Decision Theory
Why Your Brain Lies to You About What You Actually Want
Expected utility theory — the bedrock model of rational choice — assumes you know what you want, but the deepest problem with your decisions may be that you don't.
The Idea
At its core, expected utility theory is elegant: when facing a choice under uncertainty, multiply the value of each possible outcome by its probability, sum those products, and choose the option with the highest result. It's the mathematical skeleton behind how economists, policymakers, and AI systems model rational behaviour. The trouble isn't the maths — it's the first assumption buried inside it: that you have stable, well-ordered preferences you can actually measure. You don't. Research by Daniel Kahneman and Amos Tversky demonstrated systematically that people's stated preferences reverse depending on how options are framed. More unsettling still: people are notoriously bad at predicting how much they'll actually enjoy an outcome — what psychologist Daniel Gilbert calls 'affective forecasting error'. We overestimate how happy a promotion will make us. We underestimate how quickly we'll adapt to loss. So the 'utility' we're maximising is a fiction built from cognitive shortcuts, social comparison, and emotional noise. The real insight isn't that expected utility theory is wrong — it remains indispensable as a normative benchmark, a standard we can measure ourselves against. The insight is that the gap between how we actually choose and how this model says we should choose is where almost all the interesting action in human life takes place. Noticing that gap is, in a very real sense, the beginning of wisdom about decisions.
In the World
In the early 1950s, the French economist Maurice Allais presented a pair of gambling scenarios to a room full of distinguished economists — people who professionally championed expected utility theory — at a conference in Paris. In both scenarios, the mathematically consistent choice, under the theory, was clear. Almost every economist in the room, including several who would later win the highest prizes in the field, chose inconsistently. They preferred certainty so strongly in one scenario that they violated the very axioms they defended in lectures. Allais published this as a paradox, and it landed with the force of a small intellectual bomb. The economists weren't stupid; they were human. What Allais had exposed was a deep, systematic aversion to regret — specifically the agony of choosing a gamble and losing when a sure thing was available. Their brains were quietly adjusting the utility calculations to avoid a particular emotional outcome: the feeling of being the fool who gambled and lost. This is not irrational in any simple sense. Regret is painful and real. But it means the model these economists were using to describe choice had a ghost in it — an emotional variable the equations couldn't see. Allais never won the prize he thought he deserved for this. He remained convinced, until his death in 2010, that mainstream economics had never fully reckoned with what he found in that Paris room.
Why It Matters
Most of us move through decisions — career pivots, relationship choices, how to spend a Sunday afternoon — as if we already know what we want and are simply calculating how to get it. Expected utility theory, even in its broken human form, is the invisible architecture of that assumption. What this lesson quietly challenges is the confidence we bring to knowing our own preferences in the first place. Before you optimise, it's worth asking whether what you're optimising for is actually what matters to you, or whether it's a proxy inherited from social expectation, a misremembered past experience, or a future emotional state you're badly predicting. This isn't a reason for paralysis — decisions must still be made. But it's a compelling reason to slow down slightly before the big ones, to notice the framing effects at play, and to hold your preferences a little more lightly. The Allais paradox suggests that even experts, reasoning carefully, import hidden emotional logic into supposedly rational choices. Becoming aware of that isn't a weakness — it's a form of clarity most people never develop.
A Question to Ponder
When you imagine the best possible outcome of a decision you're currently weighing, are you imagining how good the outcome will actually feel — or how good you imagine it will feel to have chosen correctly?
Get a new one of these every morning.
Start learning with Thinkable