RealGM Top 100 List #5

[shutupandjam]

I feel like there may be at least some confusion about the different forms of RAPM in general, and this may be important with respect to how we're using it as an evaluation tool. Does everyone understand the difference between npi rapm, pi rapm, and xrapm? Does anyone have specific questions about the metric(s)? I feel like we're throwing "rapm" out there too much without considering its context.

For example, I wouldn't say "x is better than y in 200X prior informed rapm, therefore he had a better year" because prior informed rapm uses information from previous seasons. But the prior informed version might be more valuable than non prior informed when discussing a general time period (not just a single season).

I have mentioned this before, but I really think when we're talking about all-in-one metrics we should use SPM iterations more often. I recommend DSmok's ASPM, mystic's spm, and my own estimated impact. (not trying to promote myself here, but it's easy access and updated)...side note: I am in the process of renovating it a bit, but things won't change too much when I do

The reason I like SPM so much is because it tests out of sample better than anything other than blended rapm/spm stuff (like IPV or xRAPM/RPM) and they are stable and imo more reasonable than ws, per, etc. Their advantage over using a blended metric is that we can actually compare box score and plus minus stuff against each other rather than having them sewn together. Of course, you're still gonna get absurd results once in a while but obviously no metric is a perfect be all end all type of player evaluator.

[HeartBreakKid]

Seeing the two play head to head, I've always got the impression that Shaq was the more dominant player. I find comparing players who played at the same time to be very easy. Shaq would have his struggles against the Spurs, but getting doubled by defensive goats in Robinson and Duncan should do that to you, gave plenty of room for Kobe to feast though.


Peak play matters a lot more than longevity, in which Duncan has been very good the past few years, but not a true star player - so it holds little weight in a comparison of all time greats. Shaq at his best really did feel close to unstoppable, Duncan to me was always a very close #2.

Big Timmy certainly kills Shaq when it comes to defense, Tim has helped anchor some legendary defenses, though I do think the gap in their offense is quite sizeable as well. Shaq could score on much higher volume with much better efficiency. So what's more important, the offensive gap or the defensive gap?

I'm going to have to go with the offensive gap, aside from the volume scoring and efficiency advantage, the biggest thing that I thought made Shaq seem "unfair" was his ability to put an insane amount of fouls on the other team. Whether this was guys fouling him on purpose to stop him from getting an easy dunk, Shaq initiating the contact, or them just sending him to the line to miss free throws, the amount of fouls he drew showed an immense amount of impact, it is so taxing for a roster to foul the amount of times most teams had to against Shaq. Not to mention many teams starting to hire non skilled bigs, to pretty much be goons and bang with Shaq, which is impact in itself.

My vote goes to Shaquille O'Neal.

[colts18]

I mentioned before that Shaq faced some of the stiffest competition for a superstar ever. He played a lot of really tough defenses. His opponent D rating was the toughest for a superstar of the modern era.

Some of the elite defenses Shaq faced:

94 Pacers 104.2 D rating (8th)
95 Bulls 104.3 D rating (2nd)
95 Pacers 105.6 D rating (6th)
96 Pistons 105.1 D rating (7th)
96 Bulls 101.8 D rating (1st)
97 Blazers 103.3 D rating (7th)
97 Jazz 104.0 D rating (9th)
98 Blazers 102.4 D rating (8th)
98 Sonics 103.6 D rating (10th)
99 Rockets 102.9 D rating (15th)
99 Spurs 95 D rating (1st)
00 Kings 102.1 D rating (10th)
00 Suns 99 D rating (3rd)
00 Blazers 100.8 D rating (5th)
00 Pacers 103.6 D rating (13th)
01 Blazers 101.8 D rating (9th)
01 Kings 99.6 D rating (7th)
01 Spurs 98 D rating (1st)
01 Sixers 98.9 D rating (5th)
02 Blazers 104 D rating (13th)
02 Spurs 99.2 D rating (2nd)
02 Kings 101.1 D rating (6th)
02 Nets 99.5 D rating (1st)
03 Twolves 103.8 D rating (16th)
03 Spurs 99.7 D rating (3rd)
04 Rockets 99 D rating (5th)
04 Spurs 94.1 D rating (1st)
04 Twolves 99.7 D rating (6th)
04 Pistons 95.4 D rating (2nd)
05 Nets 103.1 D rating (7th)
05 Pistons 101.2 D rating (3rd)
06 Bulls 103.4 D rating (7th)
06 Nets 102.4 D rating (4th)
06 Pistons 103.1 D rating (5th)
07 Bulls 99.6 D rating (1st)

Shaq played 5 elite Duncan-led Spurs defenses including the 04 Spurs who have the record for D rating relative to league average (-8.8). He also played the 04 Pistons who have an argument as the GOAT defense. He played the 72 win 96 Bulls #1 defense. He faced off against 4 straight top 5 Ben Wallace defenses. Played against KG's toughest Minnesota defense. Played against the 4 time DPOY Dikembe Mutombo. He destroyed the #1 defense 2002 Nets. No one has ever faced such a tough gauntlet which is why Shaq's numbers are so amazing in the playoffs. He put up all time numbers and it vs such tough competition.

