If your personal prediction is terribly off, what happened?

[Admiral Stockdale]

1. Biden's purported support among >65s is lower than projected
2. Anti-government / Anti-establishment sentiment that drove 2016 voters continues/expands.

[morgankingsley]

It's why I'm not making a formal prediction until November 2.

I never see any issue with something like this. As I said back in like March, and I will spell it right this time:

The only poll that really matters is the poll on election day

[Agonized-Statism]

It's why I'm not making a formal prediction until November 2.

In 2020, yeah. This year is so crazy, who knows.

[Rules for me, but not for thee]

Democrats underestimated white Americans' apathy at BLM, it's not 2050 yet.

Antipathy* for the looting and Marxists at the heart of the movement.

[It’s so Joever]

I rightfully get mocked for years while President Biden likely wins a huge victory and gets a solid trifecta.
Also, I gotta love the overconfidence of Atlas right now. It’s just so fitting.

Okay doomer

Okay optimist

[John Dule]

I missed three states, but each was within 1%, so I don't feel like I got it that wrong. The blue Texas people are the ones who really screwed the pooch.

[Horus]

My prediction was only one state off, my own state.

It shows that I underestimated the Dem shift here, which looking back was silly but expected as I was viewing the entire election through a "Shy Trump vote" pessimistic lens. However, polls in Georgia, unlike in most other places, had been highly accurate in the past. I failed to remember that.

[Kuumo]

I only missed Georgia and North Carolina. I knew they would both be within 1 point and felt like Biden would flip one of them but guessed the wrong one. I suppose that my margin predictions of Biden +0.3 in North Carolina and Trump +0.6 in Georgia weren't really that different from the actual results.

[KaiserDave]

My Presidential prediction was exactly right. The the most accurate on the site if I do say so myself. I will admit that personally I had NC going to Biden the night before. But the site prediction is there!

My congressional predictions were pretty bad, I had a 50-50 Senate but I had competitive races in SC, MT, and KS that simply didn't exist in reality. The site congressional prediction I had is from February and not accurate of what I thought by Election Day.

[🐒Gods of Prosperity🔱🐲💸]

I was close - I had GA R and NC D.   I had actually just switched GA from D, it seemed like polling was getting better for GOP there and in NC, which I thought about switching also, but didn't.

[S019ian Liberal]

Got all right but Florida so far, of the uncalled state Arizona and Georgia are looking like good calls one but North Carolina not so much.

[𝕭𝖆𝖕𝖙𝖎𝖘𝖙𝖆 𝕸𝖎𝖓𝖔𝖑𝖆]

My presidential prediction only missed the Southeast, it seems, but I had all three states as tossups in the Confidence map, so I am quite contented.

[DaleCooper]

I was wrong about North Carolina and Georgia, though I expected all three of those southeastern states to be tossups.

As far as congress goes, I was wrong about the House I guess, but I was always pretty bearish on the Democrats' chances in the Senate. I expected 50/50, which can still happen, though I thought that that 50/50 would come through wins in ME and NC.

[Amenhotep Bakari-Sellers]

Let's be real these maps aren't real voting maps they are mock predicting and should be treated as such.  Overpredicting is common for D's

[pbrower2a]

Donald Trump and his campaign. Give those the credit that they deserve for making things as close as they did. Donald Trump still has his charm. Trump may be a disgusting person, but he can appeal to visceral concerns that Biden neglected -- perhaps because he can't meet those. I saw a range of results, none of which included any Trump victory, from a bare Biden victory (TL16* + MI+PA+WI) to 413 electoral votes for Biden. I thought that the states would fall in this sequence for Biden after  MI/PA/WI:

Florida
Arizona
NE-02
Iowa
ME-02
Ohio
North Carolina
Georgia
Texas

...oh did I get the sequence wrong!

*Trump Losses (20)16

[Figueira]

I didn't post my prediction here (posted it on a different forum) but I was banking on the polls underestimating Democrats in Texas (and I missed the RGV shift) and I let my optimism get ahead of me in ME-2. Other than that, I got everything right.

[S019ian Liberal]

The polls in Georgia, unlike in most other places, had been highly accurate in the past. I failed to remember that.

Not a good strongly point to make. As you could say the same for Texas to take just one example. If anything in fact the results historically have favored Dems by even more than the polls ultimately did eventually so. However this year all of those dynamics just took a dramatic turn for the worse, and so Trump way outperformed expectations there due to Dem VS collapses and lucklustre showings in terms of turnout in the suburbs.

[Hammy]

I failed to understand voting patterns at district level, and GA came down to margin of error--I said it would be +2 or less either way.

Florida and Texas (despite getting the winner right) were my biggest misses, underestimating how much Hispanics would swing towards Trump (or simply not show up at the polls, but same outcome)--I had Texas having a D+2 to R+4 range, and Florida being D+1 to R+1 range.

[Yeahsayyeah]

Polling error and some overconfidence that tempted to make bold predictions. The low early voting numbers at the Rio Grande valley should given me a pause predicting Biden the victor in Texas. I thought Kasich endorsement had some meaning in Ohio which apparently it had not. And I did not see the degree of South Florida Cubans swinging to Trump. And the "NC is more likely to flip then GA because of Kemp voter supression" talk, of course. But on the other hand I knew that everyone of these states was possible to go for Trump in the end. So TX, FL, NC, OH (+ME-2) to Biden was the Biden ceiling prediction. My trump ceiling prediction would have given him GA and the states he won. I probably should have chosen this, but well...

[bagelman]

My personal prediction was the 413 map because I deliberately prioritized mental health over getting the most accurate prediction. The only state I was truly and wrongfully confident about was North Carolina. I totally underestimated the sex scandal and believed people who claimed it would surely vote to the left of Georgia.

[MR DARK BRANDON]

I only got FL, NC and GA wrong

[S019ian Liberal]

haha. i told y'all FL wouldn't be within 3 to 5 points! get owned, atlas.

]I thought Kasich endorsement had some meaning in Ohio which apparently it had not.

we should already have learned this before when cleveland debate during the primary, then later on at the gop convention in ohio where home state the governor failed to refuse in showing. also his lack of endorsement to an official ec candidate apart from himself during the ge campaign too!