Gravity is real and we measure it terribly

The pull a shooter exerts on a defence is obvious on film and nearly invisible in every public number.

Everybody accepts that a dangerous shooter drags defenders out of position. Ask for a number and the conversation gets vague very quickly.

The usual proxy is teammate efficiency with and without the player on the floor, and it is a weak instrument. It is confounded by every other change in the lineup, it needs an enormous sample to settle down, and it cannot distinguish between a shooter who bends a defence and a passer who simply makes better decisions. It tells you something happened. It cannot tell you what.

Tracking data should have solved this and mostly has not, because the obvious measurements are the wrong ones. Average distance of the nearest defender sounds meaningful and is not, since a defender standing two feet away doing nothing useful reads the same as one who is genuinely attached. What you would want is something like the defender's willingness to leave, measured as the time taken to help and the depth of the help, compared against what that same defender does against a lesser shooter.

The reason I care is that this shapes roster decisions. Teams pay for gravity constantly, in contracts given to shooters whose own scoring numbers do not justify the money, on the theory that the effect on everybody else makes up the difference. Sometimes it does. Sometimes the team has paid a large sum for a player whose reputation exceeds his current pull, and there is no reliable way to tell those two situations apart from the outside, which is a strange gap for a sport this heavily measured.

The interior version is even less well captured. A big man who genuinely cannot be handled one on one exerts the same kind of force on a defence, arguably a stronger one, since the help must come from closer and the rotation behind it is longer. Nobody calls it gravity. It gets called an inefficient possession.

The same physical effect, described in two ways, depending on where on the floor it happens and which side of an argument the writer is on.

I would settle for one honest metric that treats both the same, and I do not expect to see it.