There is a report that Martha Coakley's internals for Sunday night put her ahead by two points, 48 to 46 percent. For what it's worth, my little birdie tells me that her Sunday night poll put her ahead by one point and that the three-night average for Friday, Saturday and Sunday also put her ahead by one. Recall that Scott Brown had pulled ahead (by as many as three points) in Coakley's polls late last week. So this might indicate potential tightening.
That said, there's a catch: Coakley also apparently has numbers that showed Brown slightly ahead both on Sunday night and over the last three nights -- a different turnout model.
Her own poll shows her failing to crack 80 percent among Democrats (while Brown is over 90% with Republicans) and losing independents by more than 30 points. A very smart Massachusetts Democrat tells me that "I'm still guessing Brown wins, but it's really a toss-up at this point."
For a look at how the race will be won and lost tomorrow, check out my guide to the three victory models for Massachusetts Republicans.