@dlowan,
Just that when I read that the man's use of numbers has been accused of padding, or at least of sharp deviations from the standard estimations, I became a lot more wary. I'd really just taken his numbers, as posted by Thomas, at face value, but apparently they're fairly controversial.
I mean, if he took Conquest's numbers on Soviet deaths, for example -- I read Conquest's "The Nation Killers", and it was expertly written, harrowing research. But yes, I mean, he's definitely on the far end of the scale in some respects -- so it doesn't, to me, give a good impression of Rummel's methodology if he did indeed just adopt Conquest's numbers. You'd want to base your numbers on a comparison and evaluation of available estimates, right, rather than just go with one at the highest end of the scale and work from there? Again, not passing judgment on what the right number
was - I'd want to go back to Anne Applebaum's brilliant
Gulag, for example, which contains a wealth of numbers that might well corroborate Conquest's, just that, yeah -- the Wikipedia para suggests that Rummel used a number that was over twice the standard assessment, and when questioned about it, referred to Conquest. That doesn't seem persuasive.
Also, I admit, it doesn't help that his Wikipedia page gave me the impression that he was very intelligent, but also fairly out of the mainstream, while his own webpage, at first glance anyway, and the one or two random articles of his I now read, gave off almost a crackpot vibe. For example
the article in which he rants on about how US academia is dominated and controlled by leftist forces that indoctrinate students and stifle conservative and libertarian dissent ... did you know that in the humanities and social sciences at US universities, "there are more communists (they call themselves Marxists), than either libertarians or conservatives together"?
(Aside from the Okie-esque contention itself, since when does Marxists = communists?)
Anyway, bla bla etc and this is fairly deep into the weeds. I don't doubt his main finding - that the Castro regime was more bloody still than the Batista dictatorship it replaced. That would certainly fit the pattern of communist dictatorships, unfortunately. None of the above even needs to mean that his numbers are necessarily incorrect at all. There's plenty of scientists and academics who are scrupulous and precise to a fault in their core research and go off on various ill-advised limbs outside it. And the snippet on Wikipedia about his Soviet numbers is too elliptic to say much about.
Just that now, I'm less likely at least to take his numbers at face value. I mean, I obviously lack the know-how to corroborate the numbers myself - so yeah, it's all about how authoritative this or that source seems. Now I'm back to longing for an alternative source to confirm or contest these numbers, as the case may be.