isn't really in the "carry job" category. He's acknowledge to have good teammates, and it's emphasized that he's been the most valuable player on teams that were all-time great, and thus "What about Green?" doesn't really resonate the same way. If you personally think Green was more valuable than Curry, by all means, speak to that, but as far as Curry's teammates being great, well yeah, that's what you need in order to get a team to the absolute pinnacle of what we've seen able to be accomplished in basketball.

Re: Shaq below Duncan/Garnett on RAPM. Hmm, I'd appreciate if you share the single year metrics you're thinking of here. My thinking was shaped when we first got access to this data when it sure looked like Shaq & KG were the kings of the metric (to be joined by LeBron eventually).

RAPMs I've seen since tend to make the gap between Garnett & Duncan more murky, but I wasn't under the impression they put Shaq clearly below those two. If they do, that's something. Frustrating the way we aren't able to standardize the stat.

Do you think ginobili was "that guy" having that kind of impact (or the abikity to have that impact) in 2003 already?

It feels like you are conflating the great help from ginobili duncan got in 2005 and forward with what he had in 2003 and is specially odd to me you compare it with curry teams

At their best warriors gave curry (in his consensus peak season) the roughly impact equivalent of peak kawhi (durant), peak ginobili (green), peak bowen (older iggy) and in a loose way, peak parker (klay). Duncan never got to overlap with more than two of those guys (at their prime) at a time, and not even that in 2003

His 2003 supporting cast is closer to 2014 warriors than 2017 warriors with how much green and thompson (parker and ginobili for this analogy) were still getting their legs on

The 2003 spurs are not even close to being the same supporting cast than the 2017 warriors

As for shaq vs duncan in rapm

I meant best 5 year stretches

https://www.thespax.com/nba/quantifying-the-nbas-greatest-five-year-peaks-since-1997/

Shaq is a tad below garnett and duncan in the leaderboard

If you want to take single year rapm fwiw, here is it too

https://basketball-analytics.gitlab.io/rapm-data/season/1999-00/regular-season/

Shaq is at 2.6 and barelt in the top 30 in 2000 regular season, duncan is top 10 in 2003 at 3.6

2003 Duncan at 2.3 (2nd to ginobili) vs 2000 shaq at 1.0 (14th, virtually tied with teammate horry) fpr playoffs

Duncan looks better than shaq in rapm and has arguably better boxscore numbers in the head to head matchups before even considering defensive gap

Why should we be so confident shaq is clearly better than duncan?

Him being bigger in the basketball media and community amd compared more with jordan (somethingh you mention in another post) is very obvious when he has a much more loud/charismatix personality and played in the much bigger market

Kobe got more comparisions to peak jordan as a player than duncan too and i wouldnt use that for a kobe>duncan argument

[trex_8063]

Well, we seem to look at plus-minus and rapm more now, and those metrics are incredibly kind to duncan

We also may have looked at numbers mpre closely and realize that for how much more easthetical and impressive hakeem moves look, his actual scoring, creation, efficiency are fairly close to duncan numbera

I think where I struggle with this is that it's not like RAPM has told us that Duncan looks clearly better than his rivals. Shaq & KG look great by these metrics after all. So while it's fine to say that the RAPM is another feather in Duncan's cap in general, the point where it's being used against rival bigs from other eras for whom we don't have all the same data makes me feel squicky.

I should also say that the fact that teammate Ginobili arguably looks better than Duncan does by these metrics seems like it should be carefully considered. Now, I've been championing Curry, and you can argue something similar about Draymond relative to Curry. So Duncan isn't the only one this applies to, and I'm not saying I'd rank Ginobili's peak ahead of Duncan's (MPG does matter), but given that so much much of the '02-03 argument for Duncan is the "carry job" phenomenon people are fond of talking about nowadays, it's weird that it's so easy to find metrics where it's not exactly clear who is doing the "carrying".

I don't really get the idea of bringing up Manu in this capacity wrt '03. In '03 Manu was a very promising but still.....unpredictable rookie. His PI RAPM was +0.8 (somewhat good, but not "carrying" anything), and that was in just 20.7 mpg in the rs and 27.5 mpg in the playoffs.
Overall....

