#1 Greatest Defensive PF of All-Time - Top 10 Defense at each position project

[iggymcfrack]

Voting Kevin Garnett. Much more dominant peak when he was in the true rim protector role as he had better DRAPMs in ‘08, ‘09, and ‘11 than Duncan ever had in his career. He was also versatile enough to be very effective in other roles as well. Has a legit argument as the best defensive player of all-time.

[penbeast0]

I'd certainly have him above Rodman if you don't count defensive rebounding as defense. If you do, Rodman takes a massive jump in utility.

I'll never understand the argument on how defensive rebounding is not defense...it's an action that prevents the opposing team from scoring almost by its very definition.

I mean, it can be, but it's not always that. Most of the time I agree with you and I count defensive rebounding as defense, but I don't know that I count a lot of Rodman's defensive rebounding as defense, considering how much time he spent chasing boards at the expense of actual defense.

I think you still count them, you just also count the open shots that Rodman's man got while he was chasing rebounds against him too.

I do count defensive rebounds as defense unless the OP asks us not to do so; I also count players who don't communicate or ignore defensive signaling as bad defense.

[Notanoob]

Looking at the previous results, I feel like there ought to be two different projects going on here - Top 10 defenders among star players (not sure on how best to define that, but we can work on that) and top 10 among non-stars. I think it's hard for us to mentally separate a guy's defensive impact from his overall impact, and I think that's reflected in the results. This should be less of an issue for centers than wings, since we judge centers heavily on their defense, but I still think that it'll be an issue.

I also think that we should probably have split this into guys who played their primes before advanced stats and guys who played after they became a thing. I think it helps push a recency bias since we cannot bring up Bobby Jones' DRAPM to make the case for him over Draymond Green, for instance.

This isn't to say that I expect you to just drop this project and start over with multiple concurrent projects asking mostly the same question, I'm just throwing my opinions out there.

[cecilthesheep]

Looking at the previous results, I feel like there ought to be two different projects going on here - Top 10 defenders among star players (not sure on how best to define that, but we can work on that) and top 10 among non-stars. I think it's hard for us to mentally separate a guy's defensive impact from his overall impact, and I think that's reflected in the results. This should be less of an issue for centers than wings, since we judge centers heavily on their defense, but I still think that it'll be an issue.

I also think that we should probably have split this into guys who played their primes before advanced stats and guys who played after they became a thing. I think it helps push a recency bias since we cannot bring up Bobby Jones' DRAPM to make the case for him over Draymond Green, for instance.

This isn't to say that I expect you to just drop this project and start over with multiple concurrent projects asking mostly the same question, I'm just throwing my opinions out there.

There are some advanced stats that go back a long way, though, like PIPM and WOWYR. I try to use those as much as I can with guys like Jones.

Out of curiosity, which results do you think are evidence of star bias? I thought we did a pretty good job with the guard positions, and while I do think the SF list is completely wack, I think that's more because of an excessive emphasis on longevity over peak and obscure advanced metrics over eye test than anything else.

[LA Bird]

Past his first few seasons, I would say Duncan was more consistent over the course of his prime whereas Garnett had higher highs but also lower lows. KG in Boston was higher in terms of per possession defensive impact than Duncan ever was but he was also averaging less minutes by then so his overall impact wasn't that much better than Duncan's. There is a clear stylistic difference between the two and while it is easy to say Duncan plays a more traditionally center role with his shotblocking and rim protection (whether that is good or bad for a GOAT PF list), Garnett also had some of his best defensive years as a center towards the end of his prime. All in all, a extremely close comparison but I would go with KG by the smallest of margins.

Vote: Kevin Garnett

(BTW, I think we should skip the #2 thread and go straight to #3. It's obvious one of Duncan/Garnett is going to win in a landslide)

[kendogg]

Given equal defense, I do give more weight to the better overall player, because they are asked to do more on the offensive end and thus can't give the large majority of their energy on defense. Guys like MJ in key possessions have an "extra gear" they can shift to on defense where as the defensive specialists do not have this gear. And I do feel like MJ was a very consistent defender for example and his rank is warranted. LeBron's rank, on the other hand, is a bit high. There is just better defenders at forward so the competition is tougher. And the best defenders play center so the competition is even tougher. If we took all 5 lists and ranked them together the centers would mostly all be at the top and the guards mostly at the bottom.

