Advanced Stats and Dennis Rodman

[EvanZ]

Was Jayson Williams also a HOF-caliber player?

Top offensive rebounders in the 3-pt era

The game has changed a lot since the mid-90's when Rodman (and Williams) were at their prime. If you look at that list above, Kevin Love's 2009 season is the most recent after the top 50. There seems to have been a wholesale shift in the philosophy about offensive rebounding. This is not to diminish the achievements of Rodman (and Jayson Williams?), but it should be put in this perspective.

EDIT: Actually, DeJuan Blair and Jeff Foster show up higher up than Love, but the point remains.

[Chicago76]

Was Jayson Williams also a HOF-caliber player?

The game has changed a lot since the mid-90's when Rodman (and Williams) were at their prime. If you look at that list above, Kevin Love's 2009 season is the most recent after the top 50. There seems to have been a wholesale shift in the philosophy about offensive rebounding. This is not to diminish the achievements of Rodman (and Jayson Williams?), but it should be put in this perspective.

EDIT: Actually, DeJuan Blair and Jeff Foster show up higher up than Love, but the point remains.

What you've stumbled across is probably more of a random statistical oddity than an indication that ORB rates are declining for interior players. F-C/C-F/Cs playing 500+ min ( or their 98/99 shortened season equiv) provides some indication of this. ORB rates broken down by ORB era and indexed to 100, here are the average ORB rates for those "bigs" (weighting them by ORB opps for each player).

87-89 100.0
90-94 101.3
95-99 105.6
00-05 103.5
06-11 107.3

So rates have actually gone up for bigs.

Taking it a step further and removing perimeter shooting bigs (further from bucket, so tougher to get ORBs). Define perimeter shooting bigs as >=8% of their scoring chances from 3 (3a/(fga+0.44FTA):

87-89 100.0
90-94 102.0
95-99 108.4
00-05 106.0
06-11 111.3

Two obvious trends: 1) as a premium was placed upon defending transition O, the ORB rates of bigs actually went up a bit. 2) When the 3 pt line moved in, this altered spacing and appeared to pull perimeter defenders away from help rebounding situations in the middle of the floor.

1994: ORB index for both measures above were 101.5 and 102.9, respectively.
3 yr of shortened line: both measures were 104.2 and 107.5, respectively.

Interior big ORB rates moved even higher in 98 when the line moved back out. This was likely a 1 yr defensive strategy lag where perimeter guys got used to the idea of defending the 3 more closely, so they were pulling themselves even further from the basket. This bounce completely evaporated in 1999 and persisted until the end of the handcheck era through 2004.

One more thing to consider: maybe even though ORB rates for interior bigs have actually gone up a bit, there are other things at play which reduce the ORB "ceiling" for the best rebounders. 95th percentile ORB rate index:

87-89 100.0
90-94 102.9
95-99 104.7
00-05 100.1
06-11 102.3

Excluding the line adjustment era, it looks like there may be something of a ceiling, but it basically washes out w/ the benefits of the focus on defending transition baskets opening up the middle for ORBs and bigs. The differences are pretty small. 13.6 ORB rate in 87-89 and 00-05. 14.0 and 14.2 in 90-94 and 95-99. 13.9 in most recent period.

There may be some team-specific things going on, but I'm thinking that Rodman would be posting almost identical ORB rates even if his career came later. The avg 95th percentile ORB rate for his 14 year career was 13.8. The average over the last 14 full seasons was 13.9.

[EvanZ]

League average ORB% has been steadily declining for a long time. How do you account for that?

http://thecity2.com/2011/08/04/so-what- ... -30-years/

[Chicago76]

League average ORB% has been steadily declining for a long time. How do you account for that?

http://thecity2.com/2011/08/04/so-what- ... -30-years/

I realize that ORB rates have declined on a team level, but looking at the data above, the answer for why interior-oriented 4s and 5s aren't imapcted should be pretty obvious: the decline in ORB rates on the team level is not evenly distributed across all positions.

