538 Model Megathread

[muon2]

If you use HuffPo 3-way and limit the polls to nonpartisan pollsters and likely voters the current numbers from their trend lines are Clinton 45.1%, Trump 40.7%, Johnson 5.1%. For the 2-way it gives 45.8% to 42.4% (moderate smoothing)

Why would you do any of this?

This in response to posts on both sides complaining about inclusion of partisan pollsters. HuffPo lets me filter them out so I did this as a point of comparison.

By further comparison that same exercise today has Clinton 45.2%, Trump 42.5%, Johnson 4.9% in the 3-way and Clinton 45.4% to 42.8%  in the 2-way. Clinton's share in both is essentially unchanged from my post 4 days ago.

The Trump line in my exercise is still rising but more slowly than it did last week.

3-way: Clinton 44.5%, Trump 43.6%, Johnson 4.4%
2-way: Clinton 45.3%, Trump 43.6%

The Clinton data is still very steady as it has been over the last month. Note that when Johnson is not included Hillary gets a quarter of his vote, but the rest goes to undecided. To eliminate recent noise and movement between undecided/third party I can put on more smoothing on the two-way results and she is at 46.0%. However, over time Trump continues to gain at the expense of Johnson, but the trend lines don't show him catching her by Tuesday.

I think 538's model is reacting like the trendlines I'm seeing with HuffPo's tool. The Trump poll numbers have stopped moving up so his trendline comes down and hers moves up. Here's what it gives me this afternoon.

3-way: Clinton 45.5%, Trump 42.9%, Johnson 4.5%
2-way: Clinton 46.3%, Trump 43.3%

[Admiral Kizaru]

Nate flipped Nevada, NC, and Florida all to Clinton.

Fat lady is onstage, waiting for her cue.

lol watch the sycophants proclaim now that he accurately predicted the election two cycles in a row now!

[Erich Maria Remarque]

Nate "Discredited Fraud" Silver herding to the consensus the day before the election? Sad!

This thread (and FBI thread of course) let me understand why Trump is so popular (he'd never get >5% in Western Europe).

This thread is very anti-scientific, full of "experts", full of the conspiracy theories, full of partisan hackery that agree with each other (like Trump's voter mirror his views about rigged election etc) and name-calling. FBI thread has all this, but also reveals how partisans are ready to believe everything that media/their "former FBI sources" say.

Haha, you deserve it.

OnTopic: I think that marginwise this model will be one of the best (probably in Nevada it will have extra big error).

[Crumpets]

I think Nate should let his model run through election night, changing each state's chances to 100% as the calls come in. It would be great to watch the probabilities over the course of the night like ESPN does for football games.

[Smeulders]

Nate "Discredited Fraud" Silver herding to the consensus the day before the election? Sad!

Indeed!  Poor Nate doesn't know what to do and is herding towards the consensus mighty quick!  And to think he was attacking PPP for doing so in 2012 and praising those brave enough to show an outlier at the end.  Shame he wasn't brave enough himself.  Sad!

The 538 model is at its core a weighted-polling average. I'm not sure how you like your weighted polling averages, but I'd be running for the hills if they mine were showing an outlier...

[Smeulders]

I think Nate should let his model run through election night, changing each state's chances to 100% as the calls come in. It would be great to watch the probabilities over the course of the night like ESPN does for football games.

I really want to see the model with some likely scenarios given the early voting. Nevada is still a toss-up, as there has been few November polls. Given what we know of Nevada early voting, odds are massively in Clinton's favour. Asking to build an "early vote model" on top of the polling model is probably asking a bit much though.

[100% pro-life no matter what]

I think Nate should let his model run through election night, changing each state's chances to 100% as the calls come in. It would be great to watch the probabilities over the course of the night like ESPN does for football games.

He did that during the midterms.  It also adjusted other states when calls were made as it became apparent that there was a pro-Republican sentiment that the polls didn't capture.

[Meclazine for Israel]

As I  noted before, the sole reason the model had Trump up in NC, FL and NV was the "trend line adjustment" which was pushing  every state a bit more than 2 points in Trump's direction.  Today that has been cut in half and that flips those three states back to Clinton.  Now 538 aligns with all other models in terms of state wins, but still is more bullish on Trump in terms of margins and odds .

Was it purely data that changed the trend line in the last few hours?

The trend line will become neglible as the election nears. And there was seemingly an adjustment up for Trump.

As for the other comments, Nate is not going anywhere.

His website is top of Google. Well presented and easy to understand.

He is making a fortune out of website advertising alone.

[rafta_rafta]

His website is top of Google. Well presented and easy to understand.

He is making a fortune out of website advertising alone.

Wang needs to work on this. His name recognition is negligible compared to Silver. And the PEC site has a poor user interface

[GeorgiaModerate]

Nowcast is back up to 70% even for Clinton, with the other two models right behind at 69.5.

