Views from Medina Road

I have just 2 more stories to read this week to have read all the short fiction on this year's Hugo ballot; then I will submit my electronic votes. I've read 2 of the novels, and parts of 2 others; 1 I don't care to read. Can't say I've read through any of the Related Book nominees, though I own 5 of them (out of 6) and have certainly read portions, or browsed significantly, in 4 of them.

I've seen 3 of the films, none of the dramatic short forms.

In case either of you noticed, I didn't post a 'New Books' page this past weekend (though I've been doing quite well for three months now posting a page each week), simply because there was nothing to list. Nothing came in the mail; I did a Barnes & Noble visit, and saw nothing new. This week a bunch of stuff has appeared: Robert Borski's Gene Wolfe book came in the mail, as did Datlow, Link & Grant's Year's Best F&H (thank you publishers); my orders for new books by Jack Dann, Kage Baker, and China M came from Amazon; and today at Borders I saw the new Haldeman, Irvine, Stewart (which I've now ordered), and Theodore Judson. So there'll be lots to list this weekend.

I am going to Worldcon (I've gone every year since 1988), arriving Wednesday evening, leaving Monday afternoon. I'm on no panels (I didn't volunteer and they didn't ask) and have no specific appointments aside from a Locus business meeting Friday morning followed by the official presentation of the Locus Awards Friday at 11 a.m., itself followed by a champagne reception. If not wandering around the con, I can probably be found hanging out at the Locus table in the dealers' room.

I still plan to derive some interesting statistics about Locus Poll voting patterns, and will endeavour to do so in the next week or so...

Here're new thumbnails of what I've been reading lately (if it works).

Despite the fact that, very occasionally, I get an irate email such as one that came over the weekend, set in pink font and reading in part
 

we are an aging population-that means that an evergrowing percent of the population is getting older and,consequently, not as interested in change for its' own beau self.if you are going to futz around endlessly with your emag's layout,give us some kind of a guide or template to your layout staff's latest little brain-fart,please! after spending 20 minutes deciphering the latest yellow brick road,i am usually pissed at you for weeks. [sic]

 
--Despite this rare kinda thing, I did in fact tweak the layout of the homepage yesterday and today. Partly I wanted to revive the 'highlights' box at the top of the center column, to keep visible substantial content that has scrolled down below more ordinary posts; and partly to try to clean up the clutter of various archive and info links that were accumulating at the top of various columns. You can compare the new layout to the previous layout on this now archived page.
 
Only the placement of various links has changed, not the underlying structure, so I hope it won't require 20 minutes for anyone to adjust to the change. I'm still not entirely happy with the placement of the Google search box, but I haven't been able to think of anywhere better to put it. (I moved it from the left column so that Future events and Blinks could remain as high as possible at all times.) One other change I made is to place my own name on the homepage, as a result of doing a recent Google ego-search and finding my ISFDB entry, and my own long-long-neglected Compuserve webpage, coming in at the top of the rankings; even a 2001 review I wrote for Locus magazine and reprinted on the website comes in ahead of any indication that I'm responsible for the entire Locus Online site itself. Ironically, the Locus Online homepage does turn up in the results, because of
 

King, Karen Joy Fowler, Ray Bradbury, R. Scott Bakker. ... H. Greenberg, Alexander Potter,and Mark Tier; novels by ... John Kessel & James Patrick Kelly, Larissa Lai

 
Isn't that funny?

An anonymous reader asks,

Can you describe what constitutes a "valid ballot?"

I've heard a lot of noise over the years about "unexpected" results being discarded. The Sharon Lee and Steve Miller controversy comes to mind.

Reply: There are roughly three types of *invalid* ballots. The most common is when someone filling out the online ballot inadvertantly hits the 'enter' key (rather than the tab key) and causes their partially filled-out ballot to be submitted via email. They realize their error, complete their ballot, and submit again. This results in 2 email ballot submissions from the same person within a couple minutes of each other. In these cases, the latter is counted as valid; the earlier (partially completed) is discarded.

