“I thought that we caught a trend and apparently it did not materialize.” Those are the words of Zogby, one day after the election.

His prediction was Kerry 311 electoral votes and Bush 213.
The actual popular vote would be 49.4 for Bush and 49.1 for Kerry. They over sampled from previous dates when some 600 people was used with a 4% margin of error. This last prediction of 955 likely voters was with a 3.2 margin of error.

Unfortunately, he was wrong. Bush won 51.1% and Kerry 48%. In the Electoral College, at this writing Iowa and Nevada are too close to call. However, Bush got 274 to Kerry’s 238.

On final predictions, the most correct was Pew Research who had a 3% spread. Battleground had a 3.3% spread. The most wrong was NEWSWEEK.

Starting on October 13/2004, Zogby had essentially a dead heat until the 16th when there was a 4 point spread in favor of Bush. Then the numbers fell back into the margin of error until one day before the election. In most instances, Bush was ahead.

Zogby may be wrong this time, but don’t count this pollster out. Relative to the popular winner, he had correctly predicted Bush.

(See Zogby.com and Polling Report.com)