Three new polls from New Jersey, all show a closer race than September.

Marist
Oct 6th, 321 LV, 600 RV
	     LV      RV
Corzine       44      45
Forrester     43      38

That is an extraordinarily small likely voter sample, and I would hope that I hardly need to repeat the problems with conducting likely voter polls more than two weeks away from an election (scroll about halfway down the link for the discussionof early LV models). It also conflicts with every other poll out right now:

Star-Ledger
638 LV
	     10/6    Sept
Corzine       44      49
Forrester     37      31

Rasmussen 
500 LV	 
	     10/4    9/15
Corzine       45      47
Forrester     38      36

While I don't buy the Marist likely voter model, it does seem pretty clear that the race is getting closer. Forrester is within single digits and seems to have the mo'. Go to Jon Corzine's website today and see how you can help bring this one home.