Another kind of invalid vote is where someone deliberately votes for the same item more than once in the same category. The rules say not to do this, but some voters try it anyway (sometimes 5 times in a category!). Needless to say, we here at Locus Online have sophisticated database queries to detect such multiple votes -- since we can't personally examine every one of the 5 or 6 hundred ballots -- and all such invalid votes are discarded. We could be much harsher than this, and discard the entire ballots of those who commit this violation, but we don't--we count one vote per item per category, in the lower/lowest position the multiple voter placed it, and disregard the others.

A variant of violation #1 is where a voter submits a second ballot several weeks after submitting their first. Sometimes they even fill in their name and subscriber number both times--as if to flaunt their violation. (Well, maybe they forgot they'd voted before.) Other times they don't, but the pattern of votes (along with survey responses) is obvious. Again, we could be harsh and disregard *both* submissions, but we don't; we disregard obvious repeat votes, but count the first (or most complete) from each voter.

The final category of suspect ballots is where numerous submissions are received within a short period of time that all vote for the same item--often only a particular item, leaving the rest of the ballot blank. This presumably occurs when a notice is posted about the Locus Poll ballot on some writer's site, whose fans are prompted to link to the online Locus Poll and vote for their favorite writer's currently eligible book. To some degree, this happens almost every year. In most cases, it doesn't affect the final outcome. This happened a few years ago described as the "Sharon Lee and Steve Miller controversy", though in that case the online poll was a parallel poll, and those votes didn't affect the official Locus Poll results whether or not those votes were included or not.

Another anonymous reader asks,

What's the print run of LOCUS?

Answer: roughly 9000. Less than 10K. Locus Online, meanwhile, has 5-6K unique visitors per day, or 30+K unique visitors per week.

I said back at the beginning of this year's Locus Poll process that I would post some observations about voting patterns -- how many voters vote in how many categories for how many items, etc. -- without discussing the actual results themselves. Now that all the results are in, and the winners have been announced (and published), I'll return to this theme -- again, without implying anything about who or how many voted for what.

There were 629 valid ballots. On the ballot, there were 14 categories in which voters could vote for up to 5 items per category -- a total of 70 votes.

How many of the 629 submitted all 70 votes?

Answer: 29. Out of 629. Not very many.

Average? 29.78. I.e., of all ballots, of all 629, on average the number of votes in those 14 categories was 30 out of 70. So lots of people vote, but not in every category.

More stats over the next few days. If you have any particular query you'd like to pose, that I don't supply the answer for, let me know.

Wow--Blogger recognized me for some reason; I didn't have to re-enter my username and password. Is it finally working as advertised?

Sitting home alone on a Friday night updating the site and catching up on email. That's what we dedicated fen do. What do people do on Friday nights who don't have websites to update? Sometimes I wonder. (This is an old CNB joke.)

--Actually, my partner Yeong has flown off to China, to the city of Qingdao (aka Tsing Tao, like the beer), for two weeks to visit his father, cousins, and siblings. During which time I'll catch up on this and that--the website, reading, and my ongoing project to expand the Awards Index and database to include all the Canonical tallies and chronological profiles I've been planning for, well, years now. An ultimate guide to the top 100 novels, novellas, novelettes, short stories; the best in each category in every year. Soon; any year now. On CD ROM.

Cheryl Morgan has already blogged about the peculiar results of the Wooden Rocket Awards, though she didn't note the result that struck me -- Best Print Publisher to Tor Books. Tor may be the best SF publisher, but its site is notably lacking in detail and currency. I note this as someone who compiles descriptions of Notable New Books regularly (every week now) for the Locus Online website, looking for links to excerpts and descriptions; for my time, Baen's is by far the best publisher site, with descriptions and generous excerpts for every single new title. Eos is a fair second place; Warner and Bantam are in the running. But Tor... Now that I check it, they do have content for a couple current titles, but.. but... (And I really should notify them that they have a long obsolete link to the Locus Poll index.)

Which raises the question of how valid any awards results are. A question which, as a compiler of awards results, I am careful not to probe. And as a recipient of a Wooden Rocket Award (they sent me a certificate last year), I would be foolish to dispute.