 |
 |
It has been some time since I’ve done a general election match-up using state head-to-head polls. Way back last October Hillary Clinton held a good lead over Donald Trump and would have defeated Trump with 96.6% probability. But these results were from back in the days when we had few polls matching up Trump and Clinton. We have many more now.
From a analysis of state polls through yesterday, and after 100,000 simulated elections, Clinton takes 100,000 wins. On average, Clinton received 354 to Trump’s 184 electoral votes. In other words, if the election was held now, Clinton would certainly win, and likely in a landslide.
Here is the distribution of electoral votes [FAQ] from the simulations:
The electoral map for Trump is grim. He loses almost all the swing states: Florida, Virginia, North Carolina, Pennsylvania, Ohio, Michigan, Wisconsin, Missouri, and Iowa. On top of that, he loses Kansas, New Mexico, Arizona, and Kentucky!
The Arizona and Kansas results may be an artifact of one large outlier poll in each state. New Mexico and Kentucky are both based on single PPP polls. Still, Trump can take these four states and still loses handily.
One interesting finding is that Maine splits its districts, although the state goes for Clinton overall.
Today, Donald Trump met with G.O.P. lawmakers, telling them he would put Oregon, Washington, Connecticut and Michigan “in play.” But apparently not. Aside from turning Nevada from blue to red, and one CD in Maine from blue to red, Trump isn’t putting purple states in play at this point. It seems unlikely that solid blue states like Oregon, Washington and Connecticut are going to turn red.
The long term trends in this race can be seen from a series of elections simulated every seven days using polls from 07 Jul 2015 to 07 Jul 2016, and including polls from the preceding month (FAQ). Clearly, Trump has never had more than a small chance (perhaps 15%) of beating Clinton. Even the “rebound” that followed him becoming the presumptive G.O.P. nominee was not much of a rebound and transient.
An animated sequence of maps and electoral vote distributions can be seen here
Ten most probable electoral vote outcomes for Clinton (full distribution here):
- 356 electoral votes with a 3.65% probability
- 362 electoral votes with a 3.63% probability
- 355 electoral votes with a 3.56% probability
- 366 electoral votes with a 3.53% probability
- 361 electoral votes with a 3.26% probability
- 357 electoral votes with a 3.21% probability
- 367 electoral votes with a 3.00% probability
- 351 electoral votes with a 2.95% probability
- 352 electoral votes with a 2.88% probability
- 349 electoral votes with a 2.74% probability
After 100,000 simulations:
- Clinton wins 100.0%, Trump wins 0.0%.
- Average (SE) EC votes for Clinton: 354.4 (13.9)
- Average (SE) EC votes for Trump: 183.6 (13.9)
- Median (95% CI) EC votes for Clinton: 355 (325, 379)
- Median (95% CI) EC votes for Trump: 183 (159, 213)
Each column of this table shows the electoral vote total aggregated by different criteria for the probability of winning a state (Safe=100%, Strong=90%+, Leans=60%+, Weak=50%+):
| Threshold | Safe | + Strong | + Leans | + Weak |
|---|---|---|---|---|
| Safe Clinton | 131 | |||
| Strong Clinton | 173 | 304 | ||
| Leans Clinton | 62 | 62 | 366 | |
| Weak Clinton | 0 | 0 | 0 | 366 |
| Weak Trump | 7 | 7 | 7 | 172 |
| Leans Trump | 12 | 12 | 165 | |
| Strong Trump | 100 | 153 | ||
| Safe Trump | 53 |
This table summarizes results by state. Click on the poll count to see the individual polls included for the state.