[Clyde Frazier]

I was planning to do a more in depth comparison on magic vs. duncan vs. shaq to bolster my case for magic, but I ran out of time. I'll post a revised version of it in the next thread. I'd still like to get a runoff vote in, though.

Runoff vote - Duncan

http://bkref.com/tiny/7w0Tb

This comes off as one of the hardest choices so far (and maybe in the entire top 100). The accolades have already been detailed, so I'll just say that they're essentially identical with duncan ahead by 1 ring and 1 regular season MVP. He also has 10 more all defensive selections, but I wouldn't necessarily use that as a huge advantage over shaq, as he was one of the most dominant offensive forces in NBA history.

I certainly value prime, peak and longevity, but I'd say how I weigh them depends on the player. In this case, you have shaq's peak with a relatively clear advantage to duncan's comparable prime and (somewhat) better longevity. Shaq had better production than given credit for towards the end of his career, but I don't think he was contributing to team success as strongly as duncan has the last few seasons.

Duncan certainly came into an ideal situation with teammates, coach, and front office, but he played a large role in contributing to that culture over the course of his career. Recency bias really frustrates me, but I think you can avoid that and still appreciate what he was able to accomplish this past season. It's hard to fathom the kind of focus it took from him and the rest of the spurs to actually pull it off.

Shaq's 3 peat in LA was one of the most impressive runs we've ever seen from 1 player. I firmly believe he and kobe could've won at least 1 more ring together (possibly 2) if they put their differences aside. Also, I'll come right out and say that I take issue with his ring in 06, and my opinion on that will never change. It has nothing to do with him, but the heat had no business winning that championship.

Yeah, I know this is a bit of a ramble, and there's clear arguments for both players. I think when it's all said and done, duncan had the better career, and in this case where it's so close, that's the deciding factor for me.

[ardee]

The only Twin towers situations we have seen are Duncan/Robinson and Hakeem/Sampson

That isn't true. There have been plenty of good 7 fters playing together

Howard/Asik: Asik traded after a year because they could never stay on the court
Howard/Gasol: Significant issues during the one season together
Gasol/Bynum: Played well together for 4 years
Shaq/Campbell: I don't remember how they played together
Wilt/Thurmond: Also unsure of how this worked out.

Wilt/Thurmond played roughly 20 or so mpg together... And it led to the best non-Celtics defense of the 60s in 1964.

[O_6]

Spoiler:

I've already made my vote for Shaq and talked about why I favor him over Hakeem at this position, but I didn't really go in-depth talking about Duncan. And since this seems to be heading towards a Duncan vs. Shaq run-off I felt the need to post. I certainly see Duncan's argument, he has one of the most complete and flawless cases of anyone in NBA history.

Duncan's longevity certainly makes him special, but in his prime I don't think he was nearly as historically unique or irreplaceable as Shaq. And Shaq was a valuable player from '93-'09 (17 years), so I don't think longevity is enough of an edge for Duncan since Duncan only has 17 years in this league.

Duncan: We've seen his species before, he might not even be the best of his kind
Hakeem
Ewing
Robinson
Mourning
Garnett

All of these post-3pt line era bigs had a very comparable impact and style to Duncan. Historically dominant DPOY-caliber Bigs who also doubled as 20 PPG offensive MVPs for their teams.

People might think Ewing and Mourning don't belong, but both of those players led 60 win Championship contenders when they were paired up with Pat Riley. It wasn't like those teams were absolutely loaded either, they were dominant defensively and led by a great coach (Spurs-like). Hakeem's game blossomed under Rudy T. and he led his teams to 2 straight titles at the ages of 31 and 32. At the age of 31, Garnett is finally on a team with real talent + smart coaching and he leads the '08 Celtics to a 66 win title season on the back of an all-time great defense. David Robinson was routinely winning 55+ games with poor talent around him and he finally won a ring once he was able to be paired up with Duncan and Pop.

Look, Duncan very well may be the best out of all of these players. But I think it's unquestionable that he enjoyed the best situation and environment of all of those guys. With the Spurs he was surrounded by a great team with a great coach day one.

We have people on this thread arguing for Prime Hakeem and Prime Garnett being better than Prime Duncan. We've had countless "what happens if Robinson and Duncan switch timeline" threads over the years. We've had "where does Ewing rank if he has a better 2nd scorer than John Starks in his prime" threads before.

I know a lot of what I wrote above is based on complete hypotheticals and impossible to say for sure. I get that, but the point of it all is to show how great of a situation Duncan has had with the Spurs compared with similar types of players as him.

