- Quick polling data for today. The Poughkeepsie Post reports on some polling data saying that in NYS Senator Hillary Clinton is the favorite amongst Democrats for the nomination in 2008 while Former NY mayor Giuliani had virtually the same support 1/3rd of those polled, as Arizona Senator John McCain. Though it should be remembered McCain has been in the presidential limelight a lot more these past few months.

Trailing in the polling was outgoing NYS Governor Pataki who only had 7% of the support. In a match up between Clinton and Giuliani the poll has Clinton by a wide margin, however two earlier polls showed different numbers. This is coming off a landslide election win for Clinton so in the months to come those numbers could well even out.
As for some polling issues there is actually a pretty interesting debate happening about polling at the moment which strikes at the heart of big media vs. Web and independence. Many traditional pollsters and newspaper critics and the like aren’t very happy about the way online poll watchers like electoral-vote.com and pollster.com tabulate their data.

Basically these online sites use public polling data, take the newest polls, average out the numbers, then use somewhat similar methodology to make a prediction about winners. It’s a pretty basic, somewhat common sense and logical way to do things. It also isn’t a) the only way and b) in my opinion it demystifies polling data to a level most people can follow as opposed to how it is often presented in more traditional press.

Most important of all it is giving traditional poll watchers a bit of a run for their money.

In 2004 and 2006 I watched Electoral-Vote.com closely. In 2004 the site showed what most pollsters were predicting, it was too close to call for the President but it was leaning toward Kerry. Voting irregularities aside most of those public polls were flawed in that election because they were looking at likely voters and missing unlikely voters who turned out in a big way, so the site went along with conventional wisdom and was wrong in its prediction.

The problem was, most people on election day looked at the exit polls, looked at the public polls and were wrong too, including those people in the traditional, old school media. So it is hard to just fault the Web sites, everyone was looking at the same thing and finding the same conclusions.

This year the predictions at that site were pretty dead on. The site made a lot more accurate predictions then many in the media and than Karl “The Architect” Rove. I had only become familiar with Pollster.com and others after the fact but as the vote tabulations started coming online I compared much of it to what was on Electoral-Vote.com and sure enough a good chunk of the data was dead on with the rest pretty close.

This is going to be an issue that will undoubtedly continue to play out over the next two years and especially in 2008. So consider this post an introduction to the issue and one I will probably look into more.

Carl ‘The Numbers Guy’ Bialik has an article in the Wall Street Journal with some insight into the issue and into who was right and wrong. When I find more written I will post about it.