[Angel of Death]

How can you people be angry at an algorithm? Seriously, does Obamacare cover Nate Silver Derangement Syndrome?

[‼realJohnEwards‼]

As I  noted before, the sole reason the model had Trump up in NC, FL and NV was the "trend line adjustment" which was pushing  every state a bit more than 2 points in Trump's direction.  Today that has been cut in half and that flips those three states back to Clinton.  Now 538 aligns with all other models in terms of state wins, but still is more bullish on Trump in terms of margins and odds .

Was it purely data that changed the trend line in the last few hours?

The trend line will become neglible as the election nears. And there was seemingly an adjustment up for Trump.

As for the other comments, Nate is not going anywhere.

His website is top of Google. Well presented and easy to understand.

He is making a fortune out of website advertising alone.

That is not how Trend Line works. It compares the national polls then to the national polls now, and adjusts all polls by how much they've moved.

[Lief 🗽]

Shouldn't all three models be the same by now?

[GeorgiaModerate]

Shouldn't all three models be the same by now?

They should be converging, but there seem to be some weird effects.  There's been another update since my earlier post, and the changes since then are:

Nowcast: 70.0 -> 70.2 ( +0.2 )
Polls-only: 69.5 -> 69.2 ( -0.3 )
Polls-plus: 69.5 -> 70.3 ( +0.8 )
 
I've given up trying to understand it.

[Meclazine for Israel]

News.com.au predicts a Trump win:

http://www.news.com.au/finance/work/leaders/why-trump-could-really-win-this/news-story/8e4bd4e6f5a65434724702ed8c8f221f

"one of the most respected pollsters in the United States, Nate Silver"

Haha. Told you top of Google meant something.

[emailking]

Shouldn't all three models be the same by now?

Tomorrow yes. But the fact they're diverging now proves that Nate is manually manipulating the data and the algorithms to herd towards the consensus that a solid Clinton victory implies her pre-election odds should be 98%+. To be honest, I would expect nothing less from such a discredited hack fraud.

[Crumpets]

Clinton back over 70% in polls-plus.

[The Other Castro]

Ugh now I'm in the awkward position of hoping Clinton loses North Carolina just to keep everyone from going OMG HE DID IT AGAIN! 50/50 STATES!!!!!!!1!

[LLR]

News.com.au predicts a Trump win:

http://www.news.com.au/finance/work/leaders/why-trump-could-really-win-this/news-story/8e4bd4e6f5a65434724702ed8c8f221f

"one of the most respected pollsters in the United States, Nate Silver"

Haha. Told you top of Google meant something.

Except that he's not a pollster

[Ebsy]

News.com.au predicts a Trump win:

http://www.news.com.au/finance/work/leaders/why-trump-could-really-win-this/news-story/8e4bd4e6f5a65434724702ed8c8f221f

"one of the most respected pollsters in the United States, Nate Silver"

Haha. Told you top of Google meant something.

Except that he's not a pollster

Nate Silver hates when people call him a pollster.

[Fmr President & Senator Polnut]

News.com.au predicts a Trump win:

http://www.news.com.au/finance/work/leaders/why-trump-could-really-win-this/news-story/8e4bd4e6f5a65434724702ed8c8f221f

"one of the most respected pollsters in the United States, Nate Silver"

Haha. Told you top of Google meant something.

Except that he's not a pollster

Stop posting links to any Australian journalists... this season has shown they know/understand f*** all.

[Lachi]

News.com.au predicts a Trump win:

http://www.news.com.au/finance/work/leaders/why-trump-could-really-win-this/news-story/8e4bd4e6f5a65434724702ed8c8f221f

"one of the most respected pollsters in the United States, Nate Silver"

Haha. Told you top of Google meant something.

Expected from a right-leaning website.

[riceowl]

Ugh now I'm in the awkward position of hoping Clinton loses North Carolina just to keep everyone from going OMG HE DID IT AGAIN! 50/50 STATES!!!!!!!1!

Ohio!

[The Other Castro]

Ugh now I'm in the awkward position of hoping Clinton loses North Carolina just to keep everyone from going OMG HE DID IT AGAIN! 50/50 STATES!!!!!!!1!

Ohio!

That would be the ideal separator, though it's probably easier to lose NC than to win OH. This is all in jest so whatever.

[User157088589849]

Does this sum up the state of the race

Clinton has a better chance in Alaska than Trump does in Michigan, Colorado, New Mexico, Wisconsin and Minnesota.

Clinton 24.5% chance of winning Alaska
Trump 25.0% chance of winning Pennslyvania

So Trump more likely to win PA by 0.5% than Clinton of Alaska.
http://projects.fivethirtyeight.com/2016-election-forecast/