[penbeast0]

I do agree with LA Bird that it won't be close for #2, and yes, my #2 would be Garnett.

[cecilthesheep]

(BTW, I think we should skip the #2 thread and go straight to #3. It's obvious one of Duncan/Garnett is going to win in a landslide)

You'd think this, but look what happened to Kawhi and AK47 in the small forward voting. They were in contention as high as number 2 and then everyone just abandoned them.

[penbeast0]

Oh, we have to have the vote, just agreeing with LA Bird that it's going to be one of those two.

[trex_8063]

Thru post #49:

Tim Duncan - 7 (HurricaneKid, 70sFan, penbeast0, pandrade83, kendogg, trex_8063, WestGOAT)
Kevin Garnett - 7 (Dr Positivity, KnickFan33, lebron3-14-3, cecilthesheep, HeartBreakKid, iggymcfrack, LA Bird)
Bobby Jones - 2 (SkyHookFTW, JordansBulls)
Draymond Green - 1 (Heej)

We gotta go into runoff between KG and TD. If you did NOT already vote for one of them, please state your pick between those two. Let's try to conclude by this evening.


Hi guys

Spoiler:

a

a

a

a

a

a

a

a

a

a

a

a

a

.

.

a

a

a

I'

a

.

a

a

a

a

a

a

a

a

a

a

a

)

)

)

)

)

)

)

)

)

)

)

)

)

)

)

)

)

[

)

)

)

)

)

)

)

/

/

[PigsOnTheWing]

I'm siding with Duncan on this one. I've always slightly preferred Tim's defense to Garnett's though the difference in their impact is minimal. Just to give a reference, I ran their numbers into my metric and here's the output:
CAREER: Duncan 4117.0, Garnett 3702.2
5-YEAR PRIME: Duncan 1389.7, Garnett 1246.8
PEAK: Garnett 304.1 (2004), Duncan 290.3 (2001)
Duncan's lead in career value comes from their last three seasons, in which he was one or two tier above a washed up Garnett and thus he gained that small avìdvantage. Going with RAPM gives basically the same results: from '97 to '14 Duncan has been 3010 points better than average while KG 3035.

So my decision is based on personal preference and I think Timmy was the main reason why the Spurs' D has been consistently among the best in last 20 years, much more than the so called "Spurs culture", the system, Pop and his teammates. I'm not saying that they didn't have an impact on his career but to me it's usually overstated while a few consider the opposite (i.e. that was Tim the culture in itself) which has also been said by Pop after all. Even in his last few years, despite the rise of Kawhi and his lack of mobility, I've always felt that he was anchoring that team thanks to his unparalleled BBIQ and instincts.

Now, the counterpoint is that the roster turnover was constantly really low so he didn't have to adjust his game to different teammates or strategies like Garnett did as he happened to be stuck in a, let's say, badly run franchise and thus, particularly in his last seasons, needed to drag the team by himself on offense at the expense of defense. I really wish we could go back in time and swap the franchises and see what happens. But, after all, KG got his revenge in Boston and in the few remaining years of his prime he showed how good he was as a defensive anchor. I can only imagine the impact he could have had in a situation like that if he was five years younger. But nontheless we are left with enough to make a fair judgement and, in the end, I prefer Duncan by a hair.

[Samurai]

As a PF, I would take Garnett over Duncan on defense since I view Duncan as playing defense more in the manner of a center than a PF. And I've disagreed with listing Duncan as a PF in this exercise. But since those things don't matter, and since I think Duncan plays defense as a C better than Garnett, and given the importance of a center's defense in the era they played in, my runoff vote is for Tim Duncan. Of course, if we counted Bill Russell as a PF he would get my vote. Like I said, doesn't feel right but given the rules that were set, that's the result.