| 1 | 0 | EC | # | Total | % | % | Clinton | Trump | |
|---|---|---|---|---|---|---|---|---|---|
| 2 | 8 | Votes | polls | Votes | Clinton | Trump | % wins | % wins | |
| AL | 9 | 0 | (0) | (100) | |||||
| AK | 3 | 0 | (0) | (100) | |||||
| AZ | 11 | 2 | 1520 | 50.6 | 49.4 | 61.8 | 38.2 | ||
| AR | 6 | 1 | 623 | 43.3 | 56.7 | 0.9 | 99.1 | ||
| CA | 55 | 1 | 1211 | 65.4 | 34.6 | 100.0 | 0.0 | ||
| CO | 9 | 1* | 1073 | 43.5 | 56.5 | 0.2 | 99.8 | ||
| CT | 7 | 1* | 1024 | 53.2 | 46.8 | 92.7 | 7.3 | ||
| DE | 3 | 0* | (100) | (0) | |||||
| DC | 3 | 0* | (100) | (0) | |||||
| FL | 29 | 4 | 3069 | 54.3 | 45.7 | 100.0 | 0.0 | ||
| GA | 16 | 2* | 1100 | 46.8 | 53.2 | 6.9 | 93.1 | ||
| HI | 4 | 0* | (100) | (0) | |||||
| ID | 4 | 1* | 475 | 37.1 | 62.9 | 0.0 | 100.0 | ||
| IL | 20 | 1 | 570 | 62.1 | 37.9 | 100.0 | 0.0 | ||
| IN | 11 | 0 | (0) | (100) | |||||
| IA | 6 | 3 | 1697 | 51.4 | 48.6 | 79.7 | 20.3 | ||
| KS | 6 | 1* | 342 | 54.4 | 45.6 | 87.4 | 12.6 | ||
| KY | 8 | 1* | 964 | 51.8 | 48.2 | 78.4 | 21.6 | ||
| LA | 8 | 1* | 549 | 41.0 | 59.0 | 0.1 | 99.9 | ||
| ME | 2 | 1 | 366 | 54.6 | 45.4 | 89.4 | 10.6 | ||
| ME1 | 1 | 1 | 201 | 59.2 | 40.8 | 96.8 | 3.2 | ||
| ME2 | 1 | 1 | 162 | 49.4 | 50.6 | 45.5 | 54.5 | ||
| MD | 10 | 2* | 3089 | 69.6 | 30.4 | 100.0 | 0.0 | ||
| MA | 11 | 1* | 431 | 63.6 | 36.4 | 100.0 | 0.0 | ||
| MI | 16 | 1 | 472 | 61.0 | 39.0 | 100.0 | 0.0 | ||
| MN | 10 | 1* | 664 | 57.8 | 42.2 | 99.8 | 0.2 | ||
| MS | 6 | 1* | 488 | 50.0 | 50.0 | 49.3 | 50.7 | ||
| MO | 10 | 1* | 553 | 51.2 | 48.8 | 65.2 | 34.8 | ||
| MT | 3 | 1* | 354 | 36.7 | 63.3 | 0.0 | 100.0 | ||
| NE | 2 | 0* | (0) | (100) | |||||
| NE1 | 1 | 0* | (0) | (100) | |||||
| NE2 | 1 | 0* | (0) | (100) | |||||
| NE3 | 1 | 0* | (0) | (100) | |||||
| NV | 6 | 1* | 539 | 48.2 | 51.8 | 28.4 | 71.6 | ||
| NH | 4 | 1 | 474 | 52.5 | 47.5 | 78.5 | 21.5 | ||
| NJ | 14 | 3* | 2726 | 54.7 | 45.3 | 100.0 | 0.0 | ||
| NM | 5 | 1* | 594 | 55.4 | 44.6 | 96.9 | 3.1 | ||
| NY | 29 | 1* | 685 | 62.6 | 37.4 | 100.0 | 0.0 | ||
| NC | 15 | 2 | 1302 | 51.6 | 48.4 | 79.1 | 20.9 | ||
| ND | 3 | 0 | (0) | (100) | |||||
| OH | 18 | 3 | 1777 | 52.6 | 47.4 | 93.4 | 6.6 | ||
| OK | 7 | 0 | (0) | (100) | |||||
| OR | 7 | 2* | 1243 | 53.3 | 46.7 | 95.7 | 4.3 | ||
| PA | 20 | 3 | 2043 | 53.7 | 46.3 | 99.1 | 0.9 | ||
| RI | 4 | 0 | (100) | (0) | |||||
| SC | 9 | 3* | 2672 | 46.1 | 53.9 | 0.2 | 99.8 | ||
| SD | 3 | 0* | (0) | (100) | |||||
| TN | 11 | 1* | 790 | 44.3 | 55.7 | 1.1 | 98.9 | ||
| TX | 38 | 2 | 1515 | 44.9 | 55.1 | 0.2 | 99.8 | ||
| UT | 6 | 2 | 1254 | 47.7 | 52.3 | 12.5 | 87.5 | ||
| VT | 3 | 0 | (100) | (0) | |||||
| VA | 13 | 2 | 1312 | 53.0 | 47.0 | 94.2 | 5.8 | ||
| WA | 12 | 1 | 584 | 57.0 | 43.0 | 99.3 | 0.7 | ||
| WV | 5 | 1* | 1045 | 34.4 | 65.6 | 0.0 | 100.0 | ||
| WI | 10 | 3 | 2007 | 54.3 | 45.7 | 99.7 | 0.3 | ||
| WY | 3 | 0 | (0) | (100) |
* An older poll was used (i.e. no recent polls exist).
