Tuesday, January 18, 2005

One More Rebounding Table and Conclusions

Brophog over at phog.net suggested that I also look at the number of offensive rebounds Kansas would expect to allow given their season average defensive rebounding rate. This eliminates some of the outliers caused by the variety of the quality of the opponents' opponents and sets a baseline for normal distribution of offensive rebounds allowed by Kansas, thus highlighting the games they allow an unusually high percentage of offensive rebounds by their own standards. Very clever, and again, I did not think of this on my own.

For clarity's sake, the columns indicate opponent, opponent's expected offensive rebounds, Kansas's expected offensive rebounds allowed, actual offensive rebounds by the opponent in the game.

TEAMexpORexpOR allOR

Looking at things in the context of Kansas's expected offensive rebounds allowed, the Colorado and Texas A&M performances look even worse (Kansas came up 3 defensive rebounds short against the Aggies and 5 against Colorado) and their good performances in the context of the other team's expected offensive rebounds look more average (+3 rather than +5 against St. Joe's, +1 rather +4 against Nevada, +2 rather than +5 against UW-Milwaukee).

Since we're talking about five or fewer rebounds per game, I hesitate to put much too much emphasis on this for the simple reason that I don't know how much a rebound is worth.

The only game for which I have data for second chance points is the Colorado game (and I have no idea how accurate that data is, unofficial stat and all...). The Buffaloes converted 21 offensive rebounds into 20 second chance points. In that game, Kansas gave up four or five extra points due to poor defensive rebounding.

If I knew Colorado's season average ratio of second chance points to offensive rebounds, I could determine if that was a good or bad performance by them in converting those chances. Unfortunately, I'm not aware of any consistent, exhaustive source for such information.

Other folks on phog.net made the observation that a lot of the offensive rebounds allowed by Kansas were long rebounds. A quick study of opponent's three-point attempts and shooting percentages categorized by the (assumed) quality of Kansas's defensive rebounding performance.

Games where Kansas saved at least one offensive rebound by both measures: Vermont, St. Joseph's, Nevada, Pacific, UW-Milwaukee, Georgia Tech. In those games, the opponents combined for a 40.5 eFG% and took 20.2 three-pointers per game which accounted for one-third of their total shot attempts.

Games where Kansas allowed the expected number of offensive rebounds: TCU, S. Carolina, Kentucky, Iowa State. In those games, the opponents combined for a 42.8 eFG% and took 21.2 three-pointers per game which accounted for 36% of their total shot attempts.

Games where Kansas allowed at least one more offensive rebound than expected by either measure: Louisiana Lafayette, Texas A&M, Colorado. In those games, the opponents combined for a 39.6 eFG% and took 22.3 three-pointers per game which accounted for 37.6% of their total shot attempts.

If one assumes that three-pointers, on average, create longer rebounds, the above suggests that Kansas does in fact struggle most to gather those long rebounds (see Bill Self's emphasis on guards doing a better job of rebounding) as they force their opponents to attempt more long shots. Presumably, those long offensive rebounds, though they prolong the possession are not put-backs are not being converted at near the rate of two points per offensive rebound.

All in all, what looked to me like poor defensive rebounding does not appear to be a big deal.

No comments: