Prediction Error Detective

Why it works

The impulse to increase control under uncertainty loses its edge once it is recognised as an understandable neurological reaction rather than a character weakness.

Psychologist Lisa Feldman Barrett describes a phenomenon she calls “affective realism”: the brain treats its own prediction as reality. When it feels threatened, the situation is experienced as threatening, regardless of whether an objective danger actually exists. A strong reaction is therefore not necessarily information about the situation. It is often information about our own prediction model being disappointed (Barrett & Simmons, 2015).

When a prediction error occurs, the system attempts to close it – either by updating its model or by acting to align the world with that model (Clark, 2013). This second mechanism explains why otherwise strengths-based leaders can revert to control and micromanagement under pressure: the nervous system is trying to close the prediction error with the means available. It is not a relapse into poor leadership.

At a glance

How to proceed

1. Set the frame

   Remind yourself: “I am not investigating whether my feeling is right or wrong. My feeling is real. I am only investigating what it is responding to.”

2. Choose a situation

   Think of a specific situation that is still tangible – a meeting, email or piece of feedback that triggered a strong reaction.

3. Work through the four detective questions

   What exactly did I expect in that moment? What actually happened? Where precisely is the difference? And the crucial question: Is what I feel a real threat – or a deviation from my expectation?

4. Allow time for the final question

   Most people answer “threat” too quickly because a prediction error feels exactly like one physically. Ask: “What would change if I treated this as a prediction error rather than as truth?”

5. Choose an action

   Either consciously update your model (“I know more now; this changes my expectation”) or, if a real threat exists, respond appropriately and deliberately rather than reflexively.

Reflection questions

Give these questions deliberate time – ideally in writing, alone or after the exercise.

1. What exactly did I expect before the reaction began?

2. How do I notice physically that a prediction error has occurred?

3. Is my initial reaction information about the situation – or about my model of it?

4. If I did not have this reaction, how would I assess the situation instead?

5. Whose expectation was disappointed – my own, or one I unconsciously adopted?

6. What would a colleague see who viewed the situation without my history?

7. Have I experienced this type of reaction before in another similarly unclear situation?

8. What happens if I do not act on this impulse immediately?

9. What would help me update my model here instead of acting?

10. Is what I feel a real danger – or unresolved uncertainty?

What people often discover

The most common insight is: “I did not react to the situation. I reacted to my own expectation.” Many people subsequently recognise a recurring reaction pattern – the same irritation appears in different situations whenever a particular kind of expectation is violated. The exercise also frequently reduces self-reproach: understanding that the reaction is an understandable system process allows people to judge themselves less harshly.

For everyday use

Use it as a brief “detective check” immediately before reacting in a meeting; the four questions can be run through internally in under a minute. Established teams can even develop a shared vocabulary: “Is this a prediction error?” can become a de-escalating, relieving question in tense situations.

Variants