Details of the methods are given in the FAQ.
The most recent analysis in this match-up can be found from this page.
Republicans’ desperation is in plain sight for all to see. They’re demanding the FBI investigate whether Clinton lied under oath, and insisting she be denied classified briefings, even as they plot to ditch their own nominee. These are the dying gasps of a party that’s about to go under.
Oregon. Sure he will.
Where Cheeto Jesus managed to lose a third of Republican votes despite running unopposed. Where he had to run and hide in Eugene to avoid the embarrassment of a rally in Beaverton where protesters would outnumber supporters. Where motor-voter has already, and will continue to shift the vote further away from the Republican base. Sure he will.
#NoRagrets
While perusing these numbers….
I just came back from a quick trip to the Post Office and on the way home whatever not-Rush RW-er on the radio here in Puget Sound said, “Blah, blah, blah, the American people blah, blah, blah…sick and tired, blah, blah, blah…That’s why Trump is doing well.”
Doing well by what measure? Like the Cincinnati Reds have won two in a row so they are doing well. They only have the SECOND worst record in the league but they are doing well.
Well like, The patient is terminal and the cancer has spread to several internal organs but TODAY, his fever is down half a degree so he’s doing well.
And The Liar listens and believes. Skewed Polls, Trump’s doing well, He’s going win!
@3,
The idiots will demand a recount.
And when the results are certified they’ll insist that the dirty hippies who control everything stole the election.
Part of it is an ongoing theme intended to trigger the dispossessed angry white racists. But probably the biggest part of it is all about delegitimizing the next Clinton Presidency. They still hope to retain control of Congress. And they want to go right on shutting down the government to stir fund raising. They need excuses. That’s the new Republican party motto.
The only good thing about Hillary Clinton is that she’s not Donald Trump, and the only good thing about Donald Trump is he’s not Hillary Clinton.
With the 2016 presidential contest, we have the worst set of candidates since 1972.
@5 That’s all a despicable lie.
So I guess Cheeto Jesus took some of his free time between meetings to head over to the Senate and loudly threaten Jeff Flake’s re-election campaign.
Only Flake’s not up for re-election this year.
Attaboy! Way to help us nail down AZ!
It’s also being reported in the WAPO that during his brief but belligerent encounter with Republican Senators, el Douche predicted he would win Illinois in a landslide – a state where he trails by 24%.
Very persuasive!
I’m pretty sure this analysis underestimates Crooked Hillary Clinton’s chances of victory.
She’s only “Crooked Hillary” in the sick, twisted minds of Trumpty Dumbty and his mentally impaired supporters.
Sloppy, what are you afraid that Hillary will do as president, that would be different from any other Democratic president?
Sensing the inevitability of his defeat, I predict Cheeto Jesus will fake a disabling injury and sit out the last two or three weeks.
But it’ll have to be the most awesome and macho of injuries.
Priapism?
@11,
she’ll insist on leaving the seat down.
The horror!
Looks like Trump is going to poison his ticket for independents and moderates.
http://www.aol.com/article/201...../21426051/
Roger Rabbit Commentary: Awfully hard to see any significant number of Sanders supporters going over to the GOP in this scenario, no matter how much they dislike Hillary.
A Trump-Gingrich ticket has one advantage, though: Between them they have enough ex-wives to fill all the cabinet positions.
(a) Glad you’re doing these again. I know it’s a lot of work, so I’m glad you finally had the time.
(b) I’m surprised how many states remain unpolled. I know it’s still early, but if I was at a polling firm, I’d want to be the first to be able say Indiana (or whatever) is in play.
A little different electoral perspective… The one Puddy is following… http://electiongraphs.com/2016ec/
@16 What’s different about it? Even your own graph shows Clinton kicking Drumpf’s butt. And that’s before the voters see the first mushroom cloud ads.
@17
Yes, an analysis that says the “BEST” trump could possibly do is lose.
And yet….
We still have four months of the Liar trying to convince himself that Hillary Clinton isn’t his next president.
Too damn funny.
What’s different with that site R senile? Their analysis oh senile one! More thorough and very different! It changes almost every day!
Did checkmate leave a smelly load again? With the information left by James Comey, the lies from Careless Crooked Heilary make excellent commercials!
Drumpf is throwing the game. He doesn’t want to be president.
BTW, guys, is this the same Monte Carlo analysis you ran back in 2012? You know, the one that showed Romney getting trounced even as the GOPeckers believed with all their hearts he was going to win?
