APBRmetrics

I think the most overlooked aspect of this debate is the significantly greater quantity of free throws generated by shots in the paint compared to shots from behind the three point line. I see many people comparing the value of three point shots to the value of two point shots by using the metric of points per shot. This is incredibly flawed because someone who is fouled while in the act of shooting does not get credited for a shot attempt. The value of shots in the paint is incredibly high primarily due to three reasons, including the high shooting percentage, high offensive rebounding rate, and free throw attempts generated. It just seems that people forget about the third reason I mentioned because the shot attempts that generate free throws are not registered as official shot attempts, leading to the points per shot values for shots in the paint to be artificially deflated. Superstar players who are especially effective at driving to the rim and generating free throws attempts must shoot a higher percentage from behind the arc than many people expect to justify shooting three pointers. For example, the 3PTFG% that Kevin Durant must shoot to justify 3PTFGA is much higher than the 3PTFG% that Klay Thompson must shoot to justify 3PTFGA.

We have this old article (8 years old) on sample data http://www.82games.com/locations.htm
2 pt shots tend to have both more FTAs and more turnovers. The three types of 3 pointers vary widely in net efficiency but all beat all 2 pointers here except those from low paint area.

Should have something newer for all data using the new databases. Do we?

A simple response would be for someone (Evan??) to figure out the PPP of a possession with at least 1 2PA vs PP of a poss with at least one 3PA.

The problem with this missing piece though is that limiting 3s doesn't mean more shots at the basket either. Nylon Calculus just put up a piece yesterday:
http://nyloncalculus.com/2014/10/22/num ... mid-range/

Likely nearly all those 3s will be midrange shots, which wouldn't have a higher PPP expected at all.

bondom343 wrote:The problem with this missing piece though is that limiting 3s doesn't mean more shots at the basket either. Nylon Calculus just put up a piece yesterday:
http://nyloncalculus.com/2014/10/22/num ... mid-range/

Likely nearly all those 3s will be midrange shots, which wouldn't have a higher PPP expected at all.

Good point. When we consider the fact that, offensive rebounding percentage of threes are higher than mid-range jumpers too, it becomes more important. As a side note, I want to add that turnovers by steal and blocked shots are worse than other TOs.

Anyways, Byron Scott is a bad coach. No coach should say something like that. Every game might bring different situations and players will make terrible mistakes if they try to listen to him. Think about a situation where there is an oppurtunity for an half-open three in the 4th quarter but a good 3-point shooter won't take the shot because Lakers already shot more than 20 in that game.

It's really absurd. You can't limit 3-pointers like minutes.

Crow wrote:We have this old article (8 years old) on sample data http://www.82games.com/locations.htm
2 pt shots tend to have both more FTAs and more turnovers. The three types of 3 pointers vary widely in net efficiency but all beat all 2 pointers here except those from low paint area.

Should have something newer for all data using the new databases. Do we?

Yeah, I should be able to look at this. Would probably just use league average FT% and multiply by 2 or 3 depending on the location of the foul (which Vantage has) to get a PPS. Could then add this in for each location.