I see via links on other sites that the complete Hugo Awards
voting breakdown, along with the list of
nominees and honorable mentions, are available online. (I understand these files were provided on CD ROM by the con committee to members of the press, but apparently
Locus Online was not on their list.) Thus there's no need to describe the results in detail; I'll just point out a few things that strike me as significant.
Novel-- Despite predictions that Martin might win the award, as a sort of cumulative honor for his series to date, his book came in last. Many people predicted Stross would win, and he came in second, but Wilson's lead was solid from the first round. It's interesting, tracking the 'no award' totals in each round, that while 17 voted for 'no award' as first place, and 18 others ranked it ahead of the second, third, and fourth placing novels, some 43 others ranked 'no award' ahead of Martin's novel.
Novella-- Link tied with Willis in the first round of counting, then dropped behind as second-place votes from eliminated nominees were added to the totals. Several commentators said that, however much they liked reading Link's story, it was difficult afterwards to say just what it was about, whether it was SF or fantasy or what. That may have turned off some voters, but I would guess that Willis' story benefited from her personal popularity to win, as more than one good but not great previous Willis story has done.
Novelette-- I thought Doctorow's online and personal popularity would swing the result his way, even though I liked Beagle's story better. As in the novel voting, there were a chunk of voters who chose 'no award' ahead of the nominee that eventually placed last, the story by Michael A. Burstein.
Short story-- Lanagan led in the initial count of first place votes, then dropped behind as others were eliminated. In this category the jump in 'no award' votes came ahead of the Resnick and Burstein stories. I'd actually thought Resnick would win -- it's the sort of sentimental story Hugo voters have gone for in the past, and Resnick is a popular presence at conventions, and nothing else on the ballot seemed really strong (except for Lanagan's story, but it was non-sf/f).
Related Book-- Gary Wolfe's book, interestingly, had the most nominations (29) in the category; Wilhelm's had only 19. I have no explanation for Wilhelm's win, except to wonder if the number of Clarion students who've learned from her, or perhaps simply the general availability of her book (compared to Wolfe's or Ashley's), tipped the votes her way.
Editor-- Hartwell led all the way. Why did voters finally come around his way this year and not before? Perhaps Dozois' absence from the ballot (he did place 7th in number of nominations) made the category more of a mix-up than it has been in many years. Or perhaps, to give voters a bit more credit, the repeated observations year after year that the same candidates win in certain categories year and year, despite other worthies, finally sunk in. [Update: Or more likely the ongoing debate about splitting the category raised awareness of who should win; see Patrick Nielsen Hayden's comment below.]
Semiprozine-- Almost the same candidates as last year (with Emerald City replacing The Third Alternative). Ansible's win last year was due to the local UK vote, then, presumably? This year Locus led all the way.
Campbell Award-- As everyone predicted, Scalzi won, and he led commandingly from the first round. A better writer than Roberson or Monette or Swainston or Bishop or Sanderson? Maybe, maybe not, but none of them write popular blogs.
Interactive Video Game-- This year's special category on the nomination ballot, deleted from the final ballot for 'lack of interest', drew no more than 13 nominations for any particular item. (The next lowest category was 28, for Resnick's short story.) The 13 were for 'World of Warcraft'. The only thing I nominated, 'Myst V: End of Ages' got only one other nomination.