Shaq: Historically Unique Style
Since '80, there have been a handful of Bigs who impacted the game in a similar style as Duncan
Since '80, there have been ZERO Bigs who impacted the game in a similar style as Shaq

A player who had an OPOY-caliber impact almost entirely from the low block. A player who was not a consistent DPOY-caliber defender but could flash DPOY ability during stretches and was generally a well-above average defensive anchor.

The only two Bigs in NBA history that were similar to Shaq were Wilt and Kareem. But Wilt played his entire career pre-3pt line and Kareem played 90% of his prime pre-3pt line. The 3pt line changed the game and made it more guard dependent. And even still, you could argue that Wilt was never as consistently dominant on offense as Shaq despite the giant volume numbers. So in reality you are talking about Kareem as the only plus defensive rim-protector who also doubled as a consistent OPOY candidate.

There is value in this uniqueness imo. It made Shaq much harder to gameplan for, he caused more coaching adjustments than any player since 1980. Since he couldn't be stopped one-on-one, he had to be guarded by entire teams. No one was doubled more in modern history.

I feel like Shaq's greatness and impact had less to do with his environment and situation than Duncan's. To me that is a valuable edge when everything else is so close.

Shaq vs. Duncan in the RAPM age
I wanted to compare Prime Shaq with Prime Duncan. I'm going to compare their 7 best years. Why 7 years? Because Shaq only has 7 prime years in the RAPM era (1998-00, 02-05). By 2006, his production had taken a drop and I think we all know that atleast one of his pre-98 seasons or his 2001 season would've absolutely been better than his '06 in terms of RAPM. '06 wasn't his true prime despite the ring he won playing 2nd fiddle to Wade, so I didn't feel like using it in this study because it wasn't his real prime and it would unfairly bring down his averages.

*USING DOC MJ's SPREADSHEET
Shaq vs. Duncan: Top 7 years RAPM --- O and D values (since 1998... no 2001)
Player ---- Avg. RAPM --- Avg. ORAPM ---- Avg. DRAPM ---- *%GP
Shaq -------- 9.81 ------------- 6.98 ------------- 2.83 --------- 85.2%
Duncan ----- 8.88 ------------- 3.50 ------------- 5.38 --------- 93.4%
*%GP = Player GP / Team GP....

Both guys come out looking extremely good in this measure, backing up their All-Time Great reputations with All-Time Great RAPM results. Shaq has the major edge offensively and Duncan has the major edge defensively. Shaq has the edge in value added per minute, Duncan has the edge in minutes and games played. I think all of that is what we expected. But I feel like Shaq's edges are more impressive.

Shaq's average ORAPM over his 7 best years comes out to 6.98
Duncan's average DRAPM over his 7 best years comes out to 5.38

- 8 Offensive Players have topped 6.98 ORAPM atleast twice (2 bigs... Dirk/Barkley)
- 10 Defensive Players have topped 5.38 DRAPM atleast twice (1 wing... Jaren Jackson lolwut?)

Duncan's defensive dominance is quite impressive but there have been several Bigs (including those mentioned by me earlier in his comps section) who have had multiple defensive seasons that would rate as Prime Duncan-caliber. The only non-Big who scored a super-high DRAPM was Jaren Jackson, a 6'4" guard who played next to Duncan/Robinson during the '98-'99 seasons. I have been thoroughly impressed with the RAPM stat now that I've looked into it, but clearly it isn't a foolproof stat and the inclusion of Jaren Jackson is proof. But the point is that we've seen several great defensive bigs on Duncan's level as a defender post-80/3pt line.

The only comparable offensive bigs to Shaq since '80 according to RAPM are Dirk/Barkley and maybe Karl Malone. That's not surprising to hear, it's about what we expected. But Shaq is ahead of Dirk and probably ahead of Malone according to RAPM. Barkley is probably the only guy who may or may not truly be considered a better pure offensive big than Shaq. But Barkley was also an atrocious defender and that's the main reason why his name will likely not be called in the Top 15 of this project. Shaq still comes out as a very positive defender, with a +2.83 grade in his prime that's not that far off from Tim Duncan's +3.50 offensive grade. In terms of per minute production, Shaq's RAPM numbers are simply more impressive.

So Timmy's main edges according to this are his teams Games Played% and DRAPM. He played in roughly 93% of his teams regular season games while Shaq played in only 85% of his teams regular season games during these two 7 year prime runs. That comes out to 6-7 games played over the course of a season. That's nothing to sneeze at. Those games could matter when it comes to seeding. Shaq was also a good but far less valuable defender during this stretch.

But the problem with Duncan is that his games played edge completely disappears in the playoffs. Shaq played 110/112 playoff games during this 7 year run, with the only two games he missed coming in his final season of this run in 2005 with the Heat. So while those 6 regular season games per year matter, the "Durability Value" of Duncan is limited because Shaq was just as durable in the playoffs when the games were significantly more important.

