A base rate is how often something happens across all comparable cases. It is the single most useful and most neglected number in forecasting, because a specific compelling story feels more informative than a general frequency and almost never is.
A test for a condition affecting 1 in 1,000 people is 99% accurate. Someone tests positive. Most people estimate the probability they have the condition at around 99%. The correct answer is about 9%. In 10,000 people there are 10 true cases and roughly 100 false positives, so a positive result is a true case about one time in eleven.
Applied to markets: a pattern that "predicted the last three crashes" needs the denominator. How many times did the pattern occur without a crash? A signal firing 40 times to catch 3 events is not a crash predictor, however impressive the three hits look when displayed alone.
Common belief
"This indicator called the last four major turns."
What is actually true
Presented without the false positive count, this is unfalsifiable. The relevant figures are how often it fired, how often it was right, and how it compares to the base rate of the event. Almost no popular indicator is presented with those three numbers, and the omission is the finding.
This is why VILIQ stores every prediction with its outcome definition fixed in advance, and reports calibration with the sample size attached. A hit rate without a denominator is not a measurement, and a model that only shows its hits cannot be evaluated at all.