 |
 |
My previous analysis of the Biden-Trump head-to-head polls was way back in early June, in part because the dynamics of the race didn’t seem to be changing much. Trump is slowly slipping in the polls, as the COVID-19 epidemic worsens in the U.S., as Trump resists movements for police reforms, and as we learn bizarre things like Trump not responding to Russian bounties on American soldiers in Afghanistan.
After 100,000 simulated elections, the results are not surprising. Biden wins 100,000 times and Trump wins 0 times. Biden received (on average) 375 to Trump’s 163 electoral votes. In an election held now, Biden would have a greater than 99.9% probability of winning and Trump would have a less than 0.1% probability of winning.
Basically, if the election was today, Biden would win in a landslide.
Trump has a serious electoral college problem. Many swing states are now trending strongly for Biden, including Florida, Michigan, and Wisconsin. Other swing states are pretty solidly in Biden’s camp, like Arizona, Colorado, New Hampshire, New Mexico, North Carolina, Pennsylvania, and Virginia. Even Ohio trends slightly toward Biden. Iowa and Texas are in toss-up territory, but leaning slightly toward Trump. Remarkable!
The trend over the past couple months are clear from from a series of elections simulated every seven days using polls from Jul 2019 to the present, and, at each point in time, including polls from the preceding month (FAQ).
There is, of course, plenty of time to turn this trend around. Still, Trump’s bizarre and impulsive behavior will make doing that very difficult.
Back to the analysis for what would happen in an election today, the ten most probable electoral vote outcomes for Biden (full distribution here) are:
- 375 electoral votes with a 5.26% probability
- 369 electoral votes with a 4.68% probability
- 351 electoral votes with a 3.62% probability
- 357 electoral votes with a 3.61% probability
- 374 electoral votes with a 3.21% probability
- 368 electoral votes with a 2.87% probability
- 413 electoral votes with a 2.57% probability
- 363 electoral votes with a 2.49% probability
- 356 electoral votes with a 2.38% probability
- 378 electoral votes with a 2.24% probability
Here is the entire distribution of electoral votes [FAQ] from the simulations:
After 100,000 simulations:
- Biden wins greater than 99.9%, Trump wins less than 0.1%.
- Average (SE) EC votes for Biden: 374.6 (22.1)
- Average (SE) EC votes for Trump: 163.4 (22.1)
- Median (95% CI) EC votes for Biden: 372 (336, 418)
- Median (95% CI) EC votes for Trump: 166 (120, 202)
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 Biden | 234 | |||
| Strong Biden | 94 | 328 | ||
| Leans Biden | 23 | 23 | 351 | |
| Weak Biden | 18 | 18 | 18 | 369 |
| Weak Trump | 6 | 6 | 6 | 169 |
| Leans Trump | 57 | 57 | 163 | |
| Strong Trump | 53 | 106 | ||
| Safe Trump | 53 |
This table summarizes results by state. Click on the poll count to see the individual polls included for the state.
