Tuesday, January 19, 2010

The end of a very short era

9:47 -- Oops, one more thing... interestingly, the 2006 Senate race results were a MUCH better predictor of tonight's results than the 2002 Governor race, at least in the 23 towns I chose: r-value of 0.69 vs. 0.03.

9:26 -- Well, Martha, you reap what you sow. Blergh. G'night.

9:21 -- AP calls it for Brown; Coakley has conceded.

9:19 -- With 14 of my 23 bellwethers reporting, I have Coakley running a median 5% behind (which nicely mirrors the actual state-wide returns so far). 71% in. Unless the final half of Boston has a HUGE swing, both in terms of trends and in terms of overall votes, this is over.

9:17 -- Framingham (decent size) actual +3, need a tie

9:14 -- Whitman actual -7, need -2. Total of 65% of precincts reported.

9:11 -- Here are two biggies: Quincy actual -4, need +2; Revere actual -3, need +4. Get out the hammers, that coffin needs a'nailin'

9:09 -- Palmer (small town) actual -11, need -2

9:08 -- Norah O'Donnell really sucks

9:07 -- Gardner actual -7; needed even (maybe not so close)

9:06 -- Dedham actual -6; needed -2

9:04 -- Brockton actual +5; needed +5. This could be close...

9:02 -- Coakley won Acton +8. Needed a draw. Again, the blues are deeper, the reds are deeper...

8:57 -- Coakley won Sharon +5. I had her as needing +5 (PS my Concord number was off by a factor of two... she won +12.5, i.e. 12.5 points above 50%)

8:55 -- Well, Coakley won Nantucket. And Edgartown. No returns from Chappaquiddick yet...

8:49 -- 37% in, Globe site still largely in FAIL mode

8:46 -- Looking at the map, my take is that blue areas of the state are really blue, while red areas are really red. I wonder if that is my visual bias, or will actually be borne out by the final results...

8:43 -- Most of the bigger towns, not surprisingly, are not in, or not fully in yet.

8:37 -- 18% in, Brown still up statewide by 5%. Checking individual towns...

8:30 -- According to Nate Silver's quick analysis, for towns with all reported, turnout is in the 60s. That's HUGE. 72% in Concord, which again, went Coakley by a lot

8:28 -- Official state elections page doesn't have anything up yet

8:26 -- According to Boston Globe (which, again, had problems on primary night) Coakley won Concord by 25%. I don't believe it. I had Concord as Coakley-needs-to-win-by (henceforth known as D+) 2%...

8:24 -- Southampton, one my 23, is in with Brown by 7%. It is the smallest town on my list, but my back-of-the-envelope math says that Coakley "wanted to lose" by less than 4%

8:18 -- Globe site crash. NECN shows Brown up 5% with 4% reporting.

8:16 -- Okay, about 7,000 votes are in, but we don't know from where. Brown up 9%

8:07 -- Rachel Maddow also at Doyle's...

8:06 -- According to Rasmussen Reports (salt taken), from their survey today, "Among those who decided how they would vote in the past few days, Coakley has a slight edge, 47% to 41%."

8:05 -- Okay, polls closed. Still looking for anything...

7:55 -- Not sure what the best site will be for real time returns. I remember that the Boston Globe's site was having problems during the primary, so that might not be your best bet. CNN tends to be pretty good.

7:47 -- According to Mike Barnacle, until two weeks ago, the Coakley campaign did not have a pollster. I'd call Coakley a turd, but for whatever reason that seems like a particularly male epithet. What is the female equivalent of turd? (I've also wondered what the female equivalent of douchebag is, but that's a discussion for another day)

7:42 -- What I was doing, but all-out. All the data you'd want to look at for recent Mass elections.

7:28 -- Beyond the fact that I want him to lose for actual policy reasons, I'd really like Brown to lose because of all the windbaggery I'm hearing on MSNBC. In addition to some Politco talking head (yes, that is redundant), Chris Matthews just noted that, were Brown to win, he'd be a good choice for Republicans to run for president in 2012. Um...

7:26 -- I'm going to buy a truck, just in case I want to run for office in 20 years. Never hurts to have an old truck.

7:23 -- Chris Matthews is hanging out down the street at Doyle's, a veritable institution in JP... hmm, that is tempting...

6:49 -- Okay, not much more to do until the polls close at 8. For what it is worth, about 40% of the people who live on my street had voted today in Jamaica Plain (it is very easy to read the voter roll and see who has a check-box next to their name, which is sorted by street). I voted on the early-end (~5:30pm) of the post-work crowd, so that number will certainly go up, although who knows by how much. Coakley will need a huge turnout in places like JP if she's going to have a shot.

6:21 -- This is vaguely interesting... I compared the results of the 2006 Kennedy-Chase Senate race to the 2002 Romney-O'Brien Governor race. On a macro level, things look like they make sense. The 23 towns I chose missed the 2002 actual result by overestimating the (D) by 1.0 percentage point; in 2006, overestimating the (D) by 0.8 percentage points. Further, the turnout was remarkably similar across those two years; while towns certainly varied a bit, the overall vote difference for those 23 towns was 434. Finally, when you look at how well the 2002 race predicted the 2006 race, well, the correlation coefficient is 0.24. Of course, the range of the data is pretty narrow, so that small an r-value is pretty much expected (if, for example, outliers like Cambridge (D) or Medfield (R) were included, then the r-value would look much better).

6:04 -- Doug Flutie is a Scott Brown supporter. I suppose I shouldn't be surprised, most (white) athletes are Republicans.

6:01pm -- Well, the good news is that, when Coakley loses this race, Joe Lieberman will no longer be the critical 60th vote.