My 2020 Model / Election Simulator

[electoralfox]

Hey folks!

I'm pleased to announce that I have created a tool that lets you change how each demographic group could hypothetically vote in the 2020 election. Furthermore, you can also edit how Trump's approval rating will change from now until election day. The simulator uses the following to churn out projected vote percentages and margins in each state. The simulator also includes a live map that will visualize the results!

The simulator weighs the following elements.
- Partisan Lean for each state
- Approval and Disapproval ratings for the incumbent president
- Weighted polling averages taking into account recency and pollster bias
- Multiple Regression method to account for polling gaps in certain states
- Demographic projections

Feel free to play around with it and post your maps along with your customized scenarios! I plan on updating this with new features including the projected third party vote and more!

Right now, the demographic percentages are set up at how each group voted in 2016. You can change this and you can also change how much Trump's approval rating will increase or decrease from present-day to election day.

https://docs.google.com/spreadsheets/d/1W8xJT4kRI5-47C-dLj9KjJgy86tSQiSYM4d74zW2XCY/edit#gid=0

Edit: If you want a personal copy, just go to File > Make A Copy. This way when you're playing around with the tool, your results won't be interrupted by others being on the spreadsheet at the same time!

[SaneDemocrat]

Decent Montana and Kansas

[electoralfox]

Decent Montana and Kansas

Thanks for checking it out! I do think the competitive Senate races in Montana and Kansas could help Biden increase his performance in these states. It's also interesting to note that Kansas becomes closer if you improve Biden's performance with the college-educated White Americans cohort.

[SaneDemocrat]

Is this assuming Biden+4.5 nationally?

[Skye]

Is there a way to reduce the third party vote? It's a tad too high IMO.

[electoralfox]

Is there a way to reduce the third party vote? It's a tad too high IMO.

I'm planning to add this feature very soon.

Right now, the third party vote is determined through the 2016 election results, the polling average, the partisan lean for each state, and the demographic inputs.

[SaneDemocrat]

Is there a way to reduce the third party vote? It's a tad too high IMO.

I'm planning to add this feature very soon.

Right now, the third party vote is determined through the 2016 election results, the polling average, the partisan lean for each state, and the demographic inputs.

Someone ruined your google sheet

[electoralfox]

Is there a way to reduce the third party vote? It's a tad too high IMO.

I'm planning to add this feature very soon.

Right now, the third party vote is determined through the 2016 election results, the polling average, the partisan lean for each state, and the demographic inputs.

Someone ruined your google sheet

Yes, I changed it. I'm only making the sheet viewable from now on. If you want to play around with it, you'd have to make your own copy from now on.

It's unfortunate some childish individuals still exist amongst us.

[G_Master]

I think third party voters are going to be one of the more interesting aspects of this election. I would like to think that the vote share would be at 2012 levels or below, but obviously it's hard to say. I honestly wonder how many people dislike BOTH Biden and Trump ENOUGH (as in at the same level) to vote third party; it probably isn't as high in 2016. Honestly, I just wish all parties except maybe the Libertarian party (the most popular third party), dropped out of the race just to make the election simpler and more streamlined and less uncertain.

[bagelman]

I set all demographics to 50/50 and Biden wins by 6 points...

[electoralfox]

I set all demographics to 50/50 and Biden wins by 6 points...

Yes, demographics are not the only thing it’s taking into account. It also takes into account approval ratings in each state, the partisan lean of each state, and “experts’’” rating as well.

Not to mention, Trump’s approval rating right now is around 40-55 on the disapproval side. I would suggest changing the approval rating to +8 or +10 if you’re trying to make it a 50-50 thing.

[ProgressiveModerate]

I've made a simulator as well, but it's certainly by no means as detailed as yours with the various demographic groups and such. Click the checkbox and a random election will be simulated based on the NPV. You may have to create a copy of the spreadsheet in order to do so. I still need to work on finding PVIs that aren't so crappy, as well as correlating certain states/regions. Here's a link if anyone wants to check it out: https://docs.google.com/spreadsheets/d/1S8gjPcyRSa50h_cRgDPRm00aE-UfIKGnwolbW3hqNYI/edit?usp=sharing

[electoralfox]

I've made a simulator as well, but it's certainly by no means as detailed as yours with the various demographic groups and such. Click the checkbox and a random election will be simulated based on the NPV. You may have to create a copy of the spreadsheet in order to do so. I still need to work on finding PVIs that aren't so crappy, as well as correlating certain states/regions. Here's a link if anyone wants to check it out: https://docs.google.com/spreadsheets/d/1S8gjPcyRSa50h_cRgDPRm00aE-UfIKGnwolbW3hqNYI/edit?usp=sharing

Looks pretty neat! My only question is what the map is taking in to display the results. Is it showing the margin of victory? Or is it showing the percent chance of winning each state?

I really like that you can simulate random elections!

My model does not take into account elasticity. That is something interesting that I want to look into.

[ProgressiveModerate]

I've made a simulator as well, but it's certainly by no means as detailed as yours with the various demographic groups and such. Click the checkbox and a random election will be simulated based on the NPV. You may have to create a copy of the spreadsheet in order to do so. I still need to work on finding PVIs that aren't so crappy, as well as correlating certain states/regions. Here's a link if anyone wants to check it out: https://docs.google.com/spreadsheets/d/1S8gjPcyRSa50h_cRgDPRm00aE-UfIKGnwolbW3hqNYI/edit?usp=sharing

The color represents the margin but as the margin gets closer to 0, there is more differentiation in the color


Looks pretty neat! My only question is what the map is taking in to display the results. Is it showing the margin of victory? Or is it showing the percent chance of winning each state?

I really like that you can simulate random elections!

My model does not take into account elasticity. That is something interesting that I want to look into.

Here's a link to the updated version BTW: https://docs.google.com/spreadsheets/d/1gsWBec1xhH5qxSDbfwoPI7bONfyaBA_DPlNaAJgDZCc/edit?usp=sharing

[𝕭𝖆𝖕𝖙𝖎𝖘𝖙𝖆 𝕸𝖎𝖓𝖔𝖑𝖆]

Sorry but I noticed you have weighted average of polls in West Virginia saying Trump 33 Biden 66
That's batsh**t crazy, did you accidentally flip the numbers?

[electoralfox]

Sorry but I noticed you have weighted average of polls in West Virginia saying Trump 33 Biden 66
That's batsh**t crazy, did you accidentally flip the numbers?

Thank you for pointing this out! Yes, it was a bug. I have fixed it.

[𝕭𝖆𝖕𝖙𝖎𝖘𝖙𝖆 𝕸𝖎𝖓𝖔𝖑𝖆]

Sorry but I noticed you have weighted average of polls in West Virginia saying Trump 33 Biden 66
That's batsh**t crazy, did you accidentally flip the numbers?

Thank you for pointing this out! Yes, it was a bug. I have fixed it.

You're welcome!

[Progressive Pessimist]

That actually looks like a potentially accurate way for this election to end up.

My only issue would be that I think NE-2 would flip well before Georgia or North Carolina.

[electoralfox]

That actually looks like a potentially accurate way for this election to end up.

My only issue would be that I think NE-2 would flip well before Georgia or North Carolina.

I think this is because of the lack of polls congressional district-specific polls in NE-02. I do agree with you.

[SingingAnalyst]

Excel-lent tool! (Get it?)

I will definitely play around with it. In particular, I believe the difference between white college and noncollege populations will widen slightly.