This is exactly the kind of thing I was talking about, taken to an extreme. A lot of theorizing about ROC curves and optimal operating point formulae, and absolutely no empirical, qualitative evidence (such as any interviews with actual scammers).

For example, there is no actual sample data provided to substantiate even the premise, let alone the conclusion of this claim:

> In choosing a wording to dissuade all but the likeliest prospects the scammer reveals a great sensitivity to false positives.