Performance and arousal are related, and the relationship has been studied since 1908; what changes under load is not quality alone but which patterns take over.
Why it matters when the plan changes
A plan is not tested on a calm Tuesday. It is tested at the point where two priorities collide, a deadline moves and the information is incomplete. Reading how someone works under easy conditions describes a situation the plan will not encounter. Traits show most clearly in weak, ambiguous, high-stakes moments, where plans are won or lost. The decisive question is what the plan demands at the moment pressure peaks.
The tension is that this is a mechanism with converging evidence rather than settled law. The relationship between arousal and performance has been studied since the early twentieth century and is not simple, and the finding that stress shifts control toward habitual patterns has been questioned by at least one preregistered replication. A score read alone predicts weakly; the signal sits in the interaction between the person and the situation. It is strong enough to build on and not strong enough to state as proven.
In practice
A role is designed around consultation and consensus and works well for two quarters. A regulatory deadline compresses the timeline and the same role now requires decisions made with a third of the usual information. Nothing about the person changed. What the arrangement demanded at the moment pressure peaked did.
Evidence
The relationship between arousal and performance has been studied since the early twentieth century and is not a simple one.
Yerkes and Dodson law, Wikipedia (2026)A score read in isolation predicts weakly; the signal sits in the interaction between the person and the situation.
Predictive validity, Wikipedia (2026)
What it cannot tell you
Behaviour under pressure describes what a pattern produces at load; it does not describe what the same person does under easy conditions, nor does it explain why the pattern exists. It is silent on skill, motivation and the specific demand of a role, so a reading only becomes useful once matched against what that role actually requires under load.
How Atlas reads it
Behaviour under pressure is the primitive the instrument measures, because a profile in isolation is inert and behaviour read against a demand is not. The question is never who someone is in the abstract but whether the behaviour the plan demands, at the moment pressure peaks, is behaviour the person and the team can naturally carry. Where it is not, the fix is the plan, the role or the support.
Questions
Plans are tested under load, not calm conditions, so general description covers situations the plan will not meet. The Yerkes and Dodson law, describing the arousal-performance relationship since the early twentieth century, shows why: what matters is whether the behaviour a role demands at its hardest moment is behaviour the person can carry naturally.
No, and that framing is exactly what is avoided here. Pressure narrows attention and makes behaviour more patterned. That is a statement about conditions and their effects, not a claim about a person's worth, character or what they will be like in five years.
The Yerkes and Dodson law has framed the arousal-performance relationship since the early twentieth century, and narrowed attention under load is well supported, but at least one preregistered replication found the shift toward habitual control less reliable than first reported, so it stays a mechanism with support, not settled fact.
The behaviour is relatively stable; what it produces is not. Changing the demand, the sequence, the support or who holds a decision changes the outcome without requiring the person to be different, which is why the recommended moves sit in the arrangement.
It can. The research on derailment describes natural strengths overshooting when load rises: decisiveness becoming unilateral, thoroughness becoming delay. That is a statement about the match between a pattern and a demand, and it is read that way rather than as a flaw.