28.1.04

OKAY: So last night wasn't such a good night for my budding interesting in political methodology.

Or was it?

Follow me here. One of my big theses was that Kerry, Clark and Dean were fighting for the same supporters. Look at my Clark, Kerry and Dean predictions again:

Kerry: 41%
Clark: 19% (actually 20% on my model, but was changed for reasons I explained here)
Dean: 16%

which is, by my math, 76%.

The actual vote percentages for the three:

Kerry: 38%
Dean: 26%
Clark: 12%

which is, by my math, 76%.

So yes, my predictions were wrong (I did rob myself of an extra day by declaring on Friday (easier in Iowa as the caucus was on a Monday)), but I feel some vindication that I was right on the only thing I explicitly modeled.

No comments: