13
   

Data-Driven Predictions for the 2016 US Presidential Election

 
 
Reply Wed 3 Aug, 2016 05:52 pm
There's a lot of "gut" predictions thrown around here for the upcoming US presidential election, but most of them boil down to "I want X to win so I think X will win."

This thread is for more serious, data-driven predictions on the election outcome. Here are some sources for polling data, prediction models and betting markets. If you have any other good sources for such things please share!

- Here is Nate Silver and 538's prediction model
- Here is the NY Times' prediction model
- Here is Princeton's Election Consortium
- Here is Predictwise, incorporating several betting marketplaces
- Here is Predictit, a real-money prediction market
- Here is the Hypermind Prediction Market
- Here is the Cook Political Report forecast
 
Robert Gentel
 
  2  
Reply Wed 3 Aug, 2016 05:55 pm
@Robert Gentel,
At present, the consensus from the above data sources is that Clinton has over a 70% chance of winning the election. However this changes rapidly based on polling data (prediction market data has been much more stable). In the second half of July Trump had narrowed Clinton's 80% vs 20% lead to nearly a dead heat and was within a few percentage points of Clinton in many prediction models just a few days ago (they were essentially 50/50) till his Khan missteps and subsequent campaign implosion (yet again).

However Trump has imploded before and come back this can and will change week to week and on this thread I'll be updating my meta-analysis of these prediction models as things change and as time permits.
cicerone imposter
 
  2  
Reply Wed 3 Aug, 2016 06:19 pm
@Robert Gentel,
I agree it's going to fluctuate, but I don't see how Trump can gain when he continues to show his ignorance and bigotry. For some reason, those close polls doesn't seem reasonable to me. I thought the die hard Trump supporters were in the mid 30s range.
rosborne979
 
  4  
Reply Wed 3 Aug, 2016 07:11 pm
@Robert Gentel,
It's also worth remembering that it's advantageous for each side to make it seem like a close race, because each of them need to get people to actually get out and vote. A candidate who was "predicted" to have a 98% chance of winning could conceivably lose simply because people believed the prediction and decided they didn't need to cast their vote.

In past elections I have found the consensus of Vegas Bookies predictions to be the most accurate, probably because they make (or lose) money based on their objectivity.
thack45
 
  0  
Reply Wed 3 Aug, 2016 07:28 pm
@rosborne979,
It may at times be advantageous, but certainly not always.. http://www.huffingtonpost.com/ryancasey/trumps-speechwhen-he-quits_b_11112630.html
Quote:
A presidential election with real, cold results is the one thing that could cut through Trump’s bullshit veneer of awesomeness. Indeed, during the 2012 campaign, Obama held a relatively small but stable lead over Romney in state and national polls for months. As the election drew near, leading poll aggregators like Nate Silver, Sam Wang, Drew Linzer and others projected with high confidence that Obama would win easily. Yet a sizable contingent on the right simply chose to deny the empirical evidence. Joe Scarborough mocked Nate Silver just one week before the election for suggesting the data showed it was more likely Obama would win. George Will predicted a Romney landslide. Peggy Noonan predicted Romney would win because she saw he had large crowd sizes and lots of yard signs. “Unskewedpolls.com“ became every right-winger’s favorite site for a time that fall because it weighted the polls to show a huge Romney lead. Meanwhile, many Republicans simply chose to stop believing in the accuracy of statistical inference.

As a result, Republicans were positively shocked at the election results. Romney’s own pollsters had constructed an outdated model of the demographics of the American electorate to give their boss encouraging results and buoy his spirits. Consequently, Romney and Paul Ryan were left dumfounded; their high-priced campaign pollsters misled them about an election result they could have seen coming for free on Silver’s, Wang’s, or Linzer’s blogs. There was no sorcery in the math—these blogs simply projected the likely outcome of the election by aggregating publicly available, state-level polling data, which has proven to be highly accurate and reliable in modern presidential general elections, especially due to the highly polarized nature of the American electorate.

Even the great Republican Svengali Karl Rove suffered a meltdown on live television when he doubted the math of the statisticians at the Fox News decision desk. After all, the actual voting results—though entirely consistent with state-level polling leading up to the election—conflicted with Rove’s deeply held faith that Romney would win. (Lest anyone think it was only recently that Rove was out of touch with electoral math reality, it’s worth remembering that he had George W. Bush waste precious final hours campaigning in California in the waning days of the 2000 election—a state Bush lost to Gore by a whopping 12 points—as Florida hung in the balance.)
0 Replies
 
izzythepush
 
  3  
Reply Thu 4 Aug, 2016 03:28 am
@cicerone imposter,
cicerone imposter wrote:

I agree it's going to fluctuate, but I don't see how Trump can gain when he continues to show his ignorance and bigotry.


All it takes is one major terrorist attack and the voters will flood to him, believing only a strong leader like him has the will to thwart such attacks.

He won't, he'll only make it worse like Dubya did after 9/11, but that doesn't matter to those who want simple solutions.
izzythepush
 
  2  
Reply Thu 4 Aug, 2016 03:32 am
@rosborne979,
rosborne979 wrote:
In past elections I have found the consensus of Vegas Bookies predictions to be the most accurate,


Vegas bookies? Are you sure? Has the law changed?

