Welcome to the second dev log! Thank you for the positive reception from last week’s announcement! Today is all about votes, specifically, how the votes in the game are calculated and the systems that sway votes as your campaign goes along.
Before we get into the mechanics of how votes change throughout a campaign, we first need to understand where the data is coming from and how it’s loaded into the game.
The Base Electorate
The base electorate is what makes up the initial vote tallies when the scenario you’ve selected begins. Ultimately, it’s one huge US Census database with some modifications to reflect recent voting trends. For example, the 2024 election works off of the latest Census data in combination with the final results of each county. It’s the same for other elections: the 2000 election works off 2000, 1996 works off 1996, and so on. The goal is to provide the most accurate experience possible, which is why we load in more than just party registration and population.
In The Electorate, you’ll be able to see the population, registered voters, race, ethnicity, party registration, and the ideologies of each county. We’ve also determined whether a county is urban, suburban, or rural based on population size. For example, you’re more likely to see an urban county lopsidedly vote for a Democratic candidate, while the opposite is true for rural counties. We’ve also done the hard work of identifying the ideologies of counties. Los Angeles County, for example, gets the Progressive ideology, which gives bonuses to progressive candidates when campaigning there. In rural Pennsylvania, however, you’ll see Populist ideologies that vote heavily for candidates like Donald Trump, but perhaps less so for a more establishment-type Republican like Mitt Romney.
When pulled together, you get a robust set of metrics to plan your strategy around.
With all of this data, we can initialize the game by using an algorithm to set up the initial voting totals for each candidate. It starts by extracting demographic data from the database and saving them as variables for each county. Then, we multiply the registered voter totals by the party lean of each county, which is determined by a combination of party registration and recent voting trends. Once that’s finished, the votes from each county will be rolled up to create the proper vote totals for each state. Then, we’ll roll that up to the national level. Finally, we essentially cut these vote totals in half to reflect turnout, which is a core concept we’ll discuss in a moment.
Every turn, the game will recalculate these totals based on the actions taken by both you and your opponents from the previous turn, giving you a living, breathing map with millions of votes to fight for.
But how do votes change, you ask?
Turnout
At its core, an election is more about getting people to vote than it is convincing them that you’re the right candidate for the job. Party identification in the United States, and the world in most places, is as polarized as ever, and the number of people willing to switch sides has inched closer and closer to zero. If you look at the exit polls from the presidential election in 2024, only 5% of Republicans voted for Kamala Harris while 4% of Democrats voted for Donald Trump. Furthermore, 49% of Independents voted Harris while 46% voted Trump. So how did Trump win? By turning out his party. 35% of the electorate identified as Republicans, while just 31% identified as Democrats and 34% as Independents.
Because of this, the game is built around the concept of turnout. Each party has a total number of potential voters in each county represented by their registered voter totals. Let’s say you have a county with 100,000 registered republicans, 100,000 registered democrats, and 100,000 registered independents. Instead of simply saying each candidate has 33%, we instead calculate the turnout for each party. Base turnout is about half of the total number of voters for that county at the start of the game, although this may change for balancing reasons as development continues.
In our scenario, you’d now have 50,000 republicans turning out, 50,000 democrats turning out, and 50,000 independents turning out. But that’s still 33% for each candidate. That’s boring. Well, that’s why we layer on a number of other things. For example, there are 10 key issues in the game with varying levels of importance. We’ll get into more detail about issues later in the dev log, but just note that an initial calculation will divvy out the proper turnout numbers for each candidate given their position on an issue relative to the county’s position. With this, you may end up with 56,000 republicans, 49,000 democrats, and 50,000 independents turning out.
But then what about independents? How do they get divided? Independents split based on a candidate’s issue position relative to the center and the ideology. Independents, in general, are considered centrists. Of course, some independents lean left or right, but in most competitive elections, they are neatly split anywhere from 45%-55% for either candidate.
For the sake of keeping things clean, we take a candidate’s issue position and measure the difference from independents, which is set in the middle (5 is the center score). That is then multiplied by a value and scaled to not go over the total number of independent voters in a county.
The logic here is that candidates who are closer to the extremes of their party are less likely to be supported by independents, which is typically true in most elections. For example, in the 2020 election, Joe Biden was considered more moderate than Donald Trump, which led to him winning independents 54%-41%. In cases where both candidates are seen as liberal or conservative, they cancel each other out, which is what you see in our 2024 scenario. Another factor that determines independent splits is the county’s ideology compared to the candidate’s, which may boost or decrease turnout.
