APBRmetrics

[MVP]

I just had an idea: I think it would be interesting to find out how likely players are to get fouled out if they have a certain number of fouls during the first quarter..or something similar.

Or maybe a different approach than that, but finding out if it helps to bench a player if he got a few fouls too quickly could be an actionable insight.

Maybe somebody has done something like this before? Any other suggestions/ideas for this one?

[Mountain]

How likely players are to get fouled out if they have a certain number of fouls during the first quarter could be studied from the actual history of game logs if you wanted to pull a big enough sample size.

Or. alternatively, it could be estimated by using personal foul rate per minute and finding the straight, all things being steady and equal, probability of accumulating enough additional fouls to foul out from whatever first quarter foul total by running each minute of additional play and generating foul or no foul at the given probabilities. Presumably most but not all players cut back on foul generating actions as danger of fouling increases and maybe it affects opponent behavior and referee behavior too.

Comparing the probability model to the historical data could help analyze these things and perhaps estimate behavorial influences in aggregrate. Maybe there is a way to separate them, by looking at opponent rates of driving or referee frequency of blowing a whistle under different conditions and in different game parts.

I don't instantly call any major personal foul studies but maybe someone else will. If not, this could be an area where you could break some further ground.