We are often more certain than the evidence really allows.
More quotesForecasting has a crooked payment system. Confidence is rewarded today; accuracy is audited months later, usually after everyone has forgotten the bet.
The expert announces a crisp future on television. The result arrives quietly on a wet Wednesday and receives no booking. Misses become “directionally right”. Dates slide. Assumptions discovered after the event are promoted to original caveats. A prediction can fail so thoroughly that it returns as thought leadership.
Nuance travels badly. Doom books interviews. “There is a 35% chance” sounds feeble beside a man promising catastrophe before lunch, even though the first statement may be the more intelligent and useful one.
Expertise still matters. It narrows possibilities, spots mechanisms and improves probability estimates. Demanding certainty about politics, markets or next quarter’s sales is asking a map to remove the weather. It is childish, but delivered in a suit.
The remedy is closer to bookmaking than clairvoyance. Put the odds, timeframe, conditions and base rate beside the prediction. Keep the original ticket. When the event passes, settle the bet using the meaning printed at the time—before hindsight’s legal department rewrites the terms.
This changes who looks impressive. The loud prophet loses the advantage of unlimited fresh starts. The careful forecaster who updates with evidence stops looking indecisive and starts looking calibrated. Someone who knows the edge of their model becomes more valuable than someone who carries identical certainty through every result.
The point is institutional memory, not public humiliation. A prediction is useful when it improves the decision made under uncertainty. If nobody records the odds or checks the outcome, the organisation has not bought foresight. It has rented confidence by the hour.