Muppet IRV
As promised, a link to the Muppets Use IRV instant-runoff-voting demo. It's a little muddy in places, but a reasonably good primer on the subject.
Also worth looking at for those interested in the Hugo: George Flynn's article "Hugo Voting: Let’s Look at the Record (Again)," which shows all sorts of interesting stats. For example, only 11%-18% of people who've registered in advance for WorldCon in the past ten years have voted in the Hugos, which casts doubt on the notion that people don't vote because it's too expensive to do so. (Though some percentage of the advance registrants register after the Hugo deadline. But even if half of advance registrations came after the deadline (and I'm pretty sure the real number is much less than that), that would still mean that under a third of eligible pre-registered members usually vote in the Hugos.) One conclusion of that article: "Not only is the overall number of Hugo voters decreasing, but of those who do vote, fewer and fewer are voting for the categories other than Dramatic [Presentation]."
There's some discussion there of special categories added by various Hugo committees: "[Most] experimental categories have flopped badly, and didn’t even make it to the final ballot." Looks like the Best Web Site category stacks up quite well by comparison with other special categories. When Best Music was a special category, for example, "only one nominee receiv[ed] more than 7 [nominating] votes."
Another interesting comment, regarding IRV and how few voters seem to make use of its full potential:
[E]ven the winners (except for Roger Rabbit in 1989) received votes on less than 80% of the ballots voted in their categories; for the Campbell Award in 1989, the winner [received a vote] on only 51% of the category ballots. Whether out of ignorance or apathy, it is clear that many voters fail to have as much effect on the Hugo results as they are entitled to.
Down near the end, there are some stats on how often the nominee with the highest number of first-place votes fails to win (on average, it seems to happen in about two categories each year), and on how often the nominee with the highest number of nominations tends to win (on average, about half the categories). And then there's an interesting philosophical section on how well the Hugo voting system does or doesn't work.
Yes, I ought to be editing and reading submissions instead of writing this. Just taking a break. I'll go get back to it.