"Released: March 18, 2004LinkThis Zogby Interactive poll is a test and reflects only the opinions of those registered online users who have chosen to participate. The results cannot nor should be assumed to represent the opinions of U.S. voters in general or the public as a whole. We hope to be able to provide statistically reliable interactive polls in the near future."
His 1998 and 2002 records beg to differ with you about "at your peril".
And his 2000 tracking polls were not the online polls he is doing now- and I still suspect that you have the wrong numbers for a few of those states although I do suppose it is possible Newsmax does.
And I stand by my assertion- his 2000 accuracy was nothing to write home about. He completely blew California. He blew New York. His polls were making wild swings beyond the margin of error that no other pollster was seeing. He had more states outside the margin of error than can be explained by random chance-- and that was in a year he was doing well. In 2002, he missed most of the key races. Not one or two- most.
I went back and doublechecked the numbers and I did post the wrong ones for Wisconsin. The 46%-46% tie was Zogby International. This was the tracking poll:
Zogby: Bush 47%; Gore 44%
Actual: Bush 46%; Gore 46%
Once more, I haven't any problem with dismissing the methodology out of hand. It is not widely accepted and so is certainly subject to discretion. I haven't any problem with accepting it either. My personal approach is to include everything with the assumption that outlier polls will be superceded by other polls quickly enough. If Zogby gets it wrong, then that's fine by me. It's his reputation that will suffer accordingly..
In my personal approach to things, Kerry is currently winning with about 320-327 EVs to about 211-218 EVs for Bush. The president's reelection campaign would be wise to behave accordingly rather than follow a faith-based approach....