PLAY
You calculate. Just not the way you think
In 1979, Kahneman and Tversky asked people to choose between bets, then reverse-engineered the maths people were using. We do calculate expected value, just not with the real numbers: we overweight small probabilities and feel losses more than gains.
This work won the Nobel Prize in 2002. Now I'm testing if the same maths explains trust.
Sensitivity to outcomes
How we perceive gains and losses. Losses loom larger than gains (loss aversion), and marginal value decreases as amounts grow. So people are risk-averse for gains but risk-seeking for losses.
- Perfect rationality
How much you care about the amount versus simply winning or losing. Lower values = the amount matters less.
λ = 2.25 means a loss weighs 2.25 times more than a gain of the same size. Higher values = losses hurt more.
How we weight probabilities
We don't perceive probabilities linearly. Small chances feel bigger than they are, while medium and high probabilities feel less certain.
- Perfect rationality
1 means probabilities are taken as they are. The further from 1, the stronger the weighting.
Same concept as gains, but for losses. Typically similar to γ+ but empirically slightly different.
Real-world examples
- Lottery tickets: we overweight the tiny chance of winning millions
- Insurance: we overweight small disaster risks
From risk to trust
Prospect Theory works for coin flips and dice
Test it yourself
Pick a bet with the arrows. The Risk game uses the curves you shaped above; the Trust Game has its own sliders.
Trust
−1.77
A stranger decides
Risk
+1.58
A computer decides
The point
Same bet. Different decision. Your parameters change how you see it.
- Parameters = perception. Same numbers, different feelings.
- High loss aversion? Even good bets feel scary.
- Overweight small chances? Changes everything.
- My research: Do trust and risk use different parameters?