Nerd War - Nate Silver v. Sam Wang v. Drew Linzer - Electoral Vote Projection (11/6/16)

[TomCADem]

Last edited Sun Nov 6, 2016, 11:24 PM - Edit history (5)

Nate Silver tends to get most of the press, but he is hardly the only statistician around. Indeed, Sam Wang and Drew Linzer have sometimes proven more accurate in their projections. While some folks have taken to habitually refreshing 538's percentage projections of who will win, there is also the prediction of the electoral college. This year, there is some discrepancy between the three statisticians, so all three have have a lot on the line in terms of the soundness of their models.

Nate Silver (538) - http://projects.fivethirtyeight.com/2016-election-forecast/
Hillary Clinton: 291
Donald Trump: 247

Sam Wang (Princeton Election Consortium) - http://election.princeton.edu/
Hillary Clinton: 314
Donald Trump: 224

Drew Linzer (Daily KoS formerly of Votamatic.com) - http://elections.dailykos.com/app/elections/2016
Hillary Clinton: 312
Donald Trump: 226

[Cha]

1. Hadn't heard of Drew Linzer.. I already like Sam!

We'll have to take note after the election to see who was RIGHT!

Thank you, Tom

[TomCADem]

4. Two More: Josh Putnam and Desart

Given the wide disparity among the various quants, it will be interesting who among them is closest to the final electoral vote count. Nate Silver has been very cautious about Hillary Clinton's chances, which may be an over correction to his insistence during the summer that there was no way Trump would win the RNC primary.

Josh Putnam (Frontloading HQ) - http://frontloading.blogspot.com/
Hillary Clinton: 340
Donald Trump: 198

Desart (Utah Valley) - http://research.uvu.edu/DeSart/forecasting/november.html
Hillary Clinton: 347
Donald Trump: 191

[Cha]

11. Thank you!

[TomCADem]

5. Also, Natalie Jackson from Huffington Post

She actually backs out about 57 electoral votes as battleground polls, which she does not assign to either candidate, but then lists the chance they break one way or the other. So, with that in mind, here is her breakdown:

Natalie Jackson (HuffPollster) - http://elections.huffingtonpost.com/2016/forecast/president
Hillary Clinton: 317 + 24 = 341
Donald Trump: 164 + 33 = 197

[TomCADem]

8. HuffPo - Showing A Last Minute Tightening of the Face and Greater Uncertainty

The interesting thing is that HuffPo has both Clinton and Trump losing probable voters with a greater number going into the battleground category.

Natalie Jackson (HuffPollster) - http://elections.huffingtonpost.com/2016/forecast/president
Hillary Clinton: 273
Donald Trump: 155
Battleground: 110

[Cha]

12. Mahalo, Tom!

[TomCADem]

15. LA Times - "Our final map has Clinton winning with 352 electoral votes."

LA Times - http://www.latimes.com/nation/politics/trailguide/la-na-trailguide-updates-here-s-our-final-electoral-map-of-the-1478473458-htmlstory.html

Hillary Clinton: 352
Donald Trump: 186

[Cha]

16. I HOPE they are Correct!

'Cause that looks like a Blowout!

[jaceaf]

6. Benchmark Politics

Very new analysis that was very accurate in the primaries:
http://benchmark.shareblue.com/

[Rex]

7. 378

[Hortensis]

10. I reached up some time ago and pulled a card

out of the air that said 335. It doesn't matter if the delegates can't actually add up to precisely that. I'm happy to see I'm "in the game." I'll be thrilled beyond anything if you're right, though. Congress!

[andym]

9. Comey's November Surprise probably helped Wang and Linzer

It's too late to account for the effects of the positive news. But expect a 1-2% bounce for Hillary nationwide which should be amplified in the swing states.

[ThoughtCriminal]

13. Wednesday will be full of "Why our models failed to predict this landslide"

1. Mostly, changes in the profile of a "likely voter"
2. Over reliance on national polls to predict state results
3. And they will throw in the FBI October Surprise / Oops, No Surprise as an explanation too.

