APBRmetrics

[supersub15]

A lot of media people measure team rebounding prowess by using rebounding differential (Tm Reb - Opp Reb).

I found this to be lacking, in that rebounding is always dependent on shooting. For example, Toronto was outrebounded by 3.09 per game last year, the fourth-worst differential in the league. Yet, the Raptors were 12th in shooting with a FG% of .463. But isn't true that the better you shoot, the fewer the rebounds?

Is there already a way to link FG% with rebounding?

[Ben F.]

See this excellent post by Tom Ziller:

[supersub15]

[Mike G]

As noted, teams rebound missed shots by themselves and by their opponents. But teams aren't equal in how many they miss relative to opponent misses. And missed FT, while technically 'available' to either team, are much less likely to be rebounded by the shooting team.

Therefore, I've been assessing team DReb% and OReb% by estimating the % of missed FGA they get. It goes like this:

OReb% = ORFG/(ORFG+OpDRFG)
in which
ORFG = OR - .035*(FTA-FT)
and
OpDRFG = OpDReb - .53(FTA-FT)

These are estimates because I've assumed 53% of missed FT are rebounded by the defense, and 3.5% by the offense. There may be notable differences in how well teams box out and/or attack missed FT.

But there are certainly differences in how often a team shoots FT or puts the opponent on the line. On a given night, your opponent may miss many FT, thus handing you many DReb. This doesn't mean your team rebounded exceptionally well, just that you fouled the right guys a lot.

Every team has a much higher OReb% on FGA than on FTA. Last season, 'traditional' OReb% averaged .271, while from FGA it was .292. Departing from the average .021 difference were the extremes of Phx (.011: they don't foul much) and Orl (.034).

This little tweak would be more accurate if anyone knows the actual OReb and DReb% of missed FT.
_________________
40% of all statistics are wrong.

[Mike G]