Tuesday, July 25, 2006


Like most of you, I started my day by downloading the pdf of the Field Poll and giving it a look over breakfast.

My basic question is: do any of its findings merit repeating?

How does a turnout model where the Republican gets 33% in LA County and 34% in the Bay Area, adding up to an 8-point Republican lead make any sense?

My answer: it doesn’t.

I’d have to guess that this is heavily over-sampled in the remaining parts of Southern California (ie, OC, Inland Empire) and the Central Valley.

Sure, ask 992 registered voters at the Harvest Crusade who they’re supporting, and ya, I’d imagine Arnold does pretty well. Ask 992 registered voters at a Save the Whales rally, and well, probably not so well.

A wise man once told me that a poll is only as good as its sample.

This Field Poll doesn’t seem to be very good.

Remember that when the next one trickles out and shows guys like McClintock and Strickland (still) up. Granted, I think they might actually be, but this won’t be valid validation. At least not in my book.