| 1 | 1 | EC | # | Total | % | % | Biden | Trump | |
|---|---|---|---|---|---|---|---|---|---|
| 6 | 2 | Votes | polls | Votes | Biden | Trump | % wins | % wins | |
| AL | 9 | 1* | 553 | 42.3 | 57.7 | 0.4 | 99.6 | ||
| AK | 3 | 1* | 272 | 47.1 | 52.9 | 23.8 | 76.2 | ||
| AZ | 11 | 7 | 4087 | 51.7 | 48.3 | 94.6 | 5.4 | ||
| AR | 6 | 1 | 799 | 48.9 | 51.1 | 32.8 | 67.2 | ||
| CA | 55 | 3* | 2118 | 65.1 | 34.9 | 100.0 | 0.0 | ||
| CO | 9 | 1 | 798 | 58.9 | 41.1 | 100.0 | 0.0 | ||
| CT | 7 | 1* | 814 | 62.9 | 37.1 | 100.0 | 0.0 | ||
| DE | 3 | 1* | 394 | 58.1 | 41.9 | 98.7 | 1.3 | ||
| DC | 3 | 0* | (100) | (0) | |||||
| FL | 29 | 8 | 6410 | 54.1 | 45.9 | 100.0 | 0.0 | ||
| GA | 16 | 3 | 2243 | 51.4 | 48.6 | 83.7 | 16.3 | ||
| HI | 4 | 0 | (100) | (0) | |||||
| ID | 4 | 0 | (0) | (100) | |||||
| IL | 20 | 0 | (100) | (0) | |||||
| IN | 11 | 1* | 929 | 42.8 | 57.2 | 0.1 | 99.9 | ||
| IA | 6 | 2 | 1383 | 49.7 | 50.3 | 44.2 | 55.8 | ||
| KS | 6 | 1* | 643 | 43.5 | 56.5 | 1.1 | 98.9 | ||
| KY | 8 | 1 | 844 | 39.3 | 60.7 | 0.0 | 100.0 | ||
| LA | 8 | 0 | (0) | (100) | |||||
| ME | 2 | 1 | 971 | 55.8 | 44.2 | 99.6 | 0.4 | ||
| ME1 | 1 | 1* | 459 | 60.3 | 39.7 | 99.9 | 0.1 | ||
| ME2 | 1 | 1* | 438 | 51.6 | 48.4 | 69.3 | 30.7 | ||
| MD | 10 | 1* | 765 | 62.6 | 37.4 | 100.0 | 0.0 | ||
| MA | 11 | 2* | 1546 | 66.2 | 33.8 | 100.0 | 0.0 | ||
| MI | 16 | 8 | 5621 | 54.3 | 45.7 | 100.0 | 0.0 | ||
| MN | 10 | 1 | 546 | 59.3 | 40.7 | 99.9 | 0.1 | ||
| MS | 6 | 1* | 517 | 45.1 | 54.9 | 6.0 | 94.0 | ||
| MO | 10 | 2 | 1835 | 47.9 | 52.1 | 10.2 | 89.8 | ||
| MT | 3 | 1 | 466 | 42.1 | 57.9 | 0.8 | 99.2 | ||
| NE | 2 | 0 | (0) | (100) | |||||
| NE1 | 1 | 0 | (0) | (100) | |||||
| NE2 | 1 | 1* | 417 | 55.9 | 44.1 | 96.0 | 4.0 | ||
| NE3 | 1 | 0* | (0) | (100) | |||||
| NV | 6 | 1* | 717 | 52.2 | 47.8 | 79.2 | 20.8 | ||
| NH | 4 | 1 | 981 | 53.7 | 46.3 | 95.1 | 4.9 | ||
| NJ | 14 | 2* | 1458 | 61.7 | 38.3 | 100.0 | 0.0 | ||
| NM | 5 | 1 | 681 | 57.6 | 42.4 | 99.7 | 0.3 | ||
| NY | 29 | 1 | 717 | 64.0 | 36.0 | 100.0 | 0.0 | ||
| NC | 15 | 8 | 5738 | 51.4 | 48.6 | 94.1 | 5.9 | ||
| ND | 3 | 1* | 372 | 40.9 | 59.1 | 0.6 | 99.4 | ||
| OH | 18 | 1 | 1037 | 50.5 | 49.5 | 59.0 | 41.0 | ||
| OK | 7 | 2* | 930 | 36.0 | 64.0 | 0.0 | 100.0 | ||
| OR | 7 | 0* | (100) | (0) | |||||
| PA | 20 | 7 | 4924 | 53.8 | 46.2 | 100.0 | 0.0 | ||
| RI | 4 | 0 | (100) | (0) | |||||
| SC | 9 | 1* | 555 | 44.7 | 55.3 | 3.5 | 96.5 | ||
| SD | 3 | 0* | (0) | (100) | |||||
| TN | 11 | 2* | 1407 | 43.6 | 56.4 | 0.0 | 100.0 | ||
| TX | 38 | 4 | 3531 | 49.4 | 50.6 | 31.4 | 68.6 | ||
| UT | 6 | 1* | 927 | 44.2 | 55.8 | 0.8 | 99.2 | ||
| VT | 3 | 0* | (100) | (0) | |||||
| VA | 13 | 1* | 507 | 56.6 | 43.4 | 97.9 | 2.1 | ||
| WA | 12 | 2* | 1493 | 62.4 | 37.6 | 100.0 | 0.0 | ||
| WV | 5 | 1* | 495 | 33.3 | 66.7 | 0.0 | 100.0 | ||
| WI | 10 | 7 | 4131 | 54.0 | 46.0 | 100.0 | 0.0 | ||
| 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.
