US Politics Mega-Blog - Page 73

On November 07 2018 05:21 ChristianS wrote:

Justify this please.

Edit: sorry, this came across as more aggressive than I intended. But, uh, why are we discarding statistical modeling as an election predicter? Or just Silver specifically?

On November 07 2018 06:23 xDaunt wrote:

I'm singling out Silver because what he does is turning out to be particularly useless, and his hedging in this election is bordering on embarrassing. If you take a step back and look at his overall performance, he was right in one election, and has otherwise been off. He's a one-hit wonder, not someone in whose models Democrats should find solace.

On November 07 2018 07:41 LegalLord wrote:

His predictions for 2012 were pretty spot on, and his 2016 analysis was justifiably hedged enough to leave him with less egg on his face than, say, Sam Wang PhD and the like. I’ve generally been fond of his analysis, in terms of providing something interesting to follow and to gain some insight, although I must admit that his more politically charged moments have not been much fun. Two that stand out is basically bending over backwards to support Hillary’s “Comey ruined everything” assertion and the recent Kavanaugh stuff where his entire team basically allowed politics to overtake any actual analysis of what was happening. Interesting to see how his predictions play out here though.

On November 07 2018 07:41 LegalLord wrote:

His predictions for 2012 were pretty spot on, and his 2016 analysis was justifiably hedged enough to leave him with less egg on his face than, say, Sam Wang PhD and the like. I’ve generally been fond of his analysis, in terms of providing something interesting to follow and to gain some insight, although I must admit that his more politically charged moments have not been much fun. Two that stand out is basically bending over backwards to support Hillary’s “Comey ruined everything” assertion and the recent Kavanaugh stuff where his entire team basically allowed politics to overtake any actual analysis of what was happening. Interesting to see how his predictions play out here though.

ATLANTA (AP) — Problem signs that arose during weeks of early voting carried into Election Day as some voters across the country faced hours-long lines, malfunctioning voting equipment and unexpectedly closed polling places.

Some of the biggest backups were in Georgia, where the governor’s race was among the nation’s most-watched midterm contests and was generating heavy turnout.

One voter in Gwinnett County, Ontaria Woods, waited more than three hours and said she saw about two dozen people who had come to vote leave because of the lines.

“We’ve been trying to tell them to wait, but people have children,” Woods said. “People are getting hungry. People are tired.”

The good-government group Common Cause blamed high turnout combined with too few voting machines, ballots and workers.

Fulton County elections director Richard Barron acknowledged that some precincts did have lines of voters but said that was due to the length of the ballots and voting machines taken from use because of an ongoing lawsuit.

While voting went on without a hitch in many communities, voters from New York to Arizona faced long lines and malfunctioning equipment.

By Tuesday afternoon, the nonpartisan Election Protection hotline had received about 17,500 calls from voters reporting problems at their polling places. Kristen Clarke, president of the Lawyers’ Committee for Civil Rights Under Law, which helps run the hotline, said that number was well ahead of the last midterm election in 2014, when it had received about 10,400 calls by the same time.

On November 07 2018 04:21 xDaunt wrote:


Yeah, I didn't like Trump before he announced his candidacy. I didn't even like him at the beginning of his campaign. And even after he secured the nomination, I wasn't sure that he wasn't a democrat in disguise. But if I knew then what I know now, I'd have held very different opinions. The difference between us is that I have altered my opinions of Trump as additional information has become available, whereas you have not. You, like most Democrats, are still serially underestimating him, even when there is a very real and growing possibility that he's about to inflict yet another catastrophic election defeat upon you.

On November 07 2018 10:12 xDaunt wrote:
I’m not seeing a blue wave.

On November 07 2018 10:15 GoTuNk! wrote:


Florida senate and governor seem like a flip (de sanctis just tied), how important is it both by itself and as an indicator?

On November 07 2018 10:18 xDaunt wrote:

Nelson losing is a very bad sign for Democrats.

On November 07 2018 10:34 IgnE wrote:
xdaunt did you watch the david frum vs steve bannon “munk debate”?

On November 07 2018 07:55 ChristianS wrote:

I'm always a little surprised Silver's critics don't make more of a fuss about him using frequentist analogies for his Bayesian predictions (e.g. Democrats winning the House is like predicting a die won't roll a 6 - if it rolls a 6 that doesn't mean your prediction was stupid). As you explained it a bit over 2 years ago, Bayesian probabilities are more a "degree of confidence" than a "x times out of y tries" sort of thing, and if you predicted 70% one way and the other happened, there is at least some blame to be assigned (not necessarily on their model). My suspicion is that there are probably legitimate methodological objections you could make to his model, but his critics rarely make them because they're usually either mathematically illiterate or trying to appeal to the mathematically illiterate.