9:34pm. CBS: New Hampshire called for Obama. Erm, election’s probably over. And unskewers are full of it. On the same note: Florida is Obama 50.01%, Romney 49.99% with 81% counted. <0.5% is recount territory (Brennan Center).

9:25pm. Electoral-vote.com totally overwhelmed. We were gone, back now.

Indiana Senate race Donnelly over Mourdock 48-46% with 63% counted. Right in line with poll medians. We’re headed for Democratic gains in the Senate.

9:19pm. Pennsylvania for Obama. Evidently pollster skew is less than 5%. (Plot it in the Geek’s Guide!)

8:25pm. Florida appears to be 50-50 with 50% counted. If that holds up, it’s right in line with polls. Recount, ZOMG!!1!!eleven! But not much to reassure those who think pre-election polls are skewed.

8:20pm. Big margins in OH (for O, early voting) and VA (for R, GOP counties) not informative. However, Donnelly (D) 50-44 over Mourdock with 27% counted in IN? Hmmm, that’s interesting.

8:14pm. Watching ABC, reminded how much I hate TV coverage on Election Night. We know more the day before (polls) and the day after (full results). Halfway-counted Virginia? Watching Anderson Cooper’s discussion. No offense to Henrico County, don’t care…

7:40pm. OK, guys – listen…Exit polls don’t tell you anything – they can be off, and get adjusted later to match actual returns. They are a research tool for the media to construct stories.

Also, large states like VA – these are too large to give accurate results in the early evening. Watch NH as fast-counting indicator of national result. See Geek’s Guide.