Quote:
It’s illegal to bet on political races in the United States. But overseas, almost anything goes. William Hill, Britain’s second-largest betting house, has already received more than 200,000 pounds ($290,000) in bets on the presidential race, spokeswoman Serena Momberg said.


http://abcnews.go.com/Politics/story?id=122622&page=1
0 Replies
 
Setanta
 
  5  
Reply Thu 4 Aug, 2016 04:01 am
The bookies in the U.S. will give odds, they just won't take bets. As they have no money riding on it, their odds might be considered dubious. Nevertheless, they have a pretty good track record.
0 Replies
 
thack45
 
  2  
Reply Thu 4 Aug, 2016 10:12 am
@izzythepush,
I had the exact same thought last night. Wondering if there's any strategy in the works.. Daesh has sad they want war with the west, and they know who's more inclined to bring it
izzythepush
 
  0  
Reply Thu 4 Aug, 2016 01:55 pm
@thack45,
That's what worries me, if a week is a long time in politics it's an eternity to November. you're right, the Jihadis would welcome a Trump presidency.
0 Replies
 
Robert Gentel
 
  1  
Reply Mon 8 Aug, 2016 10:48 am
@Robert Gentel,
At present, the consensus from the above data sources is that Clinton has over a 80% chance of winning the election. Trump has continued to poll very badly and with 75% of Americans disapproving of his approach to the Khan family the Democrats have come out well ahead in convention bounces and continue to gain.
cicerone imposter
 
  1  
Reply Mon 8 Aug, 2016 11:32 am
@Robert Gentel,
According the all the polls, Trump is toast. That's reassuring.
http://www.cnn.com/2016/08/08/politics/hillary-clinton-donald-trump-cnn-poll-of-polls/index.html
0 Replies
 
Robert Gentel
 
  1  
Reply Tue 6 Sep, 2016 12:46 pm
@Robert Gentel,
Election Update: Clinton’s Lead Keeps Shrinking
revelette2
 
  0  
Reply Tue 6 Sep, 2016 04:38 pm
@Robert Gentel,
I think perhaps it is because she has largely disappeared and Trump went on that Mexico trip and managed not to insult the Mexico President by not bringing up the wall in the press conference and got tons of good press on account of it. Such a low bar is set for him in comparison for Hillary.
vikorr
 
  1  
Reply Tue 6 Sep, 2016 04:43 pm
@izzythepush,
Quote:
All it takes is one major terrorist attack and the voters will flood to him, believing only a strong leader like him has the will to thwart such attacks.

He won't, he'll only make it worse like Dubya did after 9/11, but that doesn't matter to those who want simple solutions.
I agree with this.
0 Replies
 
nimh
 
  2  
Reply Tue 6 Sep, 2016 05:00 pm
@Robert Gentel,
Daily Kos launched something like this last month too:

President forecast / Senate forecast / Governor forecast

More detail about their presidential forecast: Hillary Clinton currently has a 70% chance of winning the presidency

You gotta look pretty hard to find the link to their methodology, but here it is.
0 Replies
 
cicerone imposter
 
  0  
Reply Tue 6 Sep, 2016 05:47 pm
@revelette2,
It's not only the wall; it's about the criminals and rapists that Mexico sends to this country.
0 Replies
 
Miller
 
  1  
Reply Tue 6 Sep, 2016 06:02 pm
The latest pole ( reported on TV tonight): Trump=46%...Mrs. Clinton= 43%.

Not factored in are Johnson and Stein, who will be/or could be the deciding factors in determining the next President.
0 Replies
 
ehBeth
 
  2  
Reply Wed 7 Sep, 2016 10:44 am
Clinton campaign data cruncher

Elan Kriegel

longish, but fascinating read

a teaser

Quote:
overnight, in some of the few hours that headquarters isn’t whirring with activity, the team’s computers run 400,000 simulations of the fall campaign in what amounts to a massive stress-test of the possibilities on Nov. 8. That way, in morning calls with senior staff, Kriegel can deliver any key findings.

0 Replies
 
oralloy
 
  0  
Reply Mon 31 Oct, 2016 09:11 pm
@Robert Gentel,
Robert Gentel wrote:
If you have any other good sources for such things please share!

http://electionbettingodds.com
http://electionbettingodds.com/congress.html

I also like the Polls Plus section of FiveThirtyEight.
0 Replies
 
 

Related Topics

Obama '08? - Discussion by sozobe
Let's get rid of the Electoral College - Discussion by Robert Gentel
McCain's VP: - Discussion by Cycloptichorn
McCain is blowing his election chances. - Discussion by McGentrix
Food Stamp Turkeys - Discussion by H2O MAN
The 2008 Democrat Convention - Discussion by Lash
Snowdon is a dummy - Discussion by cicerone imposter
GAFFNEY: Whose side is Obama on? - Discussion by gungasnake
 
  1. Forums
  2. » Data-Driven Predictions for the 2016 US Presidential Election
Copyright © 2019 MadLab, LLC :: Terms of Service :: Privacy Policy :: Page generated in 0.08 seconds on 03/18/2019 at 12:09:26