Decision fatigue is real. The explanation we have been using for two decades is not. The brain does not deplete like a battery. It conserves like an economist — shifting from expensive, careful thinking to cheap, automatic thinking when the cost of continued effort exceeds the perceived reward. The distinction changes what a mental performance coach does about it.
The chief executive made the decision at four in the afternoon. She later described it as the worst decision of her career — a hire she knew was wrong within a week, based on an interview she could barely remember conducting. She had spent the morning in three consecutive board-level negotiations, followed by a working lunch with investors, followed by two hours of contract review. By the time the candidate sat down across from her, she was — in her words — “running on fumes.”
I hear versions of this story regularly. The leader who makes a consequential decision at the end of a day filled with smaller ones and, in retrospect, cannot explain why they chose as they did. The diagnosis is always the same: decision fatigue. And the explanation, until recently, was always the same: the brain runs on glucose, willpower is a finite resource, and when the resource is depleted, decision quality collapses.
That explanation is elegant, intuitive, and — according to the most rigorous research of the past three years — substantially wrong.
The rise and fall of the glucose model
Roy Baumeister’s strength model of self-control dominated psychology for nearly twenty years. The proposition was straightforward: self-control operates like a muscle. It fatigues with use. Glucose is the fuel. When glucose is depleted by effortful decisions, subsequent decisions deteriorate. The model generated hundreds of supporting studies and entered popular culture through books, TED talks, and corporate wellness programmes.
The first crack appeared in 2016, when a massive multi-lab preregistered replication — led by Martin Hagger and Nikos Chatzisarantis — found no reliable evidence for the ego depletion effect. The study used the standard experimental paradigm: deplete participants with one self-control task, then measure performance on a second. The result was null. The depleted group performed no differently from the control group.
The finding was contested, and the debate has continued. But in 2025, Junhua Dang and colleagues conducted a new multi-lab collaboration using a more intensive depletion manipulation — longer, harder initial tasks designed to produce genuine cognitive fatigue. This time, the depletion effect appeared. The finding suggests that ego depletion is real under conditions of sufficient intensity but does not emerge from the mild manipulations used in most previous studies.
The implication is not that Baumeister was entirely wrong. It is that the model was too simple. The brain does not deplete like a battery running from full to empty. It conserves like a system managing scarce resources — shifting strategies when the cost of continued effort exceeds the expected return.
The conservation model — what actually happens
A 2026 integrative review in Frontiers in Cognition synthesised the evidence and proposed a framework that fits the data better than the depletion model. Decision fatigue, the review concluded, emerges when individuals shift from effortful, controlled decision-making to less effortful, heuristic-based processing. The shift is not a breakdown. It is an adaptation.
The distinction matters enormously for coaching practice. If the brain is depleting, the intervention is rest — stop making decisions, eat something, wait for the resource to replenish. If the brain is conserving, the intervention is different: change the perceived value of the decision, change the cost of the effort, or change the environment so that the default option is a good one.
In my work, this reframing has been transformative. The executive who made the bad hire at four in the afternoon did not need a snack. She needed a different decision architecture. The consequential decision — who to bring into her leadership team — should not have been scheduled at the end of a day of cognitive labour, not because her glucose was low but because her brain had already shifted into conservation mode, defaulting to faster, less careful processing. The solution is not biological replenishment. It is structural redesign.
What conservation mode looks like in practice
The shift from controlled to automatic processing produces specific, identifiable patterns that a coach can learn to recognise.
The first is default bias. A person in conservation mode is more likely to accept the default option — whatever requires the least cognitive effort to select. This is why studies of judicial decisions find that parole boards grant fewer paroles as the day progresses: denial is the default, and granting parole requires effortful deviation from it. The decisions are not morally different at ten in the morning and four in the afternoon. The processing mode is different.
The second is simplification. Complex decisions are reduced to simpler ones. A leader evaluating a multi-dimensional proposal begins weighting one or two salient dimensions and ignoring the rest. The analysis narrows. The trade-offs that careful thinking would have surfaced are invisible to the simplified processing.
The third is emotional amplification. When cognitive control resources are being conserved, emotional responses — which are metabolically cheaper than analytical reasoning — play a larger role in the decision. This is why arguments escalate in the evening, why interpersonal conflicts feel more severe at the end of a long day, and why the email that seemed reasonable to draft at nine at night seems reckless in the morning.
What a mental performance coach does with this
The conservation model gives coaches a practical framework that the depletion model did not.
Sequence the day by cognitive demand. The highest-stakes decisions should occupy the morning or the first hours after genuine recovery. Not because the brain has more glucose, but because the conservation shift has not yet occurred. The executive’s hiring decision belonged at nine in the morning, when controlled processing was still the brain’s default mode.
Design decision environments, not decision stamina. Rather than training clients to “push through” fatigue — which the conservation model suggests is fighting the brain’s own resource-management system — design the environment so that the default options are acceptable. Pre-commit to criteria. Create decision frameworks that reduce the cognitive load of each individual choice. The same principle that prevents burnout applies to decision architecture: sustainability comes from design, not from willpower.
Build recovery into the architecture, not the margins. A ten-minute break between meetings does not reset the conservation shift. A genuine transition — physical movement, change of environment, a task that uses different cognitive resources — does. The coaching intervention is to protect these transitions as non-negotiable elements of the schedule, not as luxuries that get cut when the day fills up.
Monitor for the signatures of conservation mode. When a client begins defaulting to binary choices (“should I do this or not?” rather than “what are my options?”), simplifying complex situations, or making decisions faster than the situation warrants, the conservation shift is likely operating. The coaching response is not to push through but to pause — to flag the pattern, defer the decision if possible, and return to it under conditions where controlled processing is available.
The honest complexity
The conservation model is better than the depletion model, but it is not complete. The research is still evolving. Individual differences in susceptibility to decision fatigue are substantial and not fully understood. The role of motivation — whether the person cares about the decision — moderates the effect in ways that complicate simple structural recommendations. And the interaction between sleep quality, physical condition, and cognitive resource management means that decision fatigue cannot be addressed in isolation from the broader physiological context.
What the new research does provide is a more accurate and more useful framework for understanding why good thinkers make bad decisions at predictable times. The brain is not broken. It is managing itself. The coach’s job is to understand the management system well enough to work with it rather than against it.
The chief executive did not need more willpower. She needed better architecture. That is the difference the new science makes — and it is a difference that changes the daily practice of everyone who works with human performance under real conditions.
ALTIUS Note This article belongs to the ALTIUS editorial library: evidence-based essays on Human Performance, Leadership and Society.
Selected Research Foundations
Dang, J., et al. — Revisiting ego depletion: evidence from multi-lab collaborations Collabra: Psychology (2025).
Hagger, M. S., et al. — A multilab preregistered replication of the ego-depletion effect Perspectives on Psychological Science, 11(4), 546–573 (2016).
Grignoli, N., et al. — Clinical decision fatigue: a systematic and scoping review with meta-synthesis Family Medicine and Community Health, 13, e003033 (2025).
Frontiers in Cognition — An integrative review on unveiling the causes and effects of decision fatigue to develop a multi-domain conceptual framework (2026).
