This is the website of Abulsme Noibatno Itramne (also known as Sam Minter). Comments here or emails to me at are encouraged... or follow me on Twitter as @abulsme.



October 2004

Electoral Vote Roundup

Being the junkie that I am, especially with things being as close as they are, every day I check a few of the places doing predictions of the electoral college results based on state polls. A pet peeve of mine is any place that even bothers mentioning national poll results as if they are at all relevant to anything. The popular vote doesn’t matter. The electoral college does.

Luckily there are dozens of places doing electoral college analysis, ranging from major main stream news outlets to individuals tracking this on their own. Some of them have really solid methodologies, some have odd methods, some are trying to be unbiased, some are clearly partisan. But they are all interesting.

Anyway, as a reference to anybody who wants to look at this kind of stuff, here are my favorite places to check regularly:

UnFutz: Once or twice a week UnFutz comes out with an electoral college survey. This basically provides links to almost all of the places on the web doing this sort of analysis on an ongoing basis. In addition UnFutz adds meta-analysis, averages, graphs and all sorts of other interesting things aggregating the results of all the others. From here you can look at the overall trends and also click through to each of the source sites to see them in detail. If you want to look at this stuff, start here. In between the surveys UnFutz is also publishing some additional meta-analysis. For the most recent UnFutz survey as of this post, look here. This site is done by a Bush supporter, but tries to have a neutral methodology. He has an interesting method that takes into account not just state polls, but also the national polls, approval ratings, and other things. I now check other places more often, but this is one of the first I found, and I got into a habit of looking, so it is still on my list. The must check site. Not necessarily because they have a better methodology, or their prediction is most likely to be right, or anything like that, but simply because they update every day, and have some cool graphs. They also provide daily commentary on what has changed since the previous days, thoughts on the validity of the results, and that sort of thing. As of today their graph shows just how volatile the electoral college is in a race this close. Since tracking started on this site in May, the lead has changed *SEVENTEEN TIMES*.

After spending time looking at all these sorts of things, including the tracking graphs on several, and seeing the current breakdown of the predictions on UnFutz (20 Bush winning, 15 bush ahead, 8 Kerry ahead, 19 Kerry winning) only one thing is clear… barring something huge happening in the next couple weeks, we definately won’t know the results until (at least) election night. This isn’t one of those where month ahead of time you pretty much knew how it would turn out. Every daily news cycle has the possibility of shifting the results significanty… with winner take all in most states, and so many states split right down the middle and really too close to call… it’ll all come down to how a handful of states end up falling at the very last minute, and that will depend on how the small number of undecideds fall, as well as how the far off the polls end up being from reality.

It should be fun. And of course, it will be even more interesting (although not necessarily fun) if any of the things listed in this article happen:

Florida 2000: The Sequel
(Richard Hasen, Slate)

Nightmare Scenarios

  • No. 1: Litigation Following Voting Glitch
  • No. 2: Litigation Over Whose Vote Counts
  • No. 3: Litigation Over Colorado’s Amendment 36
  • No. 4: Electoral College Woes in Congress
  • No. 5: Terrorist Attack That Disrupts Voting
  • (via

    There are also issues already bubbling about absentee votes and problems there, and various irregularities in voting in various places… none of which will matter at all if one or the other candidate pulls far enough ahead, but if it is close… I suppose it is too much to ask to get fun like 2000 twice in a row????

    Leave a Reply

    This site uses Akismet to reduce spam. Learn how your comment data is processed.