â It is often claimed, as in the Seattle Post-Intelligencer, that âBefore Tuesday, gay marriage had never won a statewide vote anywhere in America.â That is false. In 2006, the voters in Arizona (curiously) rejected a constitutional amendment to ban gay marriage. This was undone two years later, but the 2006 vote remains a fact. â I donât know the mind of Angus King, of course. But there are many good reasons to assume that he will caucus with the Democrats, and mostly vote with them: (1) He is a former Democrat, and apparently somewhere left of center; this is essentially the position of the Democratic caucus, while the Senate Republican caucus is quite conservative, and only getting more so. (2) His state is left of center, and far to the left of the Republican caucus. Maine voted 56-41% for Barack Obama; Obama won every age group and both genders. Democrats won the stateâs two House seats 65-35% and 58-42%. The state also voted 53-47% for gay marriage. (3) His electoral coalition is probably not even left-of-center; it is probably just liberal. King won 53% of the Maine vote, to 31% for the Republican and just 13% for the Democrat. In napkin-math terms, King pulled 10 of Romneyâs 41% but 43 of Obamaâs 56%. King was elected by Democrats. Heâll have to vote like one to get reelected. (4) If nothing else, the Democrats are in the majority and likely to remain so. Throwing in his lot with the Democrats brings him more power, and if being in the majority is the only thing he cares about, he can switch to the Republicans if they regain the majority. In this light, it would be the best of all worlds if King had the deciding vote for the majority, Ă la Jim Jeffords. But I donât think the calculation needs to get as far as #4. The first three criteria are enough. â Daniel Engber wants to give credit for Nate Silverâs success to the pollsters. This is not fair at all. Pollsters are part of the team, no doubt, and they do the heavy lifting; but theyâre not the brains. No one pollster was reliable at all. Engber seems to recognize that it was at best a collective victory for the pollsters. But each individual pollster stuck predictably with his or her own numbers. None of them was self-effacing (or honest) enough to bring in the sampling of other pollsters to improve the reliability of the numbers. That is why poll aggregators did better than individual pollsters. Someone needed to apply superior methods (and possibly superior intelligence) to the half-cooked poll results produced by individual firms. RealClearPoliticsâ crude average was more reliable than the pollstersâ but less than the smart aggregators, and several aggregators predicted the correct result with (and before) Silver. Interestingly, Engber praises the discretion Silver uses in weighing polls, in a piece dismissing its broader value, and fails to recognize the crucial difference between Silverâs average and RCPâs. â To quote the Atlantic: âWith the Romney campaign conceding Florida, the final state that was undecided, Barack Obama captured a total of 332 electoral votes to win reelectionâ. As a reminder to them and the rest of the news media, elections are decided by voters, not by which party concedes first. I canât begin to tell you how foolish and offensive this frequent media mistake is. â O.T. Ford