Once all of these initial turnout calculations are done, it’s up to you and your campaign to have the higher numbers come Election Day.
How Votes Change Indirectly
Generally, the main way you’ll increase your turnout or decrease your opponent’s is through campaign activities, but that’s a dev log for another day.
Since the nature of this dev log is about the voting system’s underlying mechanics, I’ll spend the rest of it explaining the systems that change turnout indirectly.
Favorability
It’s been a while since Americans had a positive view of both candidates running for office. Take a look at this Gallup article to see for yourself. The takeaway here is that generally speaking, candidates with higher favorability ratings tend to win the election. The most glaring exception here is 2016, where Trump had a -11% rating compared to Clinton, but in all other elections, it was either a victory for the higher percentage or the favorabilities were so close that it pretty much didn’t matter.
With that in mind, favorability is accounted for in The Electorate and does have an impact on overall vote totals. Favorability is calculated as the sum of the candidate’s attribute scores and divided by the favorability divisor (editable, but defaults at 52), giving you a baseline favorability. For example, Donald Trump’s baseline favorability in the current build of the game is 47.5%.
Favorability doesn’t tend to swing much, which is why this system works well. However, favorability will still change throughout the game. Every candidate has a Media Spin attribute applied to them, which is considered inside this equation. Media Spin is the numerical representation of a candidate’s reputation with the media and is determined by the previous turn’s actions and events.
For example, let’s say you start the game with a 2/10 media spin. Your favorability might be around 44%, which ultimately means that the media is depicting you in a bad light. To help your image, you can schedule a town hall event with a news station and put your campaign on the line in order to boost your ratings. If all goes well, you’ll receive a boost to your media spin, which will be reflected in the next turn’s favorability rating. A +2 boost would put you at 4/10, which might get your favorability up to 45.5%.
That matters a lot, actually, because favorability has thresholds. Favorability ratings above 55% will give you a considerable boost to your campaign’s effectiveness. Rallies, speeches, ads, and other actions in the game all get an extra boost when you have high favorability. At 50-54.9%, you get a minor boost. At 45-49.9%, you get no boost. And at 44.9% and below, you receive negative effects.
Favorability ratings are not strong enough to win or lose on alone, but they certainly matter in a close election. This also makes candidates with great attributes and media spin particularly strong, but also a great target for smear campaigns.
Enthusiasm
Enthusiasm represents how excited voters are about your campaign. We get a base enthusiasm score by taking the sum of the candidate’s attributes and multiplying by their current favorability rating. This means that your enthusiasm score is more of an arbitrary number, as you can easily end up with numbers like 36.1, 23.7, etc. This is your base enthusiasm, but each state will change as you campaign in them. For example, if your base score is 30 and you rally in Pennsylvania five times, your score for the state could reach 35. That doesn’t mean you can spam states with rallies, as there’s a diminishing return, but that’s the basic concept.
What’s important is what enthusiasm actually impacts, which is Election Day turnout. If you followed the 2020 US presidential election, you’ll recall the emphasis Donald Trump placed on showing up to vote on Election Day as opposed to voting early.
While the circumstances in 2020 were particularly unique, Election Day turnout for Republicans was extraordinarily high. For the game, we try to accurately model what that looks like, and the simplest answer is through enthusiasm. All in all, the people who are going to turnout and wait in long lines on Election Day are the ones most enthusiastic about your campaign, so your enthusiasm score converts into a modifier that impacts the number of voters you receive on Election Day itself.
But you might be wondering what that means for the votes you see prior to Election Day. While the game gives you an estimated number of votes as polls, the reality is that these are just polls. We’ll cover the polling system in a later dev log, but just note that there are 10+ pollsters in the game who all have different “grades.” Those grades determine their overall accuracy on any given week. The polls are then averaged, and that becomes your headline polling average for the week.
Depending on your enthusiasm score, your Election Day might look very different from what the polls say, for better or worse. The good news is that you’ll have tools available to you in the game to get an estimated number of voters accounted for through voter rolls, text message campaigns, and other check-ins during the early voting period.
Momentum
Momentum is meant to capture the national mood of your campaign. One of the key challenges in developing this game has been finding a good way to impact votes in counties you haven’t campaigned in. Since there’s over 3,100 counties, you cannot feasibly touch even 10% of them, and that wouldn’t make sense anyway. Momentum works to solve that by acting as an underlying modifier that impacts every county, every turn.
