Let's looks back on our predictions from very early on in the cycle

[ProgressiveModerate]

This was a map I made for the Presidential Election back in May 2019. Mind you I was 13 and was brand new to this stuff:



LOL is all I can say.

I also found a screenshot from a Senate map I made right when Bullock jumped into the race. I'm sure I have one from earlier but now about to go down the rabbit hole of finding it.

[Mr.Bakari-Sellers]

GA was always a Tossup, we didn't know how it went in a Runoff

[Orser67]

I think I did pretty good, minus the Tim Ryan part:

I think Trump's presidency is more likely than not to be unpopular, and I think Dems will be able to unify in opposition to Trump and take all the 2016 battleground states except Iowa.

Kamala Harris/Tim Ryan 53
Trump/Pence 45

[dw93]

This was my first prediction from 2017:

Senator Sherrod Brown (D-OH) / Senator Kamala Harris (D-CA): 347 EV, 52% PV

President Donald Trump (R-NY) / Vice President Mike Pence (R-IN): 191 EV, 47% PV

Iowa, Ohio, North Carolina, and Maine 2nd Congressional District are decided by 2% or less.

[Hammy]

This was my first prediction from 2017:

Senator Sherrod Brown (D-OH) / Senator Kamala Harris (D-CA): 347 EV, 52% PV

President Donald Trump (R-NY) / Vice President Mike Pence (R-IN): 191 EV, 47% PV

Iowa, Ohio, North Carolina, and Maine 2nd Congressional District are decided by 2% or less.

Not bad with the PV.

[dw93]

This was my first prediction from 2017:

Senator Sherrod Brown (D-OH) / Senator Kamala Harris (D-CA): 347 EV, 52% PV

President Donald Trump (R-NY) / Vice President Mike Pence (R-IN): 191 EV, 47% PV

Iowa, Ohio, North Carolina, and Maine 2nd Congressional District are decided by 2% or less.

Not bad with the PV.

Thanks. I was honestly shocked when I revisited this post just how close the PV was to the actual result, lol.

[Mr.Bakari-Sellers]

NC was the worst trending state for D's since polls overestimated Cooper chances and he won by 4 not 10 and hurt Cunningham chances

[GP270watch]

 I didn't make a map because I thought Trump would barely win an EC victory while getting crushed in the popular vote. I am so glad to be wrong.

[Mr.Bakari-Sellers]

Predictions were off due to fact Covid hit in March and the polls were conflicted on a 413 map and a 291 map.

TX, NC and FL were wrong

[Clarko95 📚💰📈]

I was a 278 freiwall believer, give or take Arizona, Iowa, and Florida. Never saw Georgia coming. I didn't expect Democrats to lose or gain many seats in the House, but did believe Republicans would hold onto the Senate.

So I was pleasantly surprised by Biden's win being bigger than I anticipated and Democrats winning the Senate, but greatly disappointed and unhappily surprised by the House.

[DabbingSanta]

First prediction I posted on Atlas in February 2018:

[Crumpets]

I was very wrong about the Dem primary and Biden's prospects specifically - I was skeptical he'd run at all. But my guesses for how the election itself would look were at least respectable:

Harris/Franken 294
Trump/Pence 244

Trump has an approval rating around 40% in 2020. Nonetheless, he still polls ahead of Harris for most of the campaign. Around September, however, the general malaise of the country catches up to him, and Harris runs a solid campaign with practically no noticeable scandals. Trump solidifies the GOP hold on Iowa and Ohio, but Harris manages to win back Wisconsin, Michigan and Pennsylvania for the win. North Carolina and Florida are both squeakers and aren't called for days after the election, with Florida eventually going to Trump and North Carolina to Harris. Despite Democratic hopes to continue expanding the map, Arizona and Georgia both stay around the same numbers as in 2016. Texas moves ever closer to voting Dem, but Trump still wins by about 4-5 points.

[seeking mystical annihilation]

The first version of my Atlas prediction, dated 2-19-2020, looks a little bearish in hindsight:

[Adjective-Statement]

October 2018:


President Donald Trump (R-NY) / Vice President Mike Pence (R-IN) ✓
Senator Elizabeth Warren (D-MA) / Mayor Pete Buttigieg (D-IN)

I always saw it coming down to Warren and Harris.

[The People's Liberation Army of Rancho Cucamonga]

My first prediction for the election back in December of 2019 looked like this. It ended up looking exactly like the map on election night, but over time that changed.



Back in June I revised my prediction based on COVID and decided that Biden had the edge. I only called two states wrong on this one:



In retrospect I should've known better on Florida.

[bagelman]

My original map from 2019 without tossups



with

[Ferguson97]

Generic Dem vs Trump prediction from early 2018

[redjohn]

My first prediction from 2019-12-04 (meant to color IL dark red):



My most accurate prediction, from 2020-10-18:



In general, many of my predictions overestimated Biden in FL, and towards the end in TX. I didn't go into election day thinking he'd win TX, but I thought it would be a very slim victory for Trump and it wasn't close. NC was close enough to where I don't feel too dumb for missing it on election eve, but it's still really interesting that yet again Democrats sensed opportunity in NC only for it to go for the GOP again.

[ProgressiveModerate]

These were the 270towin consensus's from heading into 2020:





The funniest thing is that Democrats lost 2 of the tossup Senate seats but won 2 likely R seats