The acute chronic ratio shares its numerator with its denominator

This week's training sits on both sides of the fraction, which manufactures a relationship before any athlete has trained.

Divide this week's load by the average of the last four weeks. The last four weeks include this week. That single overlap does more damage to the ratio than anything about the athletes it is computed on.

When a quantity appears in both the top and the bottom of a fraction, the fraction is mathematically bound to that quantity in ways that have nothing to do with biology. Feed random numbers into the formula, numbers generated by nothing but a dice roll, and the resulting ratios will still show structure. Extreme weeks will still cluster at the ends. The distribution will still look meaningful. You can produce an apparently informative chart from data containing no information whatsoever, and this has been demonstrated often enough that it should have finished the argument.

It has not, because the ratio does something valuable in the room: it puts a big week and a normal week into the same sentence and gives a coach a reason to say no. That is worth having. The problem is not that people use it, it is that the specific numbers acquired an authority they were never entitled to. The sweet spot between certain values, the danger zone above another value, all of it derived from analyses in which the same coupling was busy generating pattern out of nothing.

There is a second issue underneath the first. The four-week average is a chronic figure meant to represent preparedness, and it treats a session three weeks ago as equally relevant to today as one from the day before yesterday. Bodies do not weight their history like that. Any decay you choose is arbitrary, which is uncomfortable, but a flat window is also a choice and it is the least defensible one available.

I would still look at the two numbers. Just look at them separately. This week was high; the last month was moderate; that is a jump, and the size of the jump is visible without dividing anything. Nothing is lost by refusing to combine them, and what is gained is that nobody mistakes a coupled artefact for a threshold with evidence behind it.

A ratio feels like analysis. It compresses two facts into one and the compression is where the trouble hides.

Keep the two facts.