And according to both Shaq's critics and Shaq's supporters, Shaq was a player who stepped up on defense when he tried. Guess what, he'd go hard on defense more often in the playoffs. I believe that Shaq's playoff DRAPM grade would clearly be higher than his +2.83 DRAPM regular season grade, simply based on what we all believe about Shaq's effort-level increase in important games.

So out of Duncan's 2 main edges over Shaq, durability and defense, durability was definitely NOT an edge in the playoffs and his defense was probably less of an edge in the playoffs due to Shaq's effort level picking up. While the durability thing is 100% true, the defensive edge in the playoffs is more of a question mark. But I do think Shaq played better on D in the playoffs than he did in the regular season.

Tl:dr
Duncan may be more consistent, may be more complete, may be more dependable, may have maximized his legacy more... still doesn't make up for the fact that Shaq was a more dominant force with a singular skillset that could not be duplicated by another player in modern NBA history. Shaq's regular season RAPM numbers in his prime give him an edge over Duncan, despite the fact that we don't even have some of Shaq's best prime seasons ('94-97, '01... especially '01) and have all of Duncan's best prime seasons outside of '01. And those are regular season RAPM numbers. If what Shaq's critics say about his defense are true, chances are that his RAPM in the playoffs would actually increase and his lead over Duncan would grow.

Duncan is an amazing player but Shaq is more worthy of this spot.

Spoiler:

This was an excellent, well thought-out and well-supported post. I even gave it an "And1". The problem is...with me, it may have accomplished the exact opposite of what you intended.

Coming into the run-off I was unsure which way I would vote, but I was leaning Shaq. Through the years I always wondered whether Shaq was REALLY more dominant in impact than Duncan with both at their best, or if that was just the perception from the eye-test and the scoring feats while Duncan's defense and fundamental brilliance was just too subtle for acclaim. Over the years I have (at one time or another) raised just about every point mentioned on both sides of this debate in the thread, and was unable to really decide.

Then, in recent years, the +/- data started trickling in. Finally, there was some way to quantify some bit of a player's individual impact on the team's bottom line. The first +/- numbers to come out (that I knew of) were the on/off +/- on 82games.com. The problem was, their first available season was 2003 and they didn't have the playoffs until after that. By then it was slightly past Shaq's prime. KG was the dominant player of the early years of 82games.com, with Duncan a clear second. But Shaq's peak wasn't covered. Next, we started hearing about adjusted +/- ( APM) that cleaned up some of the issues with on/off +/-. Again, started with 2003, and the leaders were KG and Duncan. Then came multi-year APM starting from 2003...and it was still KG and Duncan. But by this time this young guy named LeBron was making his name, and seemed poised to take the mega impact baton from KG for the next generation. But still, we had nothing on Shaq. APM evolved into RAPM and Dirk also got his name near the top of the leaderboard, but the years stayed the same which meant no peak Shaq. Until recently.

Now, with prior-informed RAPM back to 1998 and playoffs on/off +/- back to 2001, we finally have some +/- information from Shaq's prime. Not his complete career, but at least from just before the time usually considered his peak. And the results were outstanding. In DocMJ's now oft-referenced spreadsheet, it shows a pretty convincing separation between the three players with the highest RAPM scores and even the next level down. Shaq, along with LeBron and KG, are in that top tier of scores. Duncan, along with Dirk, is in the next group down. It's not a huge difference, but it is convincing to me even knowing the questions about the accuracy of the stat in a given year. Part of the reason that it's convincing is that it keeps happening year after year.

Shaq's highest single season score is 1.4 higher than Duncan's highest
Shaq's 2nd highest score is 1.5 higher than Duncan's 2nd highest
Shaq's 3rd highest score is 0.5 higher than Duncan's 3rd highest
Shaq's 4th highest is 1.0 higher than Duncan's 4th highest
Shaq's 5th highest is 0.7 higher than Duncan's 5th highest

And this isn't even including 2001 (where Shaq would likely have another peak score) or the years pre-1998 where Shaq may also have scored very well. The reason that experimenters do repeated studies is to lower noise effects, and the fact that Shaq beats Duncan (by a small but clear margin) on a year-to-year basis is enough to finally help put to rest my question of whether Duncan was quietly matching peak Shaq's impact. Answer: he wasn't. In the face of the excellent reviews and cases we've read here for both players, this result was enough that, as I said, I was leaning towards voting Shaq in the run-off.

BUT.

In your post you emphasize the uniqueness of Shaq's skill set, and the fact that his primary area of influence is offensive. We also have a well documented history of Shaq preferring to play his preferred method...and for good reason, I might add, because it was so effective.

BUT.

One thing that I haven't read a lot about is Duncan's advantage in portability. And as I was reading your post, that kept coming more and more to the forefront of my thoughts. And I don't think it's a small concern.

I've argued before that defense is more portable than offense, and I believe that. Which means that just on general skillset, Duncan would be easier to integrate onto more teams than Shaq would. But it goes further than that.

