Essentially, if you look at page 19, you see that the race was 40% to 31% to 29% for Perata, Quan, and Kaplan. Kaplan's voters (who hadn't over voted, under voted, or "exhausted" their ballot--which I assume means they ranked 2nd and 3rd candidates that had already been eliminated in prior passes) overwhelmingly supported Quan: Quan picked up ~18,800 votes that way to Perata's ~6,400 (see page 20).
Perata said on election night "I don't understand how ranked-choice voting works" (
http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2010/11/10/MNKV1GADKC.DTL&feed=rss.bayarea ). May be that was an off hand remark, but he needed a campaign to convince more people who didn't back Quan or Kaplan to make him their 2nd choice. Maybe he tried and failed. In any event, he came up about 1,000 votes short. (E.g., if Perata have convinced more of Joe Tuman's supporters to back him over Quan-Kaplan, he would have won...Tuman's unexhausted passed on ballots were fairly split between Perata and Quan, see page 18.)
So the real surprise is that such a consistent preference for Quan among Kaplan supporters existed (the campaign dynamic that produced the result). It seems to me, not having followed the race, that this would point to a savy campaign from Quan, or at least a very similar set of politics (or a shared distain for Perata?) between Quan and Kaplan supporters that their camps understood and ran on.
Doug Hess