A problem in traffic analysis: Estimating the number of unheard U-boats

During the Summer of 1943, some 300 radio interceptions revealed the presence of about 100 German submarines in the North Atlantic. A decision-maker might well have asked an electronic warfare analyst, ‘But how many haven't we heard?’ Brian McCue examines three applied statistics approaches to this operations research and intelligence question, noting the ways in which the validity of each approach can be checked in light of the data such analyst would have had – as well as against the now-known true number.