Subadditivity effect
Spelling out the parts of an event makes it seem more likely than the whole.
What it means
The tendency to judge the probability of a category as less than the sum of the probabilities of its mutually exclusive components — so unpacking a single event into its constituents inflates total estimated likelihood beyond 100% of the whole. The mechanism is that each named component calls to mind its own evidence and scenarios, while the packed-up whole leaves much of that detail dormant; explicit parts are simply more available than implicit ones. It is a core prediction of support theory, which holds that probability judgments attach to descriptions of events rather than to events themselves. The effect distorts risk assessment, forecasting, and legal and medical reasoning, since the very act of itemizing possibilities — accident by fire, by flood, by theft — can pump up their summed apparent probability.
Examples
Asked the chance of death 'from a natural cause,' people give a lower number than the sum of their estimates for cancer, heart disease, and other causes listed separately.
Asked what could delay the launch, a team puts 'something goes wrong' at 20%. Prompted about suppliers, testing, staffing, and approvals one by one, their odds now sum past 50%.
A patient rates the chance of any side effect as small, then reads the leaflet listing nausea, headache, dizziness, and insomnia and gives numbers adding to more than the whole.
First described in Tversky & Koehler (1994).