Since politics is on topic now, I thought I'd share a bit of what we've
discovered lately in our research. This is off-the-record, please. It's
so preliminary that I don't want anyone taking it too seriously.
Our goal was to see how the most influential people in on-line financial
forums plan to vote. I'm not going to say too much about how we measure
influence, but it's the sort of thing I've talked about here in the past --
how the community responds to postings, how widely a person posts (one's
sphere of influence is obviously limited by how widely one participates in
discussions), and how closely one's pattern of postings correlates to
external patterns, such as stock prices.
The most striking thing about a survey of the most influential people is
that while those who said they'd vote for Gore and Bush had virtually
identical influence rankings in their communities, there were big
differences for Nader's supporters and the undecided. (No other candidates
showed up in significant numbers in our survey.)
Nader's supporters were much *less* influential, while undecided voters
were much *more* influential. Although the survey showed a small
preference for Bush (not surprising among people who can afford to trade
stocks), the undecided block, at 12 percent, was clearly enough to sway
the election either way... and given that they are among the most
influential in the whole population, this election looks impossible to
predict. Really, really impossible, at least with any forecasting
technique available today.
My apologies to those who want the underlying numbers on this -- I'm not
prepared to divulge details yet.
Nick