'03 Spurs in rs (team-leading figure is bolded for each metric [RAPM is a rs/ps combined metric])
Duncan: 26.9 PER, .248 WS/48, +7.6 BPM, +18 net rating, +4.4 PI RAPM, +6.1 NPI RAPM, 3181 minutes
Parker: 16.5 PER, .134 WS/48, +0.5 BPM, +4 net rating, +2.3 PI RAPM, +2.0 NPI RAPM, 2774 minutes
Bowen: 9.1 PER, .096 WS/48, +0.2 BPM, +7 net rating, +0.9 PI RAPM, +2.6 NPI RAPM, 2566 minutes
Jackson: 14.5 PER, .102 WS/48, +0.9 BPM, +/- 0 net rating, -0.5 PI RAPM, -0.9 NPI RAPM, 2254 minutes
Rose: 16.1 PER, .130 WS/48, -0.6 BPM, +4 net rating, +/- 0 PI RAPM, +0.2 NPI RAPM, 1933 minutes
Robinson: 17.8 PER, .172 WS/48, +2.1 BPM, +14 net rating, +3.3 PI RAPM, +3.1 NPI RAPM, 1676 minutes
Ginobili: 14.7 PER, .141 WS/48, +1.9 BPM, +8 net rating, +0.8 PI RAPM, +4.6 NPI RAPM, 1431 minutes
Smith: 12.0 PER, .117 WS/48, -0.4 BPM, +6 net rating, -0.5 PI RAPM, -3.0 NPI RAPM, 1032 minutes
Kerr: 11.4 PER, .109 WS/48, +0.1 BPM, +6 net rating, -0.2 PI RAPM, -2.5 NPI RAPM, 952 minutes
Willis: 12.5 PER, .091 WS/48, -3.7 BPM, -1 net rating, -1.0 PI RAPM, -3.3 NPI RAPM, 840 minutes
Ferry: 5.1 PER, .028 WS/48, -3.2 BPM, -14 net rating, -0.9 PI RAPM, -2.4 NPI RAPM, 601 minutes
Claxton: 15.9 PER, .125 WS/48, +1.2 BPM, +4 net rating, +1.9 PI RAPM, +1.2 NPI RAPM, 471 minutes

'03 Spurs in playoffs
Duncan: 28.4 PER, *.279 WS/48, *+10.2 BPM, *+24 net rating, +4.4 PI RAPM, +6.1 NPI RAPM, 1021 minutes (*except for Steve Kerr and his 46 total minutes)
Parker: 11.9 PER, .063 WS/48, -2.4 BPM, -6 net rating, +2.3 PI RAPM, +2.0 NPI RAPM, 814 minutes
Jackson: 12.0 PER, .077 WS/48, +0.6 BPM, -4 net rating, -0.5 PI RAPM, -0.9 NPI RAPM, 811 minutes
Bowen: 8.9 PER, .104 WS/48, +1.1 BPM, +6 net rating, +0.9 PI RAPM, +2.6 NPI RAPM, 750 minutes
Ginobili: 15.0 PER, .152 WS/48, +3.7 BPM, +10 net rating, +0.8 PI RAPM, +4.6 NPI RAPM, 660 minutes
Rose: 13.3 PER, .107 WS/48, -1.8 BPM, +1 net rating, +/- 0 PI RAPM, +0.2 NPI RAPM, 560 minutes
Robinson: 17.7 PER, .204 WS/48, +3.6 BPM, +22 net rating, +3.3 PI RAPM, +3.1 NPI RAPM, 539 minutes
Claxton: 14.4 PER, .117 WS/48, +0.9 BPM, +3 net rating, +1.9 PI RAPM, +1.2 NPI RAPM, 326 minutes
Ferry: 2.7 PER, .009 WS/48, -5.7 BPM, -20 net rating, -0.9 PI RAPM, -2.4 NPI RAPM, 101 minutes
Willis: 18.7 PER, .141 WS/48, -2.2 BPM, +5 net rating, -1.0 PI RAPM, -3.3 NPI RAPM, 91 minutes
Smith: 3.5 PER, -0.042 WS/48, -5.5 BPM, -23 net rating, -0.5 PI RAPM, -3.0 NPI RAPM, 66 minutes
Kerr: 22.7 PER, *.331 WS/48, *+11.4 BPM, *+62 net rating, -0.2 PI RAPM, -2.5 NPI RAPM, 46 minutes


Steve Kerr's totally absurd, non-sustainable, and totally out-of-character metrics in a scant 46 total minutes of playoff playing-time not withstanding.....Duncan otherwise leads the team in every single metric [often by >1 standard deviations over next-best]----in both rs and playoffs----while also leading in playing time in both rs/ps [by substantial margins].

