My correspondent “Number Cruncher” assesses the Suffolk poll:

Not much to report here. The poll results were all within acceptable ranges (as you reported today)*. The one area you didn’t check are the age splits. Which in my view were perfect: 18-44 year olds represent 30% of the voters, 44 and up – 70%. This is precisely the voting turn out history of Massachusetts. Coakley is in BIG trouble! Most of my polling analysis (using adjustments) have shown her at a tie or up 1. This is the first poll where she is down 4. If this is indicative of reality, her advertising blitz and her dreadful debate performance (as compared to Brown) obviously have not helped her out.
My analysis is as follows: The Suffolk Univ. poll is a reasonable poll for this race during a presidential general election. So the question remains what about the whole “special election dynamic” I believe it favors Brown, although, I think there will be some watering down of this advantage. What we are seeing is BROWN winning this race hands down.
On Tuesday, voters should not be discouraged if they hear about higher turnout than expected, because that turnout is for Brown. Unless something changes in the next 4 days, this might not even be close.
A bit of nitpicking: The African American vote over sampled by a point, women over sampled by 1 or 2 points, Patrick a little more popular than other polls have indicated. But AGAIN nothing here stands out. If you were to point an arrow up or down on this poll I would say that it favored Coakley, albeit very very slightly. The negative leaks coming from Democrat insiders are very real! Obviously not popping the cork yet, but will be very soon!

Remember when the pollsters were shocked by Hillary Clinton’s win in the New Hampshire primary? Well the pollster that was least wrong, i.e., gave Obama the smallest lead, was Suffolk. That’s the good news. The bad news is that last year, in late October, they put Jon Corzine ahead of Chris Christie by nine percentage points.

Having said that, Suffolk University and its Political Research Center, which conducts the poll, are based in Boston. This is their backyard. It’s one thing to get a poll wrong as badly as they did in New Jersey; every pollster has bad days and bad samples once in a while. But if there’s any state that Suffolk ought to be able to weight and measure properly, it’s Massachusetts.

Based on the Suffolk poll, I can’t quite say with certainty that Brown will win. But I will be very surprised if it turns out dramatically different than their results. Barring some dramatic, race-changing event in the final days, a healthy win by Coakley appears to be no longer in the cards.