[Dr Positivity]

I agree with just skipping to #3 after this thread

[GusFring]

No denying Duncan's impact on winning 5 rings vs 1 but I just prefer KG out of pure entertainment. Favorite pf ever with the constant jaw jacking and insane defensive anchoring.

Can I stop getting notifications on these posts please?

[HeartBreakKid]

No denying Duncan's impact on winning 5 rings vs 1 but I just prefer KG out of pure entertainment. Favorite pf ever with the constant jaw jacking and insane defensive anchoring.

Can I stop getting notifications on these posts please?

Is this a vote or what?

[GusFring]

No denying Duncan's impact on winning 5 rings vs 1 but I just prefer KG out of pure entertainment. Favorite pf ever with the constant jaw jacking and insane defensive anchoring.

Can I stop getting notifications on these posts please?

Is this a vote or what?

sure, one vote for the big ticket plz

[Gibson22]

2-1 for duncan as of now, let's continue the run-off for another 5 hours (12 in total as usual)

[penbeast0]

2-1 for duncan as of now, let's continue the run-off for another 5 hours (12 in total as usual)

welcome back!

[trex_8063]

I'll go ahead and say it: shot blocking is not defense.

Does it correlate many times to great defenders? Yes. But throughout the years, we've seen that the impact of shot blocks to overall defense is not very important.

In fact they're actually less valuable than just making a guy miss a shot, because a blocked shot is more likely to be rebounded by the offense.

None of that is always true though. For instance, the Duncan signature move of blocking a shot to yourself rather than swatting it away is absolutely defense. And that's not even taking the psychological impact of blocks into account.

Fundamentally, when a player blocks a shot, he's getting in the way of the ball going in the basket; I think we have to call that defense. Saying it's not defense because you can do it in a way that doesn't help your team is like saying scoring isn't offense for the same reason.

Well yes, a block is "defense" because it is, literally, an act of defending. But is it "good" defense? Not really. No more than just getting in someone's face and contesting. When it comes to determining how good a player or team is at defending, blocks are completely meaningless.

You bring up Tim Duncan. Well he's smack dab between Brendan Haywood and Andrew Lang in career block %. That's how useless a barometer it is. Now if you wanna go watch tape of every single one of his career blocks and show me that the defensive rebounding rate of his blocks is somehow way higher than everyone else's blocks, go ahead, but I'm gonna guess that that's not the case.

In fact I've long thought that blocks that result in a turnover are the only ones that should be counted in box scores. Blocked shots that don't result in a turnover might as well just go in a "defended misses" category.

But a block shot is just a result of contesting a shot. To say block shots are useless or bad defense is the same thing as saying contesting a shot is useless.

The main gist of what he’s saying is that it's not [at least in all instances] a great measure of individual defense: an elite-level block rate does not necessarily correlate to elite-level defense for a big man (though fwiw it probably almost NEVER coincides with a legit poor defender), and a low block-rate does not necessarily correlate to a poor defensive big man. And that’s all true.

However, he seems to lump all blocks together in his initial statement, as though they’re all equal; which I don’t think is true. It’s primarily this I want to speak to, but we’ll get to that later. I’m first going to respond directly to at least one other statement....

First: “In fact they're actually less valuable than just making a guy miss a shot, because a blocked shot is more likely to be rebounded by the offense.”

