Richard Hatch wrote:If I understand your statistics, Chris, the estimated chances for an adult male in the USA in 1902 owning a copy of EATCT are no greater than 1/2000 (and probably much smaller). It follows (I think) that the odds of an adult male in the USA in 1902 owning 2 copies are (1/2000)x(1/2000)=1/4,000,000. If we apply those odds to the population of Cook County (2 million) we get an expectation of only half a person in Chicago in 1902 owning 2 copies of the book. Ed Vernello owned at least 2 copies in 1902, since he took the trouble to advertise it in the The Sphinx in November 1902, unlikely to be something he would have done had he only had one copy. In fact, he likely a dozen or more copies, to justify his effort in advertising it. What are the odds of that? Do I think Vernello was the author? No. A person of interest, who may have known the author? Sure!
Richard, maybe I didn't make this clear enough in my analysis. I am not primarily calculating the statistically expected chance to find somebody like Gallaway. I am calculating the statistically expected chance to find Erdnase. In order for your case of two copies of EATCT to work you would have to tell us why Erdnase had to had two copies and not only one. My analysis basically is for anybody with one or more EATCT, because the chance to owning two copies, as you correctly calculated, is much smaller and thus we can basically ignore it. You can of course include it, but it would only change the third decimal somewhere so it does not matter.
The problem with your earlier E.S. Andrews analysis was a similar one. We can't assume that Erdnase had that name. There is really no evidence for it. However, all the evidence I am using must also apply for Erdnase. He certainly ordered his book printed at McKinney and he most likely had one or more first editions. Perhaps the only questionable bit is the 'W' in the name, but both Andrews and Gallaway have that so it cancels in the direct comparison.