I had my doubts about the way Scott Rasmussen was conducting his robo polls. But, he, after months of static numbers, recorded an Edwards bounce that was sustained for a while. His chart is here.
But, now the race is tied again. If we accept that there was a bounce, Rasmussen provided more measurable proof than most. If we accept then that Rasmussen is on to something with his polling, we must then conclude that the Edwards bounce is gone and it is back to 50-50 — actually 46-46.
As RCP said a while back, Edwards was a poll driven mistake.