NOTE: This is NOT an open thread. Off topic comments will disappear.
What bothers me about these kinds of analyses is they don’t take into account voter suppression or Russian interference.
Voter suppression may well be overcome by the determination of voters to stand in line for many hours, if necessary, to vote this tinpot wannabe out of office.
And the public is growing so disenchanted with Trump, and Biden is opening up such a wide lead, that even a torrent of Russian disinformation will prove ineffective. But what about Russian hacking of election results?
And then there’s the question of whether Trump will leave office peacefully if he loses. Of course, the best assurance of that is a blowout that shuts the door on any debate over who won. It’s not enough for Biden to eke out a win; he has to win decisively, or this country could be in for a very ugly transition.
Still, my hopes are rising that we can soon put this long national nightmare behind us.
[OFF TOPIC]
At least five GOP senators have said they will not go to Jacksonville due to either coronavirus concerns or political reasons.
New Hampshire Gov. Chris Sununu said he’s sure Trump’s outdoor rally in Portsmouth Saturday can be pulled off safely without a mandatory mask order, but he won’t be there. “I’m not going to put myself in the middle of a crowd of thousands of people, if that’s your question specifically,” the Republican told reporters.
2,
Which to me always begs the question: Why the fans?
I called my little brother July 4th in New Jersey. He received a presidential poll call. He declined to answer. Reason, his friend answered a similar poll answering he supported Trump. The next week the friend had protestors around their house and was receiving threatening calls from BiteME! jockstraps.
As the senile idiot wabbit said in 2016, we like our candidate better than yours dot bombed.
Trust polls? NOPE!
NOTE: This is NOT an open thread. Off topic comments will disappear.
@ 2, 3, 4 are off-topic. Imagine if I had posted them.
Two comments about Trump’s predicament:
• He has been so awful that CNN’s query whether Ohio is still a swing state has been answered, as the Buckeye State has swung from red to blue. There are fewer than 400 miles separating hardscrabble Scranton, PA from Cleveland Rocks.
• Some Republicans are now beginning to focus on whether Biden’s cabinet will use the 25th Amendment to install a president Kamala Harris.
Still too early for me to draw any conclusions just yet.
One thing I intend to pay close attention to are the states that are polling close but for which there are not very many polls in the field. That partially addresses Roger’s concern, at least with respect to foreign interference.
Using private popular social media platforms like FB, twitter, Insta, etc. is obviously going to be much more difficult for them this year. Hundreds of thousands of fake accounts have been liquidated since 2016. And the platforms have redesigned their algorithms to better identify coordinated inauthentic automated activity. But that still leaves other avenues open, including apps like Parler or the Trump2020 app. Whether or not these ever become widespread in the way they would need to is probably not likely. For instance, while we know the Russian GRU waged a coordinated campaign to reduce turnout among African American voters in 2016, I doubt Parler will be an effective tool for that in 2020.
Still, as Roger points out, a little bit of unanticipated targeted vote suppression, combined with sparse polling in a few “close” states could swing the electoral college. For instance, VA is a concern to me. There are a lot of significant barriers to voting in the commonwealth. But at least Virginia Democrats in the assembly reversed a host of Rapepublican obstacles. Automatic registration, vote by mail, election day holiday, and easy ID will all be in effect in 2020. Turnout may not explode. But it will be more difficult to suppress. Although state laws can do little when local county election officials are determined to make mischief.
Colorado, Minnesota, Nevada, and New Mexico are also concerning. Hopefully we will see more polling data from more of these states after the nominating conventions. At the same time the percentages of undecideds will begin to drop reducing some of the uncertainty.
25th? Nah! Skip it. And skip Harris. We’re going straight for Julius Malema.
After all, we gave you guys Putin. It’s only fair.
My mistake. Not an open thread. My post was therefore off topic.