Duncan also has position versatility. We saw with the Duncan/Robinson frontcourt that Duncan could interact with another outstanding center with a lot of skill overlap. Now, part of the reason that worked is because Admiral is one of the most unselfish superstars in NBA history and was willing to cede leadership to Duncan and work around him. But even so, Duncan had the position versatility to win acclaim as possibly the greatest power forward of all time. But, if instead of Robinson it'd have been Malone, Duncan could have also played center for his whole career and been hailed as one of the greatest. He'd have worked next to Russell, or next to Wilt. He'd have worked next to Olajuwon, or next to KG. Next to Kareem or next to Dirk. Now, some of those fits might have taken a bit of work (like the Robinson one did) and there would be some redundancy, but it would work.

Shaq? If you pair him next to any center, it's not working. He's a pure 5. Which again, isn't a bad thing, but it means that Duncan fits on more teams. But it goes further.

Duncan also has skill-set versatility. If the team was full of defensive talent, he could be your best offensive player and spend more energy there. If your team was full of offensive talent, he could be your best defensive player and spend more energy there. He often played some combination of both roles in his prime. And his impact would stay roughly similar across those roles.

Shaq? He demonstrated that he could be the lead dog next to high scorers like Penny and Kobe. But there was friction at both stops, because Shaq wanted things run through him (for good reason). As an older player he showed the ability to defer to Wade, but that was as an older player. If Shaq would have been on a team with Jordan or with LeBron...or if he and Wade would have been more similar in age...could Shaq really modify his game, maybe make himself into a defensive anchor so that both could maximize their impacts at the same time? It seems unlikely to me. Given the way that he had to be the big dog on offense in both Orlando and LA, I don't think he'd have been receptive or able to modify his game in that way, even if it were for the good of the team.

Shaq's dominance, offensive uniqueness and attitude made him one of the highest impact players of the modern generation. But they also make him only able to really reach his potential in one way. Now, that's enough when compared against 99.9% of the players in NBA history because Shaq really was just that outstanding. But against someone like Duncan, who was almost as impactful as Shaq at his best but who would fit so much better on just about any team...a player like Duncan that could not only have his max impact under a lot more situations, but also would allow his teammates to have their maximum impact under so many more situations...against someone like that, Shaq's dominance might not be enough for me.

I'm still willing to listen to cases so I'm not going to vote just yet. But suffice it to say that if I was leaning Shaq before I read your post, I'm leaning the other way at the moment.

Trust me, I don't think this is an easy decision. I made that really long pro-Shaq post but as I was doing it I also realized some of the same pro-Duncan points that you did. I'm not really disappointed that my pro-Shaq post actually made you change your mind to pro-Duncan lol. I just wanted to make a deeper case for Shaq than the same arguments we've been seeing but I wasn't going to kill Duncan just to boost "my guy".

Just as you pointed out, I think the most important part of this debate is the value of Shaq's unique skillset. I believe there is added value to his skillset being so different from anything else we've seen in recent history while you believe that this uniqueness actually limits him compared to more traditional impact bigs.

I agree with you about Duncan's portability. I agree that his more traditional Def/Off value split for a big + his PF/C versatility makes him a more universal fit. That's certainly a plus for his argument. But I also feel like Shaq is a Center who can play next to pretty much any type of PF outside of a Reggie Evans-type goon with no jumper. So I think portability is more of a plus for Duncan than it is a minus for Shaq.

And I also feel that while Duncan could fit with more teams, Shaq could elevate his "best fit" teams to a higher level than Duncan could elevate his "best fit" teams. So while Duncan has the edge in team floor, I feel like Shaq has the edge in team ceiling. I believe that Shaq's offensive game opens up things for so much of his teammates and would just cause matchup nightmares that Duncan could never create.

Damn this is confusing. No wrong choice and I certainly see the Duncan argument. But I just feel like Shaq's uniqueness was a positive for his teams because he was just impossible to guard without sacrificing elsewhere on defense, and while he was providing this uniquely historic offensive impact he was also a very good defensive anchor in his prime who could step it up to elite levels for stretches.

[shutupandjam]

Shaq vs. Duncan is really difficult. They're contemporaries to some degree, they have both been wildly successful, and yet they're so different.

The way I think of this project (at least at this point in the rankings) is this: you get any player for his career up to today. In any given situation (random team, random era), who will give you the most titles on average?

You may be able to win more titles with Duncan. His sustained excellence is terrific, and in my opinion, superior to Shaq's. I think I would take any version of Duncan over any version of post-06 Shaq. If you split 98-05, which I actually think may favor Shaq, that gives Duncan a ~3 season advantage. But with Shaq, you get more of a sure thing. I honestly think you have as good of a chance of winning at least 1 title with Shaq than with anyone else. His peak was super elite and easily better than Duncan's in my opinion.

