actually its closer to 250,000 after war, almost all in soviet hands, are the russians angles now?
I think my post on this wasn't very clear, so maybe I should clarify a little.
A Canadian historian, James Bacque, wrote a book called "Other Losses", in which he analysed the records of German prisoners in captivity after the end of the war. In the records for prisoners held by the US Army, the data on the eventual fate of the prisoners contains a column labelled "Other losses", which gives a figure of about 660,000. Bacque concluded, despite clear evidence to the contrary, that this column referred to prisoners who had died in captivity; in fact, contemporary sources make it clear that "Other losses" refers to prisoners who, due to their youth or extreme age, were seen as incapable of having committed any war crimes, and therefore were simply discharged without further action or documentation. Bacque went on to extrapolate additional numbers for the "Other losses" column, again with little or no justification, to produce a final figure of a million German soldiers dying in US captivity. To work around the inconvenient fact that these individuals were not in fact missing after the supposed mass deaths, he then postulated that the USSR had underestimated its figures for German POW's by the same amount, an assertion repeatedly denied by the USSR. A general consensus exists among historians that Bacque's work is incompetent and draws entirely false conclusions. Buchanan appears to have taken it at face value, which has more to say about Buchanan's competence as a historian than Bacque's; Roberts also appears to have accepted the figure, which is no great testament to his reviewing skills.
Put in simple terms, it's a classic example of the conspiracist circle-jerk. One conspiracy theorist dreams up an untenable theory, a couple of other conspiracy theorists back it up, and the rank-and-file, unable to distinguish between three individual opinions and the overall consensus of historians worldwide, treat it as established fact. This is the inevitable result of selecting sources on the basis of prejudice rather than reliability.
Dave
