So I know I haven't kept this blog up much. Just have been dealing with other things and haven't been very motivated. But I do want to start my first major forecast model for the 2020 presidential election. So, unlike what I predicted last year, Biden DID win the primary, and I would argue that while it seemed like after Nevada my own prediction would come true, the democratic party did put their fingers on the scale and Biden ended up winning after all. Thanks a lot, Jim Clyburn and Barack Obama. But that's another topic for another day. I'm here primarily to model for 2020 now that we knowthat it is Biden vs Trump and getting somewhat close to the election. Given that 538 released a model today and I've been procrastinating, I decided I should get in on this and start making my own model today.
In 2016, my core model wasn't bad. I did not predict the rust belt swing, as it was not foretold in the polls, but I did notice in the 48 hours leading up to the election that polling swung hard for Trump at the last minute. And I ultimately gave Clinton like a 56% chance of winning and Trump a 44% chance. Yeesh. I mean, that was way too close for Clinton. She had this massive advantage in most of my forecasts, but then last minute, gone. Trump surged as the last minuters made up their minds and the energy favored Trump. I doubt that will happen this time. I mean, the way I see it, a lot of people in 2016 were voting against Hillary and the democrats, and I think this time more people are hyper focused on removing Trump no matter who we get. That said "vote blue no matter who" will dominate the energy of this election in my opinion and I'm not expecting a similar fluke to happen in favor of Trump.
That said, I will predict the election the same way I did in 2016. To get into the data...
Aggregate polling: Biden +7.7%
Yeesh. While Trump has a very loyal base, they are not the majority of the country. Regardless of peoples' perceptions of Biden, most people HATE Trump. I mean, between COVID, the second great depression, his authoritarian tendencies, etc., he's making enemies anywhere. Of course, you'll have around 40% of people being for Trump no matter what. You get that in almost every election. I would say Trump is screwed, but I DID live through an election recently where I said Trump was screwed then too, and things did change. So I won't be that firm on it, but I do think it's very unlikely he will win. Just my thoughts on it. Of course the electoral college is what matters. So I'm going to look at the individual states and see how he's doing there. I'll calculate rough probabilities of winning each state based on current data assuming a 4% margin of error, meaning I'm aiming to look at states within 8 points as "in play." From there I'll calculate the most likely outcome and also predict the probabilities of each candidate winning. That said...
Safe electoral votes: Biden - 212, Trump - 115
Ouch. So Biden needs 58 electoral votes to win, and Trump needs 155. This is why I found it so unlikely Trump could win in 2016. Hillary faced similar conditions and only needed to win a handful of swing states while Trump had to sweep the board. Of course, he actually DID sweep the board, which is how he won. So it is possible for Trump to win still IF he has enough of a groundswell nationwide like he did in 2016 to win all of the swing states. Now let's see what will happen this time.
| States | Winner/Margin | SD | % Biden Win | % Trump Win | Electoral Votes if Biden Wins | Electoral Votes if Trump Wins |
| New Hampshire | Biden +9.3% | -2.33 | 99.01% | 0.99% | 216 | 326 |
| Minnesota | Biden +7.0% | -1.75 | 95.99% | 4.01% | 226 | 322 |
| Michigan | Biden +6.7% | -1.68 | 95.35% | 4.65% | 242 | 312 |
| Wisconsin | Biden +6.5% | -1.62 | 94.74% | 5.26% | 252 | 296 |
| Pennsylvania | Biden +6.4% | -1.60 | 94.52% | 5.48% | 272 | 286 |
| Florida | Biden +5.0% | -1.25 | 89.43% | 10.57% | 301 | 266 |
| Nevada | Biden +4.0% | -1.00 | 84.13% | 15.87% | 307 | 237 |
| Ohio | Biden +2.3% | -0.58 | 71.90% | 28.10% | 325 | 231 |
| Arizona | Biden +2.0% | -0.50 | 69.15% | 30.85% | 336 | 213 |
| Maine CD2 | Unknown (0.0%) | 0.00 | 50.00% | 50.00% | 337 | 202 |
| Nebraska CD2 | Unknown (0.0%) | 0.00 | 50.00% | 50.00% | 338 | 201 |
| North Carolina | Trump +0.7% | +0.18 | 42.86% | 57.14% | 353 | 200 |
| Georgia | Trump +1.0% | +0.25 | 40.13% | 59.87% | 369 | 185 |
| Iowa | Trump 1.7% | +0.43 | 33.36% | 66.64% | 375 | 169 |
| Texas | Trump +2.0% | +0.50 | 30.85% | 69.15% | 413 | 163 |
| Missouri | Trump +6.4% | +1.60 | 5.48% | 94.52% | 423 | 125 |
Yikes. First of all, my prediction as of now, 3 months out. Biden has a 95% chance of winning the election as it stands and Trump has a 5% chance. The most likely electoral college map has Biden winning 337 electoral votes and Trump 201.
