Hofstadter's Law

Hofstadter’s law: It always takes longer than you expect, even when you take into account Hofstadter’s law.

The line is self-referential by design. Douglas Hofstadter put it in Gödel, Escher, Bach (1979). Hofstadter’s law on Wikipedia collects the wording and ties it to completion-time guesses that keep slipping once the task is tangled enough that “next week” means “who knows.”

He framed it around computer chess. People kept forecasting that a program would reach the world champion in ten years, and when the decade closed, the next ten-year story still sounded plausible. The machine did cross the line in the end (Deep Blue versus Kasparov in 1997), but the early forecasts were the same optimism Hofstadter was teasing, just with new dates.

Software picks up the same shape. The ninety-ninety rule names the tail as a joke about percentages; Hofstadter names the same optimism on a second pass. Integration, review churn, and bugs you only meet in production often do not fit the first spreadsheet row. A pad turns into part of the safe baseline, and the slip comes back larger than before.

The recursion is the point. “Account for Hofstadter” is not a one-time line item you check off. The law says the account was still small. That reads less like a theorem you prove and more like a warning that optimism bias tracks the same brain on the first pass and the second.

What helps is estimating discipline that splits the work, names the risks, and checks the model before you freeze the date. History and peers beat gut on tasks you have not shipped in this exact shape before.