Post-Iowa

So how did I do?

Well, it depends. On the GOP side, I'd say I did pretty well. I nailed the top two, calling it correctly for Huckabee and Romney, respectively. Thompson finished barely ahead of McCain, and then Paul did indeed come in fifth, leaving Giuliani at the back of the pack. As for percentages, here's how it came down:

Huckabee +5.6%
Romney -1.5%
Thompson +3.3%
McCain +0.8%
Paul +2.7%
Giuliani -3.4%

As for the Dems, my final prediction was the best. I wrote, "Recent reports have suggested that Richardson and Biden are encouraging their supporters to caucus for Obama, should they not be viable and Kucinich has already officially gone ahead and done this. This could cut into Edwards' lead as second choice candidate. It's hard to say how much this might matter, but...a bigger swing to Obama would put him over the top." I think it's fair to say, that's basically what happened. A lot of the Richardson, Biden and Dodd people ended up caucusing for Obama, and that, combined with the huge turnout, gave him the victory.

That being said, I still think I blew it on the Dem side. My basic (non-second choice adjusted) numbers were pretty far from the mark (relatively). There were two reasons for this. First, public opinion polling is best suited to "one person-one vote" elections, and the Democratic caucuses in Iowa are not that. Second, I should not have given the same credence to polls taken 10 days before the caucuses as I did to those taken one to five days before.

It does appear as if there was some real movement in Obama's direction over that time. I included fifteen polls in my analysis to come up with the following numbers:

Clinton – 28.8%
Obama – 27.13%
Edwards – 25.55%

However, take a look at what happens when you compare the weighted averages of the oldest five polls to the most recent five polls:

Oldest:

Clinton - 30.3%
Obama - 25.5%
Edwards - 25.1%

Newest:

Obama - 29.6%
Clinton - 26.8%
Edwards - 26%

Because I'm using five-poll weighted averages here, I can confidently say that there was some actual movement away from Clinton and towards Obama (how confident? well, statistically speaking, we'd say there is less than a .01% chance that the difference between Obama's earlier support and his later support is zero). This real movement, along with the strong turnout and second-choice breaks lead to Obama's big victory.

Looking ahead to New Hampshire, I know two things. First, I don't have to worry about second choice shenanigans, and second, I'll be more careful to look for trends even within small periods of time. More on that in the next post.