So what do you take? steady with a chance for more titles over time or the 'sure thing' (not really obv but as close as you'll get) that still gives you at least 10 really quality years? It's a really difficult question to me, but I think we tend to overemphasize Duncan's post-prime years and under-emphasize Shaq's pre-prime. In other words, I think Shaq's peak advantage is greater than Duncan's longevity advantage. For that reason, I think I'd go Shaq here.

[DQuinn1575]

I Does everyone understand the difference between npi rapm, pi rapm, and xrapm? Does anyone have specific questions about the metric(s)? I feel like we're throwing "rapm" out there too much without considering its context.

No, I do not.
If you plan on using them for evaluations, it would be helfpul for me (and probably others), if youd could start with a short paragraph or two (no more) explaining each one. Adding suggested reading will allow me (and hopefully others) to learn more.

Thanks

[microfib4thewin]

I love the people who bring up Duncan playing with Robinson as a portability argument. For one, we never saw Shaq with a center like that so we can't speculate on how it would work. More importantly it is such a rare situation that there is no reason to take it into account for portability. Good 7 footers are rare. Having 2 of them is even rarer. The only Twin towers situations we have seen are Duncan/Robinson and Hakeem/Sampson. Most of the other 1000+ teams in history didn't have 1 good center let alone 2.

It wouldn't make much sense to pair Shaq with another interior big. Shaq has never shown the range or the mobility to split time at PF.

Like others have said, there have been plenty of twin tower situations, but very few worked out as well as Duncan and D-Rob did.

[MacGill]

I love the people who bring up Duncan playing with Robinson as a portability argument. For one, we never saw Shaq with a center like that so we can't speculate on how it would work. More importantly it is such a rare situation that there is no reason to take it into account for portability. Good 7 footers are rare. Having 2 of them is even rarer. The only Twin towers situations we have seen are Duncan/Robinson and Hakeem/Sampson. Most of the other 1000+ teams in history didn't have 1 good center let alone 2.

It wouldn't make much sense to pair Shaq with another interior big. Shaq has never shown the range or the mobility to split time at PF.

Like others have said, there have been plenty of twin tower situations, but very few worked out as well as Duncan and D-Rob did.

Well to be honest here, again, look at how much bigger Shaq was. He has size 23 shoes. D-Rob was very agile for a 7'1 big but next to Shaq, you certainly aren't suprised that Shaq was less mobile. It's great that it worked for them, but this shouldn't be a negative against Shaq. His style of game was made for the middle and what poster's should be doing is deciding on if variety exceeds more one dimensional at an insane clip?

If you're that good at doing your thing, and it works, what does it matter if others do other things differently but not on the same level?

[Reservoirdawgs]

I feel like there may be at least some confusion about the different forms of RAPM in general, and this may be important with respect to how we're using it as an evaluation tool. Does everyone understand the difference between npi rapm, pi rapm, and xrapm?

I think it would be valuable to discuss the differences. I have an understanding of all three (I have a background in data analysis and before he was banned I was talking with mysticbb about going over his datasets so I can better understand his methods) but for my own benefit (and others) I think it would be good to have some paragraphs explaining the differences.

[penbeast0]

Technically I should note that in the last vote, with Wilt, the contest was declared over prior to pre-declared 24 hour period, as soon as Wilt obtained a majority of existing votes (with UBF's sealing vote), but whatevs.

Actually, I aim for "2 days/1day." I have to work around my schedule to post things. If I'm going out, that means posting around 5-6P Eastern or, if I can't get it done, when I return like last night. So tonight, I'll post around 6P if I can even though I only officially posted the runoff around midnight last night.

[Notanoob]

Straight up, I am very skeptical on those who are running off Duncan over Shaq yet allowed Wilt in at #4? It's not the voting placement but the reasoning that I am listening to here. Wilt/Shaq have very similar weaknesses and in all honesty, I'm surprised then that Duncan wasn't put in a run-off versus Wilt, if this is how many feel about attitude and longevity. Not to harp on the specific voting (as we all have our favorities) but we have some posters borderline bashing players where supporters are becoming defensive (myself included) looking to have balance restored.

If Duncan's career is good enough to rival with Shaq, which it is btw, and attitude and longevity are now so important, which he beats the other 2 in both, then Wilt's peak while higher, like Shaq's, and lesser years of some higher play, like Shaq, shouldn't have mattered then.

All I am seeing now is some voters picking and choosing who they want where and not being consistent down the board. Because for everything I have read in what our panel is valuing (and it's not all of you btw) Duncan should then have been slotted 4th.

**Again, while the placement doesn't matter, it is what we are trying to objectionally do here. And if we are going to go through all the work of it, we all might as well be as consistent as possible in the ranking system. If peak and attitude weren't as much issues before against better attitude and longevity, then they shouldn't all of a sudden become ones now.***

I just hope the rest of the project doesn't follow this path. Cause guys like Hakem/LBJ/KG deserve a lot of praise for their individual play.

I think that this post bears repeating.

