Base Rate
Also called: base rate, unconditional probability, benchmark rate, what happened anyway.
How it is read
Establishing the base rate first is what turns a percentage into a finding. Without it, the percentage turns into nothing. Say a market closed higher on 55% of days. A rule that is right 60% of the time has then added five percentage points. Quoted without that benchmark, the same 60% reads as a strong result. The comparison has to be made on the right population. That is the set of days the rule could actually have fired on. A rule that only triggers on unusual days must be compared against the base rate for those days. It must not be compared against the rate for all days.
What it does not tell you
A base rate is itself an estimate. It has a sample size and a period behind it, and it moves. A figure computed over one regime does not necessarily describe another. It is also easy to compute against the wrong population. That produces an apparent edge. The edge is only the difference between two kinds of day. A rule beating its base rate is a necessary condition for being useful. It is never a sufficient one. The gap still has to clear the cost of trading it.
See it in an experiment
Unfamiliar market — Let a hidden mixture of regimes unfold. Decide whether any construction fits the observations.
Related concepts
Base Rate First: What the Market Did Anyway
A rule that works 60% of the time against a market that rose 55% of days has found five points, not sixty.
Open the lesson