Re: Rebound Effect
- --- In APBR_analysis@y..., "Dean Oliver" <deano@r...> wrote:
> Rodman was the worst in that sense. His claim as the greatestHmm, I suppose I never thought about that. Obviously, though they're
> rebounder of all time, which makes sense from a purely individual
> perspective, can be questioned based on what you found.
> The other way to look at it is how his teams rebounded with and
> without him (which will show that his teams were good rebounding
> teams, but the best? I dunno.)
treated the same in the stats, not all rebounds are created equally.
Some just bounce into a player's hands, while others require a great
deal of effort in terms of positioning and boxing out. Is it
possible that the reason we see what we see here is that a higher
percentage of Rodman's rebounds were these "easy" rebounds that any
of his teammates could have claimed had he not been there?
Either way, assuming the Rodman effect holds in general on more
observation, it's another factor diminishing the value of rebounding
specialists like Fortson. Rebound rate would overstate their real
contribution to the team in that area (and, in the same sense,
really bad rebounders might be undervalued, I suppose, if they're
leaving more boards for teammates. Which makes me wonder -- did
Wallace's rebound rate go up when he got Clifford Robinson as a
teammate? To answer my own question - no, he went down from 18.3
rp48 to 17.1. Though I'm not sure how exactly to explain his off-the-
charts 20.2 rp48 this season save sample size.)
- --- In APBR_analysis@y..., "monepeterson" <mone@s...> wrote:
> > To try to get some idea, I took a look at players' rebound rates'Available rebounds' is surely a key concept in such discussions.
> > , with, and post-Wallace on Detroit and Orlando:
> > Corliss Williamson:
> > pre (2000-01 w/Toronto): 8.29 rp48
> > with (2000-01 w/Detroit): 10.08 rp48
> > Well, that's strange. A major boost.
> I gotta think that these results (and some of your other ones) are
> affected by the number of rebounds available on each team. Perhaps
> they're moving to a situation where more rebounds are available?
In 1991, Toronto's team rebound differential was 44.3-41.5, so they
were a good rebounding team. Detroit averaged 45.0-44.5, just above
average, and with more rebounds around.
Saying there were 85.8 rpg in Toronto and 89.5 rpg in Detroit would
imply that one's rebounds per minute should be higher in Detroit.
A further distinction might be made between rebounds gathered by your
own team vs. opponents. I have come to believe in scaling only the
opponents' rebound totals.
In this case, considering total rebounds gives Corliss a per-36 rate
in Toronto of 5.75, and 7.29 in Detroit.
Considering only opponent rpg yields a change from 5.94 to 7.33.
The difference is now less dramatic ( +23% rather than +27%)
It's my belief that making this adjustment would tend to smooth the
majority of transitions from one team to another, for most players.
And I would consider this evidence that it's a viable adjustment.
In this case, the Pistons apparently needed scoring even more than
they needed rebounding, and this is what the guy does (did) well.
Even his turnovers dropped about 20%, despite the increased offense.
Such across-the-board improvement can never be predicted by mere
About 10 years ago, I did look at Rodman's effect on his teammates.
I'm sure it was on paper and now lost; not sure how I did it, but it
seemed clear he was depriving teammates of rebounds, once he decided
to be a 'specialist'.