causality
How high must the dike be?
It is February 1953, and the North Sea has just come over the Dutch dikes in the night — eighteen hundred dead, a fifth of the country's farmland under salt water. The Delta Commission must now set the height of the new defenses, which means answering a question the historical record barely touches: the magnitude of the ten-thousand-year storm surge, extrapolated from a century of tide-gauge readings. Ordinary statistics lives in the middle of distributions; this decision lives in the extreme tail. Build the inference: the limiting theory of maxima, fitted to annual extremes, with uncertainty honest enough to price concrete against catastrophe. Too low and the sea returns; too high and the nation bankrupts itself on sand and stone.
Who this problem belongs to
The two figures whose methods fit it best, out of 61 in contention.
Fisher, together with L.H.C. Tippett, proved the foundational 1928 theorem establishing that the maximum of a large sample from a wide class of distributions converges to one of only three limiting families, the mathematical bedrock of extreme value theory that any dike-height calculation for a ten-thousand-year storm surge must ultimately rest on. His broader career-long insistence on rigorous small-sample and limiting-distribution theory, developed for agricultural and biological data, gave statistics the exact-inference machinery needed to extrapolate meaningfully beyond a century of tide-gauge readings into the genuine tail. This is precisely the theorem the Delta Commission's actual statisticians built on when they set the new Dutch dike heights, making Fisher's contribution the direct mathematical ancestor of the entire problem rather than an analogy borrowed from elsewhere.
Wald's statistical decision theory, developed during WWII, formalized exactly the problem's second half: choosing an action, in this case a dike height, that minimizes expected loss under deep uncertainty about which outcome will actually occur, with asymmetric and catastrophic costs on either side of the choice. His framework for reasoning rigorously about decisions when the underlying probability of the worst case is itself uncertain, rather than assuming it away, is precisely the decision-under-deep-uncertainty apparatus the Delta Commission needed once the tail probability of the storm surge was estimated. His survivorship-bias insight also cautions against underestimating the true frequency of catastrophic events from a limited historical record, exactly the trap the dike-height problem must avoid. His applications were wartime ordnance rather than coastal engineering, but the mathematics transfers almost completely.
In the mind map
The same ideas, as concepts rather than history — in John's ML knowledge map.
61 figures are scored on this problem. Draw it in a battle to see where you land.