A base rate is how often an outcome happens in general, the background frequency before any specific evidence about a particular case is added.
Last reviewed 15 September 2025 · Educational, not advice
A base rate is the underlying frequency of an outcome across many similar cases, the rate at which something happens in general before you look at the specifics of one situation. If a certain kind of event has occurred in roughly one year in five historically, that one in five is the base rate, and it is the sensible starting point for any estimate about a new instance of the same kind of event.
Base rates matter because people routinely ignore them. Faced with a vivid story or fresh detail about a specific case, it is natural to lean on that and forget how often the outcome occurs in the first place. Psychologists call this base rate neglect. The result is estimates that swing too far on thin evidence and away from the steady anchor that history provides.
The disciplined approach is to start from the base rate and then adjust. You begin with how often the outcome happens across comparable cases, then move that figure up or down for the specific evidence in front of you, in proportion to how strong that evidence really is. Strong evidence justifies a large move. Weak or anecdotal evidence justifies only a small one.
Choosing the right reference class is the hard part. A base rate is only as useful as the set of past cases you compare against, and reasonable people can disagree about which cases are truly similar. Too broad a class washes out real differences. Too narrow a class leaves too few cases to learn from. Naming the reference class openly is part of an honest estimate.
In prediction markets a base rate is a check against being swept up by a single story. Before trading on a confident view, asking how often this kind of thing actually happens grounds the estimate in evidence rather than narrative. The market price is itself partly a crowd estimate, and comparing it to a sober base rate is one way to test whether your own view is really better.
Suppose a market asks whether a particular kind of event will happen this year, and over the past forty years it has happened in eight of them. The base rate is eight in forty, or twenty percent. If a contract trades at fifty cents, the market implies far higher than the base rate, so you would want specific evidence strong enough to justify that gap before agreeing with the price.
Illustrative only. Numbers are examples, not a quote or a prediction, and exclude fees.
A base rate is a starting anchor, not a final answer, and the wrong reference class can mislead as easily as ignoring base rates entirely. It improves estimates but never removes uncertainty or the risk of loss. Stake only what you can afford to lose, never to chase a loss, and never on borrowed money. If it stops feeling like a free choice, step back. In the United States you can call or text the helpline on 1-800-GAMBLER or visit ncpgambling.org.
It is how often an outcome occurs across many similar cases, the background frequency before specific evidence about one case is considered. It serves as the sensible starting point for a probability estimate.
It is the common error of ignoring the background frequency and leaning too heavily on a vivid detail about a specific case. It leads to estimates that move too far on weak evidence.
Start from the base rate, then adjust up or down for the specific evidence, in proportion to how strong that evidence is. Strong evidence warrants a large adjustment, weak evidence only a small one.
Because a base rate is only as good as the set of past cases you compare against. Too broad a class hides real differences, too narrow a class gives too few cases. Naming the class openly keeps the estimate honest.
The rules change fast. Get the changes that affect you, plain and current, not tips.
Independent. Every claim dated and sourced. No platform pays for its place.