The
popular vote share market at Iowa Electronic Markets and Rasumussen's
presidential tracking poll, in my estimation, do the best job of projecting Bush's share of the two-party popular vote. (Bush's share of the Rasmussen poll = percent for Bush/(percent for Bush + percent for Kerry.) Here's how the numbers looked on the eve of the second Bush-Kerry debate:
Bush's showing in IEM's popular vote share market yields 337 to 388 electoral votes. His share of the Rasmussen poll yields a slightly lower estimate: 319 to 369 electoral votes. (The conversion of popular vote share to electoral votes is explained
here (see method 3).)
