One of the major success stories of the NBA's statistical revolution, has been the league's gradual relinquishing of the Mid-Range shot, in favor of the 3-Pointer. An average 3-Point attempt will yield more points-per-possession then its Mid-Range counterpart, and intuitively, team's are shooting more 3's then ever. However, I believe its still not enough, and that team's could significantly increase their Offensive-Prowess by trading in even more Mid-Ranger's for 3's.
The first reason teams could stand to shoot more 3's, has been well known in the analytic community for years. Relative to Mid-Range jumpers, 3-Point attempts have a significantly higher chance of being rebounded by the offense (as seen by CounttheBasket's heat-map). Similar results have been found by Kirk Goldsberry.
Now that's nice, but one of the counter-arguments has always gone something like this. The "average" 3-Pointer may be more valuable then the "average" Mid-Range shot, but if a drastic change in shot-selection occurred, the conditions that produced the above results may change. In short, teams would be trying to force up increasingly contested 3-Point attempts, while discarding open Mid-Range shots. And, as we all know, the best shot in basketball is an open shot. Right?
Now that's nice, but one of the counter-arguments has always gone something like this. The "average" 3-Pointer may be more valuable then the "average" Mid-Range shot, but if a drastic change in shot-selection occurred, the conditions that produced the above results may change. In short, teams would be trying to force up increasingly contested 3-Point attempts, while discarding open Mid-Range shots. And, as we all know, the best shot in basketball is an open shot. Right?
Wrong. Apparently, the worst shot in basketball is an open Mid-Range shot, with both contested 3-Point and Close attempts being more effective. True, this seems to change when the shot-clock winds down to 4 seconds or less, and these league averages are obviously dependent on the quality of player taking the shots, but wow. This graph (via Krishna Narsu) is a somewhat radical shift in what Basketball-Heads consider a "good shot".
The league-average FG% on long-Mid-Range jumpers (16-23) in 2013, was 38.3%. When taking into account the value of a 3-Pointer (3 points vs. 2), that's equivalent to 25.5% shooting from the 3. Meaning, even if shooting more 3's resulted in a significantly lower 3-Point%, as long as you were shooting above 25% (which 289 NBA players did last year), the offense wouldn't suffer. When taking into account the benefits of Floor-Spacing and Offensive Rebounding differences, the break-even point could be even lower.
That's all for now. Teams are shooting more 3's then ever, and on the surface, it doesn't seem to be enough.
The league-average FG% on long-Mid-Range jumpers (16-23) in 2013, was 38.3%. When taking into account the value of a 3-Pointer (3 points vs. 2), that's equivalent to 25.5% shooting from the 3. Meaning, even if shooting more 3's resulted in a significantly lower 3-Point%, as long as you were shooting above 25% (which 289 NBA players did last year), the offense wouldn't suffer. When taking into account the benefits of Floor-Spacing and Offensive Rebounding differences, the break-even point could be even lower.
That's all for now. Teams are shooting more 3's then ever, and on the surface, it doesn't seem to be enough.