That's why I aggregated the years together the way I did too. There seem to be little "steps" down over the last 25 years, so I clustered years on the same "step". Rough lg ORB rate by cluster:

87-89 ~0.330
90-94 ~0.323
95-99 ~0.308
00-04 ~0.286
05-11 ~0.270

Reasons ORB rate has declined:

1) to prevent transition O the other way
2) to keep the floor spread for a kick out, and because perimeter players in general are probably 3-4 feet further out on O these days thanks to the three anyway. This means the perimeter defenders are covering a man on the floor who is further from the basket too.

Both of these events are more heavily felt by players playing PG, SG, and SF. Net result: lower team ORB rates because fewer players are contesting those boards, BUT a slightly higher interior defender ORB rate on average, because there is less competition for the boards that fall closer to the basket.

Part of the reason why ORB rates are lower is also thanks to the distance shooting PF or C who plays offensively further from the basket. Which isn't Rodman. In today's game, he may be setting some screens a bit higher here or there, but the portion of the floor he's playing on O will be the same.

That's what I was correcting for on the study--looking at the guys who generally aren't playing outside. It's roughly 1950 player seasons worth of data over that period--about 1850 player seasons when you remove the more perimeter-oriented guys who shoot a lot of threes.

And when you look at one particular type of player (interior offensive big), the ORBs have at worst held constant for guys rebounding at the upper level and have actually gone up on average for all players within this type.

[EvanZ]

What you've stumbled across is probably more of a random statistical oddity than an indication that ORB rates are declining for interior players. F-C/C-F/Cs playing 500+ min ( or their 98/99 shortened season equiv) provides some indication of this. ORB rates broken down by ORB era and indexed to 100, here are the average ORB rates for those "bigs" (weighting them by ORB opps for each player).

If you just included those 3 groups, your study may be inaccurate. For example, Dennis Rodman, Reggie Evans, and Kevin Love are categorized under the "F" group, none of the three you mentioned ("F-C/C-F/C"). I'd think it would be important to include these players (and many of the others who show up). Can you address this?

[Chicago76]

Can't say much now. Intoxicated at airport bar, but its accurate. 1-Evans and rdmn are in dif eras and directionally push things the same way. There are always pos classification issues, but as a percent of floor time it's not significant.

2-stuff I know off top of my head: guards from 87 to 11 dropped 4.1 to 2.4. That's 3.4 of a 6 pt decline (multiple for2 g spots). Small forwards dropped abt 2. That's 5. stretch 4s and 5s...who didn't really exist much in 87 were the other 1. Interior o P fwds and centers have no change or slight improvement.

Any b-r position class issues isn't going to be significant.


Hope this helps.

[Chicago76]

One more thing. Your post on orbs on 3s being 40 percent yet o reb rate down. Due to more time post release for players to collapse on basket + long boards on 3s. It's basically 80s reb with longer boards so added chaos. On2s there is less time so perim d and o stays home. More off Rebs for interior o but fewer team orbs.

[EvanZ]

Can't say much now. Intoxicated at airport bar, but its accurate. 1-Evans and rdmn are in dif eras and directionally push things the same way. There are always pos classification issues, but as a percent of floor time it's not significant.

2-stuff I know off top of my head: guards from 87 to 11 dropped 4.1 to 2.4. That's 3.4 of a 6 pt decline (multiple for2 g spots). Small forwards dropped abt 2. That's 5. stretch 4s and 5s...who didn't really exist much in 87 were the other 1. Interior o P fwds and centers have no change or slight improvement.

Any b-r position class issues isn't going to be significant.


Hope this helps.

It does. Your intoxication levels notwithstanding, I think you've made a convincing case here.

[Chicago76]

Glad to help. A few of my brain cells are idling right now, so I'm glad this made sense. I've been trying to figure out a reasonable basis to normalize stats for quite awhile and the more I dig, the more I realize standard linear adj for pace, reb rates, etc just don't work. Beyond the obvious (3s), there are a whole host of differences based upon relative production for a given position/role.