Light sessions vanish from load totals, so the week that felt heaviest can register as the easiest one.
A recovery session logged as zero load is a lie in the spreadsheet
Forty minutes of pool work, some bike, some mobility. Ask the monitoring system what that cost and the answer, in a lot of set-ups, is nothing at all. It was a recovery day. Recovery days do not go in the load column.
That is a convention, not a fact, and it distorts everything downstream. Total weekly load is the denominator for half the ratios anyone calculates, so quietly removing sessions from it changes those ratios without changing the training. A week with four hard sessions and three recovery sessions and a week with four hard sessions and three days off produce the same number. They are obviously not the same week. One of those athletes has been on his feet for seven days and one has not.
The effect compounds when a squad increases its recovery volume, which is what usually happens when the schedule congests. More travel, more pool sessions, more soft-tissue work, more time in the building. The recorded load stays flat or drops. The chronic load, the rolling average people use as the marker of what a body is prepared for, drifts downward because the light days are feeding it zeroes. Then a normal hard session arrives and gets flagged as a spike against a baseline that was deflated by all the work nobody counted.
I am not arguing that a mobility session equals a match. I am arguing that it is not zero, and that the difference between not much and nothing is the difference between an average that tracks reality and one that slides.
Part of the reason for the zero is practical. Nobody wants to hand out rating forms after a stretching session, and the GPS units stay in the drawer for pool work. So the sessions hardest to measure are the ones assigned the value that requires no measuring. The missing data becomes a number, and the number is the most convenient one available.
A flat nominal figure per attended session would beat it. Two arbitrary units for anything on the schedule, applied consistently, at least keeps the week honest about how many days had something in them.
Arbitrary and consistent is a real methodological position. Zero is not neutral, it is a claim, and it is usually wrong.

