Idea: chaikin statistical plus-minus

[nbacouchside]

so i was playing around with some of bchaikin's different stats (touches, touches / min, %pass, %shot) and i wondered if someone with more programming or excel know how than me (basically any of you, i'm totally useless with that stuff) might think of a way to regress those stats against longterm RAPM to see what their coefficients might be with offensive RAPM.

maybe this isn't possible or maybe not all of bob's stats that he uses are publicly available, but just from looking at it, it seems like players that have a lot of touches per minute would likely be highly correlated with high offensive impact, as they are usually a team's best player. Derrick Rose for instance, when I calculated it out, had over 2 touches a minute in his last two seasons prior to this one, where the Bulls as a team only averaged around 5 touches a minute. So he saw the ball, quite a bit.

[nbacouchside]

oh speaking of bob,

bob, have you ever described the formula you use for estimating zero point possessions? i tried to figure it out based on the description in this article: http://web.archive.org/web/201210311651 ... aikin1.htm

ZeroPtPoss - this is zero point possessions, or how many times a player missed a shot that was rebounded by the defense, missed a FT that was rebounded by the defense, or committed a turnover, i.e. how many times a player was personally responsible for his team not scoring on a team possession...

A player's points scored per zero point team possessions is an attempt to determine a player's scoring efficiency. The logic here is relating a player's points scored to how many times he himself is personally responsible for his team not scoring on a team possesion - i.e. the ratio of a player's points scored to how often the team does not score due to something he personally did (miss a FGA or FT rebounded by the defense, or committing a turnover)...

but unfortunately i got results that seemed very off for the team i was estimating (bulls), so i think i screwed things up... probably in the estimating of how many missed shots or free throws were rebounded by the defense.

[bchaikin]

or maybe not all of bob's stats that he uses are publicly available...

at http://www.bballsports.com the "superdb.exe" free download has an nba stats database that lists all of these numbers (so you don't have to calculate them). you can even add calculated columns (like in a spreadsheet) that subtotal when you subtotal teams, players, etc...

have you ever described the formula you use for estimating zero point possessions?...

nothing sophisticated - this measure of offensive efficiency of points scored per zero point team possession personally responsible for (pts/0ptposs) is simply a player's points scored divided by a denominator of zero point possessions that adds up: his missed shots multiplied by percent of defensive rebounds (typically 0.67 to 0.74 depending on the year), his missed FTs/2 multiplied by a defensive rebounding factor, and his turnovers...

you can use a league wide defensive rebounding percentage for both to get a ball park estimate, or be more accurate using a team's specific defensive rebounding percentage, or even moreso using a missed FTs defensive rebounding percentage (typically high, 80%-85%) and thus a missed FGAs defensive rebounding percentage a bit lower than the others mentioned...