A new poll is out in Massachusetts, done by Research2000.

They have Coakley up 49 percent to 41 percent, which I find pretty plausible, with a few caveats. They have independent, self-proclaimed Tea Party candidate Joe Kennedy at 5 percent, which seems a little high. I think he’ll probably be similar to Chris Daggett’s third-party bid in New Jersey, fizzling on Election Day.

Second, in this poll, independents split 49 percent to 36 percent in favor of Brown. Considering the mood of the electorate, smaller turnout in a special election, and what we’ve seen from independents in other states, I wouldn’t be surprised if Brown did better – maybe a lot better – on Election Day.

UPDATE: The guys at RedMassGroup note that Democrats often don’t perform quite as well on Election Day as they do in Research2000’s final polls. Of course, this doesn’t necessarily mean the pattern will hold, but based on the experience of Bob McDonnell, Chris Christie, etc., Scott Brown should feel good about performing better than the modest 41 percent he gets in the R2000 poll.

A couple of folks are arguing that the sample has way too many Democrats. I think that’s probably the case, but not definitely the case; figuring out who shows up in a special election is a big gamble. Do I think the electorate on Tuesday will be fewer Democrats and more Republicans and independents than usual? Sure. But with this last-minute ad blitz, Brown fans can’t count on that.