The model researchers actually use

If you read the actual literature rather than the summaries of it, one framework comes up again and again: I-PACE — Interaction of Person, Affect, Cognition and Execution, developed by Matthias Brand and colleagues¹.

It is worth knowing because it is strikingly unlike the popular story, and because it makes better predictions about what to do.

The four parts

Person. What you brought with you — temperament, mood and anxiety vulnerability, early experience, loneliness, impulsivity. These are predispositions, not destiny, and they explain why two people with identical use have different outcomes.

Affect and Cognition. A trigger arrives — internal (bored, low, tired) or external (a cue, an hour, a notification). What happens next is not automatic: it depends on how you read the trigger, and on what expectations you hold about relief. The model gives this step real weight, which the popular story does not.

Execution. Executive control — the capacity to override the pull. Depleted by tiredness, stress and being drunk, which is why the same person decides differently at 1pm and 1am. See The braking system: why self-control fails at 1am.

Reinforcement. The outcome trains the loop. Relief is learned, the cue becomes more potent, and the cycle strengthens — this is where the model connects to incentive sensitisation².

Why this matters practically

The dopamine story has one intervention in it: endure. I-PACE has at least four, because it has four places to intervene.

• Change the trigger environment → Change the room, not yourself
• Change the reading of the trigger — recognising an urge as a wave rather than an emergency → You do not have to fight it. You have to outlast it.
• Protect execution — sleep, and decisions made in advance → Almost all of this happens at night
• Interrupt reinforcement — which is what sitting through an urge without acting actually does

That is not a coincidence. Almost everything in this site's practical articles maps onto a component of this model, because the model is a decent description of the thing.

What it is, and is not

I-PACE is a theoretical framework, not a finding¹. It organises a large amount of scattered evidence into something that can be tested; it does not itself prove anything. Parts of it are better supported than others, and a critic would reasonably say it is broad enough to accommodate most results.

But it is what the researchers use, and it is notably more modest than the version aimed at the public. There is no claim in it that a substance is damaging tissue on a schedule. There is a claim that a person, a moment and a habit interact — which is both less dramatic and considerably more actionable.