Context: Wald wrote wartime memoranda on aircraft vulnerability while working in Columbia University's Statistical Research Group. His work used damage recorded on surviving aircraft to help analyze losses. CNA later reprinted these memoranda.
Context: The video uses a simplified wings-and-tail damage pattern to explain the problem. The cited records do not establish that exact pattern as Wald's measured dataset. Returning aircraft show survivable damage, while lost aircraft are absent from inspection.
Context: Visible damage alone leaves out aircraft that were lost. In the AMS postscript, SRG director W. Allen Wallis recalls military interest in protecting areas with the most observed hits. That supports the general preference, not an exact first plan naming wings and tail.
Context: Wald used statistical models to estimate risks despite missing damage records for lost aircraft. The video's spoken armor recommendation is a simplified explanation, not a verified quotation. His memoranda do not record that exchange with military officers.
Context: When hit exposure is comparable, a vulnerable area can show fewer hits among survivors because damaged aircraft are lost. Wald's worked engine example includes surviving planes with engine hits, so an engine hit is not always fatal. Its figures are explicitly hypothetical, not measured loss rates for all wartime bombers.
Context: A sample containing only survivors can hide important missing cases. In this example, examining returning aircraft leaves out those that were lost. This is a selection problem, not a claim that the observed holes were recorded incorrectly.