Shaq and Wilt are practically the same player. In terms of how you evaluate them, they both had crazy peaks, both were inconsistent and had drama issues with teammates and management, neither has a longevity advantage over Duncan. I'm confused as to why people put Wilt ahead of Duncan no problem, but now aren't putting Shaq ahead of Duncan.

I think this is the first instance of people subconsciously changing their criteria for voting in this project, which was something some on here were worried about.

[colts18]

Shaq's averages vs top 5 defenses from 93-03:

27-13-3, 54.2 FG%

Against top 3 defenses:
27-13-3, 53.9 FG%

No one dominated top defenses like that. Shaq was one of a kind.

[shutupandjam]

I. Does everyone understand the difference between npi rapm, pi rapm, and xrapm? Does anyone have specific questions about the metric(s)? I feel like we're throwing "rapm" out there too much without considering its context.

No, I do not.
If you plan on using them for evaluations, it would be helfpul for me (and probably others), if youd could start with a short paragraph or two (no more) explaining each one. Adding suggested reading will allow me (and hopefully others) to learn more.

Thanks

Ok, I'll try to keep this as basic and short as possible. Just in case it ends up being too long, I'll bold the especially pertinent sentences.

Adjusted Plus Minus ("APM")

APM is the logical place to start because rapm evolved from it. Raw plus minus is interesting in context, but it's basically useless over time because it doesn't isolate a player from his teammates. To try to correct this problem, Wayne Winston and Jeff Sagarin created APM (they called it WINVAL).

The goal is to isolate a player from his teammates by using regression analysis. To do this you take every 5v5 matchup stint that occurs during an nba season (there are usually ~30,000 different ones) and run a linear regression where the dependent variable is the point margin during each stint (home pts/100poss - away pts/100poss) and the indepedent variables are indications of whether each particular player was on the floor (1 for home players -1 for away). When you run the regression, you get coefficients that represent the ('isolated' plus minus) value of each player.

But there are problems with APM: low minute players often get absurd values (often you just take low mp guys out of APM regressions), the range of values is quite large, and there is very little consistency from year to year.

Non Prior Informed RAPM ("NPI RAPM")

In order to correct some of APM's biggest issues, Joe Sill developed RAPM, which is exactly the same as APM except that it uses a ridge regression instead of an ordinary least squares regression.

Simply put NPI RAPM is RAPM for one year only - single season RAPM

If anyone wants Sill's original paper, PM me and I'll send it along. To put it basically, ridge regression assumes all the coefficients are close to 0, and the more information we have about a coefficient (e.g., a player who played many minutes), the further it can stray from 0. The biggest advantage RAPM has over APM is that it prevents overfitting (low mp players no longer get absurd values, and the range of values is much more reasonable). It also helps multicollinearity to some degree - that is, it helps to disentangle players that play many minutes together. According to Sill's study, one year rapm is as accurate (similar r-squared) as three year apm.

The main disadvantage is that it introduces a bias - i.e., that each value is close to 0. It is also not particularly great (but obv much better than APM) at predicting future seasons and also isn't terribly stable.


Prior Informed RAPM ("PI RAPM")

I mentioned that ridge regression assumes the values are close to 0. We call 0 the "prior." You can change the prior and replace it with any value that you think is more reasonable.

PI RAPM uses RAPM from previous years in order to improve stability and predictive value.

Techincally, PI RAPM could use any prior, but in today's jargon we usually use it to refer only to RAPM where the prior is the previous season's RAPM (which used the previous season's RAPM and so on).

Again, it improves stability and predictive value (plus it looks way more reasonable), but it is influenced heavily by previous seasons (even with an aging curve, it has trouble giving up on certain players !Garnett).

xRAPM/RPM

Rather than using rapm from previous seasons as the priors, these forms use a combination of rapm from previous seasons, spm from previous seasons, and spm from the current season as a prior.

In other words, xRAPM uses RAPM and Box Score elements from current and previous seasons to improve predictive value and stability.

Again, this all helps xRAPM become the best out of sample predictor of the RAPM family, but it introduces biases of its own and heavily relies on the box score and previous seasons. And the way the spm (box score) regression is set up really matters - this is the (2014) SPM used in xRAPM, for example.

-----

All of this stuff can add value to these discussions as long as its used in context. Personally, if I say "RAPM" I'm talking about NPI RAPM, but it's probably best at this point to distinguish so everyone knows specifically which iteration you're talking about.

One important factor to keep in mind is that parsing play-by-play data is very difficult to do perfectly. As a result, different people will come up with different RAPM results for the same season (though they shouldn't be wildly different). Additionally, how people account for other factors (HCA, coaching, point margin when the stint begins, etc.) will change the results.

Hopefully this helps clear things up to some degree. If anyone has more questions, I'd be happy to answer them. AcrosstheCourt could also be helpful here, and it's really too bad mystic got banned because he's knowledgeable about this stuff too.

