lorak wrote:Owly wrote:lorak wrote:If we want to know if defense in playoffs performed under/over expectations, then we have to compare playoff series drtg to expected drtg (where exp drtg = (team RS drtg+opp RS ortg)/2).
Results for Spurs with Robinson (negative value is good):
Code: Select all
YEAR OPP DRTG
1995 LAL -9,3
1995 DEN -4,2
1993 PTB -2,5
1990 PTB -2,1
1993 PHO -0,1
1990 DEN 1,1
1996 PHO 2,4
1995 HOU 3,0
1994 UTA 4,0
1991 GSW 4,1
1996 UTA 5,7
So 11 playoffs series, 4 times Spurs defense performed better than expected, 1 time basically at expectation level (1993) and 6 times worse than expectations. Doesn't look good for Robinson. Of course we should look closely at each series (for example I wouldn't blame DRob for 1991 as it was great coaching job by Nelson, who outcoached Larry Brown), but pattern seems rather clear.
I don't know which set of Robinson defense numbers to trust, but on these ones I'd suggest that there is not a "rather clear" pattern.
So I will elaborate (and if you don't trust my numbers, then check it by yourself

). The same data as above, but with opponents regular season ortg relatively to league average:
Code: Select all
YEAR OPP DRTG opp ORTG
1996 UTA 5,7 5,7
1993 PHO -0,1 5,3
1991 GSW 4,1 4
1992 PHO 12,6 3,9
1996 PHO 2,4 2,7
1990 PTB -2,1 2,4
1994 UTA 4,0 2,3
1995 HOU 3,0 1,4
1995 LAL -9,3 0,8
1995 DEN -4,2 0,8
1993 PTB -2,5 0,3
1990 DEN 1,1 -0,1
So Spurs defense usually played above expectations only vs barely average offensive teams - with one exception in 1990 (PTB). Other than that their defense was awful against teams with offenses with at least 1 ortg point above league average. And that's that clear pattern I'm talking about, because Robinson's consistent failures against good teams is something what many people hold against him. Sure, we can say that in 1991 it was Brown's lack of adjusments or in 1995 it was Rodman's bad influence, but it still leaves us with other series and I honestly doubt we can use excuses for every one of them. That was just essential flaw in DRob's game - both on offense and defense - that he wasn't able to perform constantly at good level vs better opponents and his game was exposed in the playoffs.
To clarify (don't know if this is necessary, but feel I should) the trust thing was with regard to multiple people posting different numbers based analysis.
The fuller explanation of the pattern is appreciated because, as I noted previously, the overall impression given by the first set of numbers was that the Spurs met expectations defensively.
Now I'm still not full convinced but this is interesting. My problems/queries:
1) Is your hypothesis that his lesser defensive impact (versus good teams) is playoff specific or more general or do you think it happens in the RS too (and if so do you have evidence for this?)?
2) Isn't team Drtg a rather blunt tool for individual D on fairly small samples (which is another reason I'd be more comfortable with a "defensive impact fell versus good teams" conclusion if had a larger - i.e. regular season - sample.)
3) Do we have comparisons for other defensive anchors with reasonable sample sizes (I'd imagine a good D should be more reliable than a good O in the playoffs - good Ds typically require other positive indicators that help elsewhere good communication, teamwork etc, though perhaps this wasn't happening with Rodman, but I digress - so elite offense maybe should cede to good defense. But do we have evidence on this. Do we have evidence where (the goodness of the) defense is basically all on the anchor (which, with the exception of Rodman, it typically was).
Massive tangent here, do we know Rodman helped the Spurs on D? It's certainly imperfect because of roster turnover, but in 92-93 in a Robinson down year (including defensive statistics) and with coaching turmoil their relative Drtg is -1.2; the next year it's -1.7. Not a huge visible (defensive) impact.
I might be clutching at straws because I like Robinson; but really I'm just curious and want to know more.
And Jim, you're using raw stats again. Please, please stop. You say And DWS is flawed but even if it weren't 2.4 DWS
over 1615 minutes isn't better than 2.2 DWS
over 1122 minutesM Malone (3 year peak, playoffs): 0.071331269 DWS/48
D Robinson (3 year peak, playoffs): 0.094117647
Don't use raw Win Shares (offensive, defensive or cumulative) over playoffs because it skews hugely favourable to the guy who played on the better team (or more precisely the team with the longer playoff run in terms of minutes).
Otherwise '93 playoffs Dan Majerle (which Win Shares per 48 and PER agree was below average) was better (2.1 WS) than say '86 playoffs MJ (0.5 WS), '87 or '88 playoffs Olajuwon (2.0 and 1.3 WS) and eight of Bill Russell's thirteen playoffs.
magicmerl wrote:Through to post 323
10 Kobe Bryant (GC Pantalones, ShaqAttack3234, ardee, batmana, JordansBulls, Basketballefan, DQuinn1575, john248, lukekarts, ronnymac2)
5 Karl Malone (Baller2014, therealbig3, trex_8063, magicmerl, FJS)
2 Dirk Nowitzki (rich316, Quotatious)
2 Jerry West (penbeast0, Notanoob)
2 DrJ (Warspite, Clyde Frazier)
1 David Robinson (shutupandjam)
Leaning Kobe, but not voted (An Unbiased Fan)
Leaning West, but not voted (RayBan-Sematra)
If they had a vote
1 Karl Malone (dautjazz)
1 Isiah Thomas (RightToCensor)
1 Charles Barkley (Purch)
1 Moses Malone (Jim Naismith)
1 David Robinson (BallerTed)
Looks like a runoff may not be needed.
Post 307 I cast a tentative vote for Robinson (sorry, I didn't bold it and left it as somewhat of a footnote of a much larger post).