Decision-making glossary
Base rate
A base rate is how often something happens across a relevant group of similar cases, before you consider the details of the case in front of you. If 20% of comparable projects finish on time, 20% is the base rate. Good forecasts start from it and then adjust. Ignoring it is a well-documented judgment error.
In a 1974 Science paper, Amos Tversky and Daniel Kahneman described an experiment in which people judged whether someone was an engineer or a lawyer from a short description. Told the group held 70 engineers and 30 lawyers, or the reverse, they gave essentially the same answers either way: the description crowded out the base rate. Given no description at all, they used the base rate correctly.
In practice, the base rate answers the question “how does this usually go?” Before estimating how long a migration will take or whether a hire will work out, look at what happened on similar migrations or hires. This is sometimes called taking the outside view.
The hard part is choosing the comparison group. Too broad and it ignores real differences; too narrow and there are too few cases to count. A team that records its own decisions and outcomes builds its own base rates over time, and those are often more relevant than industry figures.
In Decize: Decize scores forecast skill against the base rate of a workspace’s own resolved forecasts: positive means better than always predicting that rate. See how →