Last year, I did an analysis of how online magazines were represented in the results of the Locus Poll and was looking forward to comparing this year’s results to last. The July issue is out and I jumped right to the numbers.
The first thing I see is that online magazines are well-represented in the magazine results: Subterranean (5th), Jim Baen’s Universe (8th), Strange Horizons (9th), Ansible (13th), Clarkesworld (14th), SFSite (15th), Fantasy Magazine (16th), SFRevu (21st), and IROSF (22nd). Locus counts only eight online magazines, where I see nine. My bet is they considered Subterranean a print magazine.
However, the next thing I see really bothers me and completely invalidates any year-to-year analysis I had planned:
“Results were tabulated using the system put together by webmaster Mark Kelly, with Locus staffers entering votes from mail-in ballots. Results were available almost as soon as the voting closed, much sooner than back in the days of hand-counting. Non-subscribers outnumbered subscribers by so much that, in an attempt to better reflect the Locus magazine readership, we decided to change the counting system, so now subscriber votes count double. (Non-subscribers still managed to out-vote subscribers in most cases where there was disagreement.)”
They changed the vote counting system after the polls closed. If they were so concerned about the results reflecting reader opinion, why allow non-subscribers the chance to vote in the first place? Doing something like this makes it seem like they were unhappy with the results and put a fix in. Given their long-standing reputation, I’m sure that wasn’t their plan, but what were they thinking?
I think this action significantly hurts the credibility of the poll and the Locus Awards. By their own admission, changing how they counted the votes took the Best First Novel Award away from Patrick Rothfuss and handed it to Joe Hill. That’s just a horrible way to lose and a horrible way to win.
I’d love to get my hands on the raw data and see how things really shook out.