Clinton 370+ and 400+ would not shock me.

[TomCADem]

14. Politico - Nate Silver rages at Huffington Post editor in 14-part tweetstorm

It will be interesting to see if the various competing poll specialists begin to gravitate towards each other on the last day. It seems like there is some movement toward this. Perhaps they are all starting to hedge. Of course, whoever nails it will have hefty bragging rights.

http://www.politico.com/story/2016/11/nate-silver-huffington-post-polls-twitter-230815

It began with “This article is so fucking idiotic and irresponsible,” and got only somewhat more polite from there.

Nate Silver unloaded Saturday on the Huffington Post’s Ryan Grim, who accused the polling guru and founder of the prediction website fivethirtyeight.com of “changing the results of polls to fit where he thinks the polls truly are, rather than simply entering the poll numbers into his model and crunching them.”

Rather than taking a simple average -- like RealClearPolitics does -- Silver’s model weights polls by his team’s assessment of their quality, and also performs several “adjustments” to account for things like the partisan “lean” of a pollster or the trend lines across different polls.

According to Grim, however, Silver is “just guessing” and his “trend line adjustment” technique is “merely political punditry dressed up as sophisticated mathematical modeling.” Grim also noted that FiveThirtyEight’s model -- due to his adjustments -- shows Trump more likely than not to win Florida, while the Huffington Post’s calculates her victory there as more likely.

[workinclasszero]

19. 4. I predict an under count of the massive/historic Hispanic turnout

for Hillary.

[Maeve]

17. Y'know...this is the kind of argument I like!

fighting over how much we're going to win by!

[bushisanidiot]

18. Nate may need to tweek his model re: weights, but he is helping drive dems to the polls

complacency is how we lose. urgency and panic is how we win.

[KittyWampus]

20. interesting, thanks for posting.

[Egnever]

22. All three are low

We are going to break at least 320

[TomCADem]

23. THE FINAL COUNTDOWN - November 7th Update

Nate Silver has been lionized as the king of the polling quants, but in past years, his predictions have actually been quite similar to other pollsters. However, this year, there seems to be a wide disparity among the quants regarding how the race might finish. So, does Nate hold on to his throne? Or, does a new quant emerge as the king or queen of the nerds?

Nate Silver (538) - http://projects.fivethirtyeight.com/2016-election-forecast/
Hillary Clinton: 300
Donald Trump: 238

Sam Wang (Princeton Election Consortium) - http://election.princeton.edu/
Hillary Clinton: 312
Donald Trump: 226

Drew Linzer (Daily KoS formerly of Votamatic.com) - http://elections.dailykos.com/app/elections/2016
Hillary Clinton: 313
Donald Trump: 225

Josh Putnam (Frontloading HQ) - http://frontloading.blogspot.com/
Hillary Clinton: 322
Donald Trump: 216

Desart (Utah Valley) - http://research.uvu.edu/DeSart/forecasting/november.html
Hillary Clinton: 347
Donald Trump: 191

Natalie Jackson (HuffPollster) - http://www.huffingtonpost.com/entry/polls-hillary-clinton-win_us_5821074ce4b0e80b02cc2a94
Hillary Clinton: 323
Donald Trump: 215

LA Times - http://www.latimes.com/nation/politics/trailguide/la-na-trailguide-updates-here-s-our-final-electoral-map-of-the-1478473458-htmlstory.html
Hillary Clinton: 352
Donald Trump: 186

Larry Sabato - http://www.centerforpolitics.org/crystalball/
Hillary Clinton: 322
Donald Trump: 216

[Quixote1818]

26. Going 321 instead of 322 because of the elector in Washington State who said she would not vote for

Clinton.

http://www.seattletimes.com/seattle-news/politics/hes-a-state-democratic-elector-but-robert-satiacum-says-he-wont-vote-for-clinton/