Equivalent facts can lead to different choices depending on how they are framed.
More quotes0.506127 becomes 0.506.
The computer ignores the missing digits. Over time, the forecast becomes dramatically different.
This is the problem with demanding confident predictions from unstable systems: the uncertainty is already present before the analyst starts looking clever. Extra decimal places in the final answer cannot recover an unmeasured starting condition.
Commercial forecasts often disguise this with ceremony. A few large decisions have uncertain timing, an early reaction may alter demand, one dependency can change several later dates—and yet the final total arrives alone, dressed as a commitment. When reality diverges, the analyst is accused of getting the number wrong. Sometimes they did. Sometimes the number was answering a question the system made impossible.
Ranges help, though a lazy range can become a shrug with two endpoints. The valuable work is sensitivity: which assumptions move the result, how much, and what new evidence would change the action? A wide uncertainty around an irrelevant variable is pretence. A clear warning that one early signal could alter the entire outcome is operationally useful.
Forecasting still matters. Decisions under uncertainty improve through estimates. The distinction is between prediction as preparation and prediction as corporate fortune-telling. One supports action. The other provides a future number so somebody can be blamed in the present.
The butterfly is mostly innocent. The manager insisting on a single-point answer from a chaotic system has less of an excuse.