This poll definitely raised my eyebrows, based on what I’ve seen in the Democratic primary for Senate District 14. It brought to mind another surprising poll from another Democratic primary race two years ago, where the pollster found Jeff Harris with a strong lead with a month to go before the primary, and Margaret Donnelly trailing badly. The final result was actually Koster eking out a win over Donnelly, with Harris trailing far behind. (In the interest of full disclosure, I consulted for Margaret in that campaign.)

That poll took some of the wind out of the Donnelly campaign’s sails, as it was a little more difficult to raise money and win endorsements in that last crucial month with independent results supposedly proving that the campaign was flagging, when the campaign itself had polls conducted by very respected pollsters that showed the opposite. In any case, the Harris campaign got enough momentum and enough votes from Donnelly to hand Koster the win. (Let’s not even talk about the Molly Williams campaign.)

And the firm who conducted this poll with such dubious results? None other than Research 2000, which Daily Kos is now suing for producing bunk polls. (Michael has some more here.) Now, I’m not alleging that Research 2000 produced a bunk poll for the 2008 Democratic primary for Missouri Attorney General, and in fact their excuse rings true for how they got the primary so wrong. But it is a delicious coincidence, nonetheless.

I’ll have more to say about the current Senate District 14 primary, but  the lesson here is really to take polls like these with a giant boulder of salt.

Oh, and there’s one other weird connection to the 2008 Democratic AG primary besides a way too early poll with surprising results. After Koster won the Democratic nomination, Maria Chappelle-Nadal endorsed his Republican opponent and even recorded an ad for the GOP nominee.