(True story: A cigar friend I know bet multiple boxes of cigars on Romney winning. And lost. Then again, this is what happens when you watch Fox News 24/7 and believe that “Unskewed Polls” idiot).
There are those lazy Democratic voters that just can’t be bothered to show up every two years to just vote. So make it look like Hillary has it in the bag, start doing a Bernie the last 8 weeks of the election maybe helped by a souring economy, and suddenly the one pole that matters…the one done with real voters in November and things may come out differently. Like Britain America may wake up and find we voted who into office some Libertarian guy?!? I noticed the polls left him out. He’s going to be in the race, and hopefully he will even make it to the debates. Or will both major parties try to keep him out. Maybe not this time, as the Republican’s do seem to want to eat their candidate.
The polls after the parties conventions will be more meaningful, and there will be a lot of noise and thunder, but it really won’t matter until after the World Series. How about them Mariners!!!
PS and like an Olympic runner who has peaked before the Olympic’s and has no room to improve that is where Hillary is at. She is at her peak and can’t improve it or not much.
NYT article that Drumpf hasn’t ruled out quiting or not taking the nomination.
I said before that this “situation” that the Repuke voters created isn’t the best thing for Drumps biography and hurts him the most. Backing out of deals and claiming bankruptcy is what he knows best in life, so quoting the race seems like a normal option for this moron.
@20. I guess I should have read all the comments before posting @21.
Seems like Drump is petering out but then again I got rid of my cable subscription 2 weeks ago and no longer watch much if any news on TV
@16/@19 Thanks for the link! I’m the guy who runs ElectionGraphs.com. Glad you like it. Darryl and I sometimes actually share notes. He just sent me a couple polls I missed that I’ll be adding in this weekend for instance…
Each of the ways we do things have their own pros and cons. I of course like mine, but the big thing that Darryl does that I do not is use the uncertainty in the polls to run a simulation and generate actual odds of each candidate winning, where I just present the range of the swing if all the states closer than 5% go one way or the other.
And both of us are looking at “if the election was held today”. People like Sam Wang and Nate Silver also factor in odds for how much things can change between now and Election day, which is also an important way to look at things. While every estimate I have seen of “election held today” essentially says there would be a nearly 100% chance of a Clinton win, all the models that allow for changes between now and November give Clinton a lead, but non-trivial odds for Trump (between 15% and 25% generally).
Rather than pick one of us as a favorite and ignore the rest, you actually can get a lot of good insight at looking at all the different folks doing election modeling and seeing how they differ.
Lots of great data and analysis out there!
Puddy, et al.
What Sam@25 said.
Sam does an excellent job of poll aggregation & analysis, and presents the results beautifully. I highly recommend it!
Generally, they will be in good agreement on the hypothetical outcome. You can learn things from Sam’s analyses that you won’t get from my analyses, and vice versa.
Darryl and Sam,
I’m starting to see polls showing results on a three-way race between Clinton, Trump and Johnson. Would it be possible for you guys to run your respective analyses using three-way numbers as opposed to just head-to-head numbers?
@27 – I can’t speak for how Darryl deals with 3P results, but I can talk to how I deal with it.
At the moment I already include polls that include Johnson (Libertarian) and Stein (Green) in my analysis. They just get folded in with all the other polls. The basic analysis remains a two party analysis though where I am looking at Republican-Democrat margins. The presence of the 3rd and 4th parties just potentially changes that margin.
If and when we start getting polls that show a third party with a poll average showing them in second place in any state, the way I display results will actually need to change, because the “margin” in that state will no longer be the Republican number minus the Democratic number (or vice versa).
If the poll average for a 3rd party candidate is actually within 5% of the lead in any state, then things on my site would have to change even more drastically, as it would need to take into account the possibility of them actually winning a state.
I have a mental plan on what I would need to do on my site if either of these things happened, but frankly it would be quite a bit of work, and would not actually make a difference in the analysis in the absence of the things above, so I won’t pull the trigger until it looks like one or both of the things above are actually going to happen.
At the moment Johnson seems to be polling around 10% in his best states, and Stein around 5% in hers… which is a long way from coming in second in any state, let alone winning any. Usually this time in the cycle is when 3rd parties peak, so unless something unusual happens, it looks like revamping the site to explicitly take into account the possibility of 3rd parties coming in 2nd place or 1st place in some states won’t actually be necessary.
Having said that, 2016 has been a highly unusual year, so I won’t rule anything out and have those contingency plans in my pocket to use if indeed Johnson or Stein start surging.