An NBA Player’s Comfort Zone

Both on player level and on team level, we see that defensive three-point shooting percentages against you are basically random[1. This is a great reason why I am skeptical of some of the defensive numbers that are floating around. It is good for Chris Paul that people shot badly from long range against him this year. But that’s most probably luck and not that his hand is more distracting than another persons hand]. This means that players take most of their shots when they feel comfortable about taking them. Thus, you can rather try to make them feel not comfortable enough to take the shot. We also saw that taller defenders had on average a further distance from the shooter than smaller defenders, without having a higher shooting percentage against them. This indicates that most players feel less comfortable to shoot against taller guys at the same distance (duh).

Given that understanding, if we look at defender distances and height difference between the shooter and the defender, we should expect to see a height difference specific plateau for shot frequency. Meaning that, as most shots happen when the shooter feels comfortable, most shots happen when the distance between player and defender is high enough for the player to be in his comfort zone[2. It is basically a more binary approach to Krishna’s more sophisticated aXPPS.] This should let us see at which defender distance a player is really happy to shoot[3. Note: I’m of course talking about average shooters. There is no defender distance where Nick Young is not happy].

To identify that plateau, I took all three pointers and filtered out all the shots that happened late in the shot clock or game clock. I binned height differences, so that all height differences from -10 to -6 counted as -8 (defender is 8 inches taller than shooter), -6 to -2 as -4 and so on.

You can clearly see the plateaus for which most shots happen. More shots happen with a defender distance between four and five feet (23 percent of all “non-heave” threes) than for a defender distance of less than four feet (16%), showing that around four feet is where most players feel like they could let one fly. And the plateaus are also shifted according to height differences. It is interesting to note that you cannot see this effect clearly without binning shots by height difference, as the combined defender distance looks more like a peak around five feet. The difference between the -8 inch plateau and the +8 inch plateau is around one foot. Given that both groups have most likely a lot of players with a 6 inch difference (because of the binning), a foot sounds pretty reasonable for a 45° shooting angle[5.

I love it when a plan comes together

.].