[colts18]

Shaq's averages vs top 5 defenses from 93-03:

27-13-3, 54.2 FG%

Against top 3 defenses:
27-13-3, 53.9 FG%

No one dominated top defenses like that. Shaq was one of a kind.

Duncan's # vs Top 5 defenses. For his whole he has just 6 series vs top 5 defenses (Shaq has 20 series vs top 5 defenses).

22-14-4, 46.4 FG%

Duncan has just 3 career series vs top 3 defenses while Shaq has 13 of them.

[drza]

And I also feel that while Duncan could fit with more teams, Shaq could elevate his "best fit" teams to a higher level than Duncan could elevate his "best fit" teams. So while Duncan has the edge in team floor, I feel like Shaq has the edge in team ceiling.

And this is why I keep holding on the vote. I believe that Shaq at his best was better than Duncan at his best, and I even put some quantitative evidence out there about that "better than Duncan" impact extending down through 5+ at least. So, if I'm convinced that Shaq had a higher impact with his "do it my way" game than Duncan did across all of the different iterations of the Spurs in the last 17 years...is it really, at this point worth choosing Duncan over Shaq just for roster flexibility?

I'm a big believer in portability, but I'm also a big believer in impact. In the last top-100 project I remember having a long debate with El Gee about Karl Malone vs Garnett. El Gee argued that Malone's extra longevity at a super high level was worth more than KG's (then) great-but-not-as-long longevity with a few years with a higher peak. Essentially, I was voting that the higher impact potential at the peak few years was more important than a bit of longevity. El Gee "won" the debate and Malone was voted in before KG, but I still think my argument was right for me.

This is a different case, but there are some siimilarities. And I worry that a vote cast for Duncan would be analogous to voting for Malone in the previous example...choosing the lower impact player that had more consistency at that lower level. And if that's really what it comes down to then I don't think that would be enough. So for now, I remain stuck on my vote as I wrestle it out.

[drza]

Straight up, I am very skeptical on those who are running off Duncan over Shaq yet allowed Wilt in at #4? It's not the voting placement but the reasoning that I am listening to here. Wilt/Shaq have very similar weaknesses and in all honesty, I'm surprised then that Duncan wasn't put in a run-off versus Wilt, if this is how many feel about attitude and longevity. Not to harp on the specific voting (as we all have our favorities) but we have some posters borderline bashing players where supporters are becoming defensive (myself included) looking to have balance restored.

If Duncan's career is good enough to rival with Shaq, which it is btw, and attitude and longevity are now so important, which he beats the other 2 in both, then Wilt's peak while higher, like Shaq's, and lesser years of some higher play, like Shaq, shouldn't have mattered then.

All I am seeing now is some voters picking and choosing who they want where and not being consistent down the board. Because for everything I have read in what our panel is valuing (and it's not all of you btw) Duncan should then have been slotted 4th.

**Again, while the placement doesn't matter, it is what we are trying to objectionally do here. And if we are going to go through all the work of it, we all might as well be as consistent as possible in the ranking system. If peak and attitude weren't as much issues before against better attitude and longevity, then they shouldn't all of a sudden become ones now.***

I just hope the rest of the project doesn't follow this path. Cause guys like Hakem/LBJ/KG deserve a lot of praise for their individual play.

I think that this post bears repeating.

Shaq and Wilt are practically the same player. In terms of how you evaluate them, they both had crazy peaks, both were inconsistent and had drama issues with teammates and management, neither has a longevity advantage over Duncan. I'm confused as to why people put Wilt ahead of Duncan no problem, but now aren't putting Shaq ahead of Duncan.

I think this is the first instance of people subconsciously changing their criteria for voting in this project, which was something some on here were worried about.

Speaking only for myself, I had both Shaq and Duncan over Wilt. I would have voted for either in the run-off over Wilt (I did vote for Shaq), and I was pretty surprised that Wilt went in when he did. Not that I think these posts were necessarily directed to me personally, but the phenomenon you're describing certainly isn't universal, if it's there at all.

[ardee]

I think it's interesting that a good part of Shaq, Duncan and KG's primes overlapped, and it would bear some fruit to have a look at how they compared at the time.

2000
Shaq
KG
Duncan

2001
Shaq
Duncan
KG

2002
Shaq
Duncan
KG

2003
Duncan
KG
Shaq

2004
KG
Duncan
Shaq

2005
Duncan
KG
Shaq

I think 2005 Duncan/KG was pretty close.

[ardee]

Shaq and Wilt are practically the same player..

Disagree here.

They may be similar athletically but stylistically and how they impact the game... no.

Wilt was much more durable, a better shooter, more consistent on defense, better rebounder, and better at running the offense from the high post ala Walton/KG.

Shaq was better in the low post, arguably better at passing out of double teams, and less durable.

Especially after Wilt got coached by Hannum he was very different from Shaq.

It's perfectly reasonable to have Duncan between Wilt and Shaq, considering both their durability and defensive edge over Shaq.