A group threshold describes an athlete who does not exist

Cut-offs built from squad averages sit at a point that may match nobody in the room they govern.

Averages have a habit of producing a creature nobody has met. Take thirty players, compute the mean loading tolerance, draw a line there, and the line describes a hypothetical athlete assembled from parts of everyone and belonging to no one.

This is not a philosophical complaint. It has consequences in both directions on the same afternoon. The threshold is too low for your hardiest players, who spend the season being held back by a rule built partly out of the fragility of teammates, and it is too high for the two or three who are genuinely near their limit, who are handed permission by the same rule. The group cut-off does not fail gracefully. It fails hardest at exactly the ends of the distribution where the decisions matter.

A related trap sits in how the mean is computed. If the squad contains a mix of positions, or a mix of ages, or four players returning from long injuries, the average is a blend of subpopulations rather than a description of a single one. A blended mean can land in a gap where no actual athlete sits, the way the average of two clusters lands in the empty space between them. You get a number that is arithmetically correct and biologically absent.

What should replace it is not another number for everybody. It is a rule for how each athlete's own line gets set and revised: start conservative, increase when the response to the last increase was uneventful, hold when it was not. That is a procedure rather than a threshold, and procedures are harder to put in a policy document, which is much of why they lose to thresholds.

I understand the appeal of the single line. It removes argument. It protects staff who have to justify a decision after something goes wrong, and it can be applied by whoever is on duty. Those are real institutional goods and they are not the same thing as being right about the athlete in front of you.

The test I would apply is simple. Ask whether any individual in the squad has ever had their line moved because of their own history. If the answer is no, the threshold is not a measurement of anyone.

It is a policy with a decimal point on it.