The parole study is the one everybody cites and it deserves both its fame and a caveat.
Danziger, Levav and Avnaim-Pesso analyzed a thousand-odd parole decisions by Israeli judges and found that the proportion of favorable rulings started at around sixty-five percent at the beginning of each session, declined steadily to near zero by the end, and jumped back to sixty-five percent immediately after a food break. Three sessions a day, three sawtooths.
The caveat: the effect size is implausibly large for a pure depletion explanation, and later work has argued that case ordering — unrepresented prisoners scheduled later, cases grouped by prison — accounts for some or much of it. The specific study is contested.
The general finding is not. Sleep deprivation degrades the use of feedback in risky decision-making and increases risk-taking. Hunger and fatigue shift judgment toward defaults and heuristics. Acute threat arousal — limbic activation outpacing prefrontal regulation — narrows attention and impairs the comparison that evaluation requires. Every one of these is well established independently.
The defensive implication is that your judgment is not a constant, and that the variation is predictable.
Which is why practiced operators schedule the ask.
The car dealership visit that runs four hours before the finance paperwork appears. The negotiation session that continues past midnight. The sales conference that starts at seven and puts the close at the end of day three. The relationship argument initiated at two in the morning. Kissinger's thirty-four days of continuous shuttling, chapter 207. Every interrogation in Part 4.
None of these are coincidences of scheduling. Depletion is the instrument, and the ask is placed in the window.
The protocol is a check performed before any commitment, and the mnemonic from addiction recovery is as good as anything: HALT. Am I hungry, angry, lonely, tired?
Add two: am I under time pressure, and have I been in this room for more than two hours?
If the answer to any is yes, the decision is deferred. Not reconsidered — deferred, as a rule, without evaluating whether this particular decision seems fine. The whole difficulty is that a depleted judgment is not experienced as depleted; it is experienced as clarity, because the reduced complexity feels like decisiveness.
The counter-response is to say it plainly: I don't make decisions at this hour, we'll do it tomorrow. The other party's reaction is diagnostic, and it is chapter 304's test.
This counters Law 8.