It’s true that blocked shots are more likely to be recovered by the offense than a missed shot: blocked shots are recovered by the offense about 40% of the time (give or take; NOTE: varies by player, more on that below); a missed shot is rebounded by the offense roughly 23% of the time in recent years (though used to be more in previous eras where crashing the offensive glass was favored over getting back on defense).
So one could say that the above statement is true if you’re guaranteed you can “make the guy miss” WITHOUT blocking the shot. But the only way to guarantee it misses is, in fact, to block it.
Otherwise, no matter how well-contested the shot is, some of them are going to go in. This is the “great defense, better offense” you’ll sometimes hear commentators say. Forcing the shooter into an extreme high degree of difficulty on the shot does NOT guarantee a miss. These are NBA players; they are [on average] both freakishly athletic AND freakishly skilled. They consequently pull off some amazing stuff. Even difficult and/or REALLY well-contested shots are going to go in at least ~20% of the time (probably higher than that, actually, but let’s use that figure for the following mathematical illustration):

*We’ll assume a well-contested shot goes in 20% of the time.
**When it misses (the other 80%), we’ll assume the offense retains possession (OReb) on 23% of those misses.
***Let’s say on the reset after retaining possession that the average 1.08 pts/possession (roughly league avg in recent years).
****A blocked shot goes in 0% of the time, but let’s assume is recovered by the offense 40% of the time (and they’ll average the above 1.08 pts on that retained possession).
*****For simplicity let’s assume the initial shot is always a 2-pointer.

Hard-contested (but not blocked) shot
(.2 * 2) + (.8 * .23 * 1.08) = 0.599 pts avg outcome for offensive team

Blocked shot
(.4 * 1.08) = 0.432 pts avg outcome for offensive team

So the avg blocked shot is the more valuable defensive play, even if we assume only 20% conversion rate on the contested shot. Basically, the shooter has to hit 10% or less on the contested shot before the contested shot becomes the more valuable defensive play. But that’s nigh on impossible with NBA players.

Now if you wanna go watch tape of every single one of his career blocks and show me that the defensive rebounding rate of his blocks is somehow way higher than everyone else's blocks, go ahead, but I'm gonna guess that that's not the case.

Actually, that DOES appear to be the case. I’d made some scouting comments back in the ‘13 season comparing Duncan and Howard in their defensive style and tactic (can’t find the post now). I’d commented how Tim Duncan blocks shots by taking advantage of footwork and positioning, being right on top of the shooter---hands at the ready---at the time the shot goes up, such that he scarcely needs to even leave his feet to go after and lightly tap the ball shortly after it leaves the shooter’s hand. And always just a light tap with a mind toward keeping it in play (often tapping to himself, as someone else mentioned).
Howard, otoh, would more just rove to a region, frequently with his hands positioned low, wait for the shot to go up, and take an enormous leap and swat at the ball as it’s nearer to its apex. And he’s often batting toward the first rows, even when it’s clearly not necessary and could be kept in play:

Watch on YouTube

This is a much “sexier” play, and MUCH more likely to make the highlight reel (as evidenced above), and MUCH more likely to grab the attention of the casual fan. But it is the [demonstrably] less valuable block (compared to what Duncan is doing): a shot that is blocked into the front row goes to the defensive team precisely 0% of the time.

Here is an article specifically pertaining to the ‘13 season, which compares several of the league’s leading shot-blockers that year and comparing the % of their blocked shots that are recovered by the offensive team, the % of their blocked shots which are blocking lay-ups (hopefully self-explanatory why blocking a lay-up is [much] more valuable than blocking a jump-shot), and their blk/PF ratio.

The author then creates an equation based upon these factors to numerically describe the value added by the blocks they made per games, called this value the “Marginal Blocking Value”. I’m not sure I agree with all the inputs, but at any rate it indicates Duncan is a cut above the rest in the avg value of his blocks. He’s one of the best in % of blocks recovered, the single-best [by a handy margin] in the % of his blocks that come on lay-up attempts, and also has the best foul ratio (by a handy margin).

Fwiw, I remember another article [which for the life of me I cannot find now] that sampled (I think) an even larger sample of games (multiple seasons involved), and which included some of Howard’s physical peak (like circa-’11). Iirc, that article similarly had the % of Duncan’s blocks that were recovered by the offense at somewhere around 35%, but for Howard it was actually >50%.


Point being that not all blocks are created equal (and Duncan seems to be among the highest tier among his contemporaries in the value of his blocks).

[trex_8063]

OK, calling it for Duncan.