Reply #2: yet another journalist playing with flame bait

[Lexingtonian]

http://www.americanresearchgroup.com

(likely voters)
New Jersey
Bush 42%
Kerry 50%
Nader 1%
Other 1%
Undecided 6%
September 13-16, 2004

Assuming a mere 2-1 split of the Undecideds against the incumbent (it's been 5-1 in recent Presidential elections says Bob Shrum) that's a 54-44 Kerry win at very worst. Actually, Bush still has ~2% (of ~4%) 'bounce' in most states, so Kerry will probably get about the same as Gore, 57%, or a little better even.

ARG is a far more accurate pollster than the ones cited- they're all assuming a massive Christian Right turnout for Republicans in their national modelling, and it hardly needs pointing out that this is not a good assumption in New Jersey.