[Goose]

I suspect both of the above result in exaggerated conclusions about how everyone should learn to be a 3-point shooter and no one should ever shoot midrange.

I mostly agree. I think the "selection bias" pretty much cancels itself out, as it applies to all kinds of shots. The degree to which a shot is contested however is very important. The problem is that usually the closer you get to the basket, the more probable it is someone will contest the shot. That may make a 3-10 foot shot in general harder to make than one somewhat further out. There is obviously a rule-induced discontinuity at the three point line. Now, if you can reliably get open looks at 3-10 foot shots, the expected value probably goes way up. The problem with a mixed strategy for the offense is that they have to be proficient at the various different strategies, just like the defense does. So it is always a balancing act.

I think analytics fails to take into consideration the fact that no two teams, and no two games, are alike. What is true for all of basketball on average may not be true for any given team. If you can't shoot the 3 very well, trying to do so isn't going to help much. OTOH, if you see Lynn Shackelford running to the baseline, you better guard him, no matter his distance from the rim.

[Goose]

I suspect both of the above result in exaggerated conclusions about how everyone should learn to be a 3-point shooter and no one should ever shoot midrange.

I mostly agree. I think the "selection bias" pretty much cancels itself out, as it applies to all kinds of shots. The degree to which a shot is contested however is very important. The problem is that usually the closer you get to the basket, the more probable it is someone will contest the shot. That may make a 3-10 foot shot in general harder to make than one somewhat further out. There is obviously a rule-induced discontinuity at the three point line. Now, if you can reliably get open looks at 3-10 foot shots, the expected value probably goes way up. The problem with a mixed strategy for the offense is that they have to be proficient at the various different strategies, just like the defense does. So it is always a balancing act.

I think analytics fails to take into consideration the fact that no two teams, and no two games, are alike. What is true for all of basketball on average may not be true for any given team. If you can't shoot the 3 very well, trying to do so isn't going to help much. OTOH, if you see Lynn Shackelford running to the baseline, you better guard him, no matter his distance from the rim.