The political junkies among us (sign me up) aren't willing to wait until all the votes are counted. We're looking for signs, early on, to tell us which way the winds are blowing. Tonight I'll be looking for those signs especially in what are expected to be the closest statewide races -- Welch/Rainville (U.S. House) and Dubie/Dunne (lieutenant governor).

Early returns typically come from the very smallest towns with few ballots to count and the cities, where machines do the counting. Conventional wisdom says the Republicans, Rainville and Dubie, need to come out of populous, red-leaning suburbs (Colchester, Essex) and cities (Rutland City and Town, Barre Town) with more than 50 percent of the vote. They need a solid cushion to balance the big blue vote for Democrats and Progressives in Burlington and in wide swathes of Windham, Windsor and Washington counties.

Of course, every election is different. Vermont hasn't had an open U.S. House seat since 1988, too long ago to make useful comparisons of voting patterns. The last truly open statewide race was the Jim Douglas/Doug Racine battle for governor in 2002. So as early towns report, I'll be comparing the Rainville/Welch results with that 2002 election (though the presence of independent Con Hogan in the 2002 race may make it too complicated).

And the race for governor? For me, the question is -- how much better a showing does Democrat Scudder Parker make against Douglas than Peter Clavelle did in 2004? Clavelle got just 38 percent of the vote. Parker is almost certain to do better, especially since this is supposed to be a big election for the Democrats. Douglas is expected to win, but if Parker hits 45 percent or better, that's a big wake-up call for the governor.

Anybody else have a favorite way of reading the early results?

-- Candace Page