APBRmetrics

I know this sounds sac-religious. Taking emphasis away from rebounding, but the way I see it offensive rebounds are separate from how efficient you are with the ball. I was wondering how other stat-heads would take to the idea of instead of incorporating rebounding into an offensive rating, rating offenses and defenses strictly on how effective they are per "offensive opportunity". Then considering rebounding a completely separate phase of the game where no team is offense or defense. (Similar to special teams in football). I know this idea is a bit abstract just thought I would get others opinions.

Outside his offensive rebounding, Dennis Rodman had no offense, in his Bulls years.
By retrieving 19% of teammates' misses, he made 47% shooters as effective as 53% shooters without him.
Actually, for a team that's protecting a lead, a missed shot followed by O-Reb can be even better than a made initial shot. It gives you more time to kill.

In football: If you line up to punt -- but it's a fake, and you run or pass it -- is that a 'special teams' play? Or is it an offensive play on 4th down? It seems to me, the punt or the FG attempt is still offense; you're on defense when the other team actually gets the ball.

If a defender has tipped the ball and it's loose, and you dive on it and call timeout, you're playing offense. If the scrum is above the rim after a missed shot, it's still offense until the other team has the ball.

I get that there are gray areas between O and D. But the D hasn't finished the job until the ball is secured, usually via the D-Reb.

I believe at least once or twice I've seen offensive ratings presented that were stripped on the rebounding element. Maybe by Evan Z. or Jon Nichols. I think it is fine to look at lots of different cuts, roll-ups and partial roll-ups. Looking at a set of the 4 Factors (raw and / or adjusted) separately is useful too.