"serve as America's safety net to keep someone like Trump becoming President." That was the original intent....the people cannot be trusted to elect the President directly. An educated and well informed electorate should eliminate the concern....uh, yeah, I take your point.

Take a look at the duration of civilizations through time...not all that long on average.

I hear you on data entry, I try and add a new set every week.

Are you working from state polling data or state probabilities of party victory? Simulation or computation?

https://twitter.com/TheFix/status/789090575249125376

Why doesn't he smile more?

Clinton 340 Trump 198 with probability of Clinton win @ 99%.

Watching Utah for fun. McMullin may take the EV.

It seems Clinton has decided not to run out clock and has moved into red states. Texas and Georgia will be on my watch list.

Not a big fan of the electoral college.

I wasn't either, and I could never figure out why we used this process to elect our President....until this election. They serve as America's safety net to keep someone like Trump becoming President.

You have probably read this, but it might be of interest to others.

http://blog.paddypower.com/politics/2016/10/18/paddy-power-pay-out-hillary-clinton/

~Phil~

The Paddy Payout puzzled me until I considered it didn't cost them much at 2/9 odds. Plus they probably figure a lot of that money will get plowed back into other Paddy wagers, or even back into their Betfair division which is still running the Clinton:Trump battle. At the very least, it got them a lot of free advertising on line and in print.

http://selwynduke.typepad.com/selwyndukecom/hillary-clintons-vulgar-m.html

I'm pretty sure the matrix is constructed post simulation and does not drive the simulation. Nate wrote an article a few months ago that seemed to imply his model WAS driven by a big state by state matrix, but with further inquiry to other sources (that answer my emails) I concluded it unlikely, a problem of mathematical under identification on the one hand, and on the other, there really isn't enough election data to estimate the correlations with much confidence.

Based on what Silver published in his blog around 2012, I think Nate was breaking model variance into a local and national component as I do, but he may have moved forward into other schemes since. I've toyed with the idea of census urban/suburban/rural codes and even the Target/Walmart ratio (slightly kidding).

The Special Case model worked well in 2012 for me late in the election, it agreed well with Silver and Prediction markets and bookies. It's working well late in this 2016 election - and I have a lot more data sets to play with in 2016. The nice thing about the special case is that you don't need to run simulations at all, the answer is a simple computation.

This idea has been the subject of CW.com in the past, I think it's the most feasible path towards something close to proportional election of the President that we are likely to see.

I'm all for it!

I think California has passed a rule that they will proportionately allocate EVs if more than 2/3 of other states agree. I think this would be great for the country because the candidates would need to appeal to everybody, regardless of which state they live in.

The 538 gave some insight into their correlation model at the end of the article below.

http://fivethirtyeight.com/features/if-trump-wins-heres-how-the-map-might-look/

Despite that, I think the press would have been far more generous to Trump had he not decided to play coy on the question of concession. The bar for him was set ridiculously low, and he managed somehow to take that bar and smash himself in the face with it anyway.

That said, he probably picked up some support from the religious right, owing to his raw description of late-term abortion, and for lending credibility to propagandist James O'Keefe's latest fake videos.

"Trump needed a big win last night. He did not get it. His closing threat of a "November Surprise" will not go unnoticed and was a "Yuugge" blunder. I don't think his odds will move much lower, but only because they can't go a lot lower."

I do think the odds of a Clinton winning in a blowout are much higher given last night's debate. I run The UpShot state probabilities from NYT, 538, HP, PW, PEC and DK through two of my own math models about once a week. Actually it's one model and a special case of the one model, but that's a detail. I then compare the Upshot six probabilities of a Clinton win. The results indicate that 5 of the six (NYT,538, HP, PW and DK) prediction shops behave as if the 50 states and DC behave in a highly correlated fashion, all states tend to drift D or R in the same direction on any given election day. In other words, the upcoming election is going to be a strong wave election, effectively decided by roll of one die. The exception among the six is PEC (Princeton Electoral Commission) which seems to evaluate local state effects as fairly important i.e. a smaller wave effect.

