Why Even Close?

Suffolk University literally sits in the shadow of the State House on Boston's Beacon Hill and frequently conducts political polls.

Their read on the race had Brown winning handily, based on a Sunday night phone poll of three "bellwhether" communities which voted within 1% on the statewide total for 2006, the last senate race outside of Presidential year.

It seems like a valid method until you take into account the personalities & the communities.

It was larger than life Ted Kennedy (then a 44 year incumbent) easily winning re-election to the Senate capturing 69 percent of the vote against Republican language school owner Kenneth Chase, a political non-entity.

There's no one that famuous in this race. It's not the same choice.

And while the communities of Fitchburg and Gardner in Worcester County (central Massachsetts) may have mirrored the state-wide percentage, they are medium sized standalone cities outside of Metropolitan Boston where the bulk of the state's population lives... Add in Peabody a bedroom community along the inner of Boston's two beltways, and the total population of the sampled cities is only 112 thousand, out of the whole Commonwealth's 6.6 million.

The poll becomes a sample of a sample but not valid by meaningful demographics. They might be right, but the method seems dicey.



The onscreen poll as Keith and Markos talked was the only one close.