If these polls are correct, this is going to be a blow out. I haven't seen a lot of deviation from this, by the way. In 2016 Hillary would suffer random dips in poll numbers at times where Trump started coming within spitting distance of winning. This is what happened leading up to the election which is why Trump won. The electoral college, as it exists, HEAVILY favors the democrats. I know the democrats have a lot of salt over 2016 and "muh popular vote", but honestly, the electoral college should theoretically ensure that a republican never wins the presidency again under the current alignment unless the democrat really screws up. Hillary really screwed up. All she needed to do was hold the rust belt on top of the states that she won. She couldn't even do that. She lost virtually all states on the chart above, only winning in Minnesota and New Hampshire. And her victories there were fairly narrow there if I recall. Trump needs to overperform polling by over 6 points nationwide to have a shot at winning. While he certainly overperformed by such a margin in some states, him doing it consistently is going to be very unlikely.
Now, do I think there is wiggle room here? Sure. But no matter how you slice it, as long as Biden holds the rust belt, New Hampshire, and Minnesota, Trump is screwed. This should be easy, but let's not hold our breath. To Trump's advantage and Biden's disadvantage, Biden does not have a very enthusiastic base. His supporters seem more brought together out of sheer hatred of Trump and wanting to get him out than love for Biden. He could overpoll on the basis of people simply not being motivated to vote for him. It happened last time. Moreover Trump seems intent on choking the post office in order to suppress the vote to reduce turnout so he can win. So could Biden's numbers be lower than what is shown? Sure. But Biden is polling so strongly his win is almost a statistically significant conclusion. On the flip side, it could work the other way. Trump has an enthusiastic base who will fall on their sword for the guy, but this isn't the majority of the country. And if a the country is motivated by hatred of Trump given his gross incompetence and malfeasance, well, Trump will lose by a landslide. Biden might even win the south. Honestly, I doubt that Biden will win Texas and Georgia, as they are always a tease for centrist democrats trying to win over alienated republican voters, but he is within striking distance. This should scare the crap out of Trump. Biden could walk away with over 400 electoral votes if he sweeps the map. Honestly, I think the numbers could change from now by up to about 4 points. I highly doubt the variance will exceed that barring major game changing events as I believe most people have already made up their minds, but honestly, that means that Biden could score anywhere from 301 to 413 electoral votes in my opinion.
I know, I know, we saw this before in 2016. I was here talking about how Trump was screwed and couldn't win and at one point I did think Clinton could walk away with 400 electoral votes too. But I highly doubt Trump can repeat 2016. He's just too hated right now, and the country is in a position they would vote for a literal crap sandwich just to get him out so we can have some sane leadership. In 2016 the democrats tried forcing said crap sandwich on us and they voted for Trump out of spite. I don't see that happening this time. Nina Turner's half a bowl of crap is starting to look too appealing for some right now.
Anyway I might do a few more of these before the election. I definitely plan to do one on election day or the day before. They won't be as frequent as 2016 but honestly I don't expect much to change between now and November anyway. But long story short unless something changes the game significantly, Trump is screwed.
No comments:
Post a Comment