A median wave Clinton EV looks to be about 350, a blow out in excess of 400 EV is not fantasy (5%). Trumps equivalent 5% dream is 300 EV, not a blow out at all.

Is there going to be anything left of Trump's vaunted multi-billion $$$ Brand Name? Once associated with luxury and an inflated price tag, but now associated with trailer park. Will this be reflected in this years Forbes top 100 American Plutocrats list?

Trump! network. Don't think so. If you can't manage 90 minutes of watchable content once every couple of weeks.....just fill in the blanks, Trump! style.

Wallace is well known for asking pointed questions, but he also tends to help an interviewee focus during follow up. Trump did not exploit this very effectively, Clinton did. Trump is simply not very fast on his feet.

All and all, I think Wallace did a very fine job. I think one moderator works better than two. This was the most "traditional" "debate" of the series.

Hillary is not the natural public speaker that her husband is, but she has clearly spent a lot of time learning to appear friendly and approachable, and mostly pulls it off, in the sense that airline attendants and grand parents do. She has a lot of debate experience, and it showed. She was very effective in rattling the Donald and much more coherent. She speaks in short paragraphs, and can manage time when she wants to. Not great, but highly serviceable, with a lawyers attention to detail.

Trump does not speak English during debates. He cannot complete a sentence, so don't even think about paragraphs. He naturally communicates in tweets. Call it "TwEnglish,"
but it's a broken form, it implies a thought, sputters some nonsense and then moves on to the next implication. It's about as solid as jello. You can read into it what you want. In that sense it's a form of dog whistle, but more fun!

Throughout these debates, Trump has shown himself to have about 30 minutes of stamina. He fades for the remaining 60 minutes, when not speaking, his eyes close and his head drops, which I suppose is meant to signal deep thinking, but also looks a lot like a guy, who doesn't sleep much, dozing. Trump has been consistently walloped in those final 60 minutes, because Clinton paces herself well and outlasts him.

Trump needed a big win last night. He did not get it. His closing threat of a "November Surprise" will not go unnoticed and was a "Yuugge" blunder. I don't think his odds will move much lower, but only because they can't go a lot lower.

Thanks. Soundo like Rudi, Roger and the boys were riffing. I bet Kellyanne was hoping that he wouldn't go all 1950s on stage.

https://twitter.com/KFILE/status/788924793936576512/photo/1

"I should have gotten it"

The GOP got what it deserves for nominating a self-absorbed reality TV diva - a penthousewife who wants to be queen.

When they go low, we go vote. #NastyWoman

"There was even a time when he didn't get an Emmy for his TV program three years in a row, where he started tweeting that the Emmys were rigged against him," Clinton noted.

"Should have gotten it," Trump interjected.

I mean, really. Seriously?

I'm rather good at explaining Americanisms to Brits (and Irish).

"Hombre" is Spanish for "man," obviously. But in Americana, it's a word from Westerns, kind of like the etymology of "pardner."

In 50's Westerns, "hombre" is almost always preceeded by "bad" (as Trump did so tonight). So it's seen as a put-down, and it's also seen as a bad example of "Spanglish." Cultural appropriation, and all that.

So my guess is that Latino groups will be at the forefront of reaction to the word -- yet another group Trump is making zero inroads with...

Does that help?

:-)

-CW

I'm reading snarky comments to debate articles now that are saying exactly the same thing: "Stay home, Trumpkins! The election is rigged, so don't even bother voting! That'll show them!"

Heh.

:-)

-CW

Our motto here has always been: "I watch the debates so you don't have to!"

:-)

Thanks for the kind words.

To everyone -

Looks like I missed a Trump gaffe, as I didn't even note down his use of "hombres." The internet did notice, however...

Heh.

-CW

All in all, Hillary just had to make sure Trump was the headliner after the debate. Planned and executed.

:-)

-CW