Momentum, like enthusiasm, is an arbitrary number that changes based on what happened last turn. It is calculated by finding the sum of your media spin, your base enthusiasm, and favorability, and then subtracting 85 and multiplying the result by 0.5, giving you a momentum score.
Let’s say your media spin in 6/10, base enthusiasm is 32, and favorability is 49.3%. The calculation would work as follows:
6 + 32 + 49.3 = 87.3
87.3 - 85 = 2.3
2.3 * .5 = 1.15
Your score in this case is 1.15. If we divide it by 100 to get a decimal value of 0.0115 and multiply your current turnout in a county, you’ll get a raw voter count that we’ll add back to the total number of voters. For example, if you have 100,000 voters turning out in Seminole County, Florida, the equation would look like this:
100,000 * 0.0115 = 1,150
1,150 + 100,000 = 101,150
Your new voter count is 101,150. It might seem like a lot when you do this for every county, but when you consider the fact that the US has over 170 million voters and Seminole County is a decently populated county, this is actually a modest gain. It should also be noted that momentum can be negative. Random events or narrative-driven content due to your campaign’s actions can result in an additional value that can add or reduce the initial equation. If a major scandal hit your campaign, you could have -10 taken from your initial momentum calculation, which could result in a score of 77.3. Here’s what those numbers look like when that happens.
77.3 - 85 = -7.7
-7.7 * .5 = -3.85
100,000 * -0.0385 = -3,850
100,000 - 3,850 = 96,150
A scandal can be so bad that your polls could drop significantly overnight. An October Surprise in The Electorate can truly destroy a campaign. Luckily, events as harsh as a -10 modifier are rare, but they can happen at inopportune moments.
County Tags
As I mentioned previously, county tags like urban, suburban, and rural are in the game. However, there are actually dozens of different tags that are used to represent the demographics and special features of a county.\[/p]
For example, counties with particularly large colleges will get a boost to democratic turnout. But tags don’t just impact turnout. As you can see in the image above, Los Angeles has the Hollywood tag, which gives a boost to democratic fundraising in the county.
Urban tags help boost democratic turnout, suburban tags increase the issue importance scales so that candidates with issue positions closer to the center benefit from more independents turning to them. Rural tags help drive republican turnout. This does not mean that counties in any of these areas are unwinnable by the opposing candidate. Because of the Census data and recent voting trends, Donald Trump can easily win Miami-Dade County, despite being urban.
Additionally, ideologies are considered county tags, and change the way votes move, or don’t move. These systems can get pretty lengthy to discuss, so we’ll leave the conversation about tags and modifiers for a later dev log.
Issues
We’re not going to cover the full issues system here, as that also deserves a dev log of its own, but I’ll give you the general concept here as they do impact vote totals at the start of the game and throughout.
The political spectrum in The Electorate is represented on a scale of 1-9. 1 being very liberal, 9 being very conservative, and 5 being the center. There are 10 issues in each scenario, all ranked by an importance modifier from 0.1 to 1. For example, the top issue in the 2024 scenario is the economy. The economy’s importance modifier is 0.9, while housing, the lowest ranked, is at 0.3.
Without getting into the weeds of the equation, which will be presented at a later date, the general idea is that the absolute delta between the candidate’s position and the county’s position on each issue is subtracted by 10 and multiplied by a modifier to create the raw voter count for each county to add at the start of the game. This means that Donald Trump will start with a completely different base electorate than Nikki Haley would, and so on.
And while I know there might be questions about why there’s only 10 issues, I’ll explain my reasoning with the context of this Pew Research article. Pew asked respondents about which of the 10 key issues identified in the 2024 election were most important to them. By the time you get to the bottom of the list, only 37% of respondents say that the issue is particularly important to them.
In terms of this game, going further than 10 issues would be unnecessary and not have any notable impact on vote totals, given the already present reduction to the importance modifier after issues 7 and 8. Therefore, I decided to stick with 10 issues, and create a stronger gameplay loop around those instead of having multiple secondary issues that don’t have any significant impact on the final outcome.
What Comes Next
That concludes our overview of the voting system. There are a number of other factors that impact voting, but the rest of them are driven directly by campaign actions. We’ll touch on those in later logs.
Once again, thank you for the kind support and engagement from last week’s launch. It’s been very rewarding to already see a community come together around The Electorate. If you haven’t already, feel free to join the Discord server to discuss the game, give suggestions, and help make the ultimate election simulator!
If you’re excited about the game, be sure to add it to your wishlist and hit the follow button.
Be sure to come back next Tuesday when we go over the Candidates and their systems.