Just for you, in March, they replace Biden With Hillary Clinton. She and the Democrats treat republicans as republicans treated democrats and go on to have 24 years of amazing progressive accomplishments despite daily howls for investigation and impeachment by the minority republicans in the house and senate.
LRepublicans are standing by Trump, with 91 percent saying they approve of the job he’s doing, up from 85 percent last month.
But the president has reached new lows among independents and Democrats. Thirty-three percent of independents say they approve of the job Trump is doing, a 6-point drop from the same survey in early June.
Only 2 percent of Democrats approve of the job Trump is doing, the lowest on record. The disparity between Trump’s job approval rating among Republicans and Democrats is the greatest Gallup has ever recorded.
>> Republicans are all in.
Puddybud @5,
Trust polls? NOPE!
Sounds like your brother is as much of an imbecile as you are!
The trick is to be smart enough to tell real pollsters from push pollsters/scammers/political campaigns.
It’s not really that hard.
@5. your brother had a buddy…. Right.
You might as well say “My brother’s best friend’s sister’s boyfriend’s brother’s girlfriend heard from this guy who knows this kid who’s going with the girl who had protestors around their house and was receiving threatening calls.”
It’s as credible.
Maybe the trump lawn signs he didn’t mention factored in?
Roger @1,
“What bothers me about these kinds of analyses is they don’t take into account voter suppression or Russian interference.”
Voter suppression is a real issue at the margins. That is part of what happened in, for example, Wisconsin in 2016. No doubt “Republicans” will be working overtime exploiting the COVID-19 epidemic to suppress voter turnout for 2020.
The polls DO take into account the type of Russian interference that we saw in 2016. Those efforts were largely social media disinformation campaigns and the strategic release of stolen emails. The objective was to persuade voters, so that polls will reflect the cumulative effect up to the point the poll is taken.
“But what about Russian hacking of election results?
The is no evidence of any foreign government hacking the election results in 2016.
That said, there were 20 or so states where the voter registration system was targeted by Russian hackers. Registration data were stolen from a small number of states. I don’t believe there is any evidence of voter registration databases being changed by hackers in 2016.
In theory, if hackers managed to corrupt voter registration databases (say, to cancel some voter registrations), provisional ballot would reduce the effect on the election.
11,
Coattails. Not always a good thing.
Republicans appearing on the ballot this November have almost no way to put space between themselves and Styupid Hitler. Leaving them to either go Full Trumpalo Styupid (like our Trolls4Trump) or stand in the corner hiding behind the curtains.
We shall see.
@6 Judd Gregg is an always wrong wing nutjob..
If progressives were truly serious about power, Nancy Pelosi wouldn’t be Speaker. Steny Hoyer wouldn’t be Majority Leader. Biden’s cabinet will be a corporate Dem replay of Obama’s.
Again, Judd Gregg is a joke.
16,
The good news, at least, is that The Q Clearance Pussy is already getting started on pre-configuring his future life in defeat.
Remember when two-term President Obama insisted on holding open ACA until the GOP caucus stormed out, and insisted on adopting the industry-approved MittensCare plan, overruling his own team of advisers to give Boehner what he wanted? Remember when Obama appointed Judd Gregg to Sec. of Commerce?
What was he again? Oh, that’s right.
A radical Mau Mau Kenyan Usurper Black Panther Death Panelist.
At some point late in the McCain disaster, as Caribou Barbie was traveling coast to coast to Aw Shucksing her way through every interview, Judd Gregg was telling a Rotary Club in Concord that Obama was going to appoint Angela Davis AG.
He’s a total piece of shit. But then again so are they all.
It isn’t so much that Judd Gregg is wrong about shit, as that he’s a completely disingenuous lying, groveling, hack. He said many of the same things about the Obama administration even after they offered him a slot in the cabinet. Watch your back, indeed.
It’s good in a way that Gregg penned this. Biden’s basic instincts after more than three decades in the Senate are bi-partisan. Here Gregg is doing our work for us, by ably demonstrating how useless and really dangerous those instincts can be.
@5 I don’t answer surveys either. Just gets you on another mailing list and most are asking for money.