And on top of that he's credited by the coach, by the GM, and by multiple players as being the culture leader of this team.

If someone is "carrying" this team, is there any doubt as to who that someone is?

Even by impact metrics ONLY, this doesn't rate out as an exceptional supporting cast for a title-winning team. Perhaps not "bad" or "weak", but not notable either. Again: that's by impact metrics ONLY. Add box-derived metrics to the evaluation, and they begin to look weak for a title winning team (especially in the playoffs).


And to be clear, do you object to the use of "carry job" in general? Or just to the use of it for '03 Duncan?
Because few players have benefited from the "carry job" narrative more than Hakeem Olajuwon (and I'm not sure that '94 cast was notably worse than Duncan's '03 cast, particularly once playoffs are weighed in).

[Dutchball97]

Sure, but in 2012 we had Duncan at 9 (while Hakeem was at 5). We've seen the two of them flip well after either of them were playing peak basketball.

I'm not saying that's wrong, but it's interesting, and worth understanding how people are reaching new and different conclusions.

Looking at the 2012 project it doesn't seem too strange to see things change a bit. There was only one vote without extra ballot options, seemingly no explanation needed and different seasons for one player needing to be tactically switched or be discounted. Lots of close calls that could've been a different outcome with the current voting system. Maybe just a different view on the game as well. Russell was 3rd despite still not being in now, while 09 LeBron made it all the way to 10th despite being in plenty of run-offs. Since it was a new concept people were also still debating the criteria in the voting threads. Not to be disrespectful but the discourse also looks a bit more anecdotally driven than statistically with a lot of news articles quoted and the most in depth stats used being things like SRS and TS%. Besides that, the voting pool seemed to shrink pretty fast compared to a more consistent turnout this edition. Smaller voting pools can also lead to more variance.

All in all plenty of reasons for things to change over time.

Okay, but that's all explanations that don't deal with the actual basketball.

In general, people used to see Olajuwon as better than Duncan, now it's flipped. Assuming this is about something other than differences in the voting base, the question is:

What has caused Duncan to be elevated while Olajuwon has fallen?

This is particularly interesting because we're not talking about drastically different types of players. If someone bumps Duncan over Magic they're almost certainly saying, "Magic is certainly better on offense, but I think Duncan's defense gives him the edge." In the case of Duncan & Olajuwon, we seem to have a situation where people have shifted on both offense and defense in the same direction at a time after the players played their prime. What is causing this change?

Fair enough. Honestly it seems to me like it is mainly a matter of changing perspectives on certain things over time. Hakeem sometimes gets lost in the shuffle among all-time greats but he also regularly got mythologized as "the one man Jordan feared to go up against". Especially his offense probably got a bit too much credit due to the dreamshake and how aesthetically pleasing his scoring repertoire was in general, much like how there are still people boosting Kobe up their lists for the sole reason of a varied scoring game that looks good. However, I don't think it's a crazy take the "boring" Duncan was actually as good, if not better, on offense as Hakeem at their peaks.

Another aspect is the 03 Spurs. On paper they're loaded with Robinson still there and Manu and Parker also joining the fray but none of them were in their prime. I think people became more aware of just how important Duncan was for the succes the team had that year.

In general though I've had 94 Hakeem and 03 Duncan right next to each other on my ballots for a couple rounds now as I have a hard time seeing much seperation between them. They're both all-time great defenders with strong offensive games who won MVP and carried solid, well rounded teams without another All-Star to hard fought titles. I can't really justify either being like 5 spots higher or lower. I'm pretty much of the opinion they simply didn't get it "right" the first time.

[ty 4191]

Fascinating to see the way Duncan's stature has grown over time.

What do you mean by this?