This help screen is still under construction. Please excuse the debris and lack of organization!
This is a computer version of our board game, Election Day. It mostly follows the rules of the board game, with a few minor differences where we could take advantage of the arithmetic skills of the computer. Because it is based on a board game, you may notice that many features act like a board game. For example, when we need a random number, we generate a number from 1 to 6, like a die roll. If this had been designed from go as a computer game, we probably would have generated a different number range in each situation. Some things are handled like a deck of cards rather than the sort of randomization that would occur to a programmer naturally. Etc.
This game runs entirely within your browser -- Edge or Firefox or Chrome or whatever browser you're using. It was, in fact, something of an experiment on our part to see if we could make a game that runs in a browser. So there's nothing to install: just go to the page with the game and start playing. In case you're wondering, once your computer (or cell phone or whatever) downloads all the components, there is no further communication with our server. The game does not run on our server and show results in your browser like many complex web sites. It runs entirely in your browser. You could almost disconnect from the Internet once you've downloaded the page. The only catch is that there are sounds and graphics that it doesn't download until it needs them. (We should make it work even if these are not available, just don't show the picture or play the music.) Oh, and lest you wonder, there's a catch to downloading all the components to your computer and running straight from your computer: browsers have a strange security rule that prevents them from loading some images and music from your hard drive. There are ways around this, but it requires messing with browser settings.
This is a game of a US presidential election. Each player (including computer players) represents a candidate for president. The screen displays a map of the 50 states plus the District of Columbia. When the players/candidates "campaign" in a state, they get "voter support points", called "voters" for short. Voters are a measure of relative support between candidates. (They do not represent a specific number of voters. Just, 2 voter support points beats 1 voter support point, etc.)
On election day, the candidate with the most voters in a state generally wins the state. But an opposing candidate may be able to pull off an upset victory. As the polls close from the east coast to the west coast, results roll in and the candidates see who actually wins each state.
As in real life, each state has a different number of electoral votes. The candidate with the majority of electoral votes wins. If no one has a majority, the election goes to the House of Representatives, where the candidate with a majority of the states wins, with each state getting one vote (regardless of population).
The game allows for 4 parties: Democrat and Republican, plus four "third parties": Green, Libertarian, Independent Democrat, and Independent Conservative. The third parties have a much better chance to win in the game than in real life. A game where a Green player has a realistic chance of winning, i.e. pretty close to zero, would be very frustrating. Assume for purposes of the game that there is some major political upheaval that gives these third parties a serious chance. You can give them a reduced chance of winning by designating them as "Minor Parties" or "Spoilers". But even at that, to make the game interesting they do much better than these parties do in real life.
Each state has a "friendliness" to each party that represents how difficult or easy it is for that party to win support in that state. Support for Greens and Libertarians is exaggerated to make it comparable to Republicans and Democrats, while maintaining relative strength between states. That is, if the Greens have twice as much support in state A as in state B, we give them twice the friendliness, but exagerrated to the same scale as the Democrats and Republicans.
Each candidate has a set of ratings that determine his chance of success at various game tasks: Media, Negative Campaigning, Debate, Fundraising, and Plan Surprise. We call this his "personality".
Each candidate also has a "party strength". This determines how many "actions" he can perform each turn, and how many states he gets for Endorsements and Voter Attention. (More on those later.)
The electoral vote counts and poll closing times in the game are those that are in effect for the 2024 and 2028 elections. The friendliness of states and PACs, and the amount of money each PAC has, are based on statistics from the 2016 and 2020 elections. If we should be so lucky that anyone is playing this game in 2032, and if your humble lead developer is still alive, we'll have to update the numbers.
The District of Columbia is, of course, not a state, but it functions like a state for almost all game purposes. All rules that apply to a state also apply to DC, except that if the vote goes to the House, DC does not get a vote. So, at risk of offending purists, in these rules any reference to "states" includes DC (except for the House vote).
You may notice that in the game there is almost no reference to the issues being debated – the only hint is the names of the PACs and the debates. Of course someone who believes in democracy might hope that the race is not about manipulating the voters but about the issues, but this is reflected in the friendliness ratings. This was deliberate, to make this a game about campaign strategy and not about controversial issues. If you’re a Republican you can play a Democrat and vice versa without having to espouse positions you hate! When playing the game, you are likely to find that you are talking, not about gun control or race relations or trade with China, but about who is ahead in this or that state and whether now is a good time for a media run.
We include some explanations and background where we feel this is helpful to understanding the rules. This information is not strictly necessary to play the game, and never supersedes actual rules. It is given in red, like this paragraph.
On the first setup screen, you describe the players. There can be up to 4 players, including both human players and candidates controlled by the computer, which we call "robocandidates". For each, select a party, a "personality", and a "party strength". If the candidate will be controlled by a human, check the "human" box. If it will be controlled by the computer, leave this box unchecked. If the candidate will participate in debates, check the "debater" box. In general, minor parties do not participate in debates. But you must have at least two candidates to debate.
"Personality" determines what the candidate is good at and what he is bad at. For example, a "Celebrity" is very good at media runs, not so good at debates. A "Bulldog" is good at debates and negative campaigning, but poor at fundraising and planning surprises. Etc.
"Party strength" determines how many actions the candidate gets each turn. Human players should generally be "major party", which gives 4 actions. In a game with multiple human players, you may want to make any Robo players "Minor party" (3 actions) or "Spoiler" (2 actions) so the robots aren't likeky to defeat all the humans. (Though I just played a game last week where two humans were beaten by a Spoiler.) If you are playing solitaire, after you've played a few games you may find that you can consistently beet a robo major party -- human beings are smarter than the AI. So you can make your opponent a Super Party to make a tougher game. I designed this game and I can usually beat a major party, but I still lose to a super party more often than I win.
By default the game sets the candidates to a Democrat and a Republican. You can switch the order or replace or add other parties. If playing with less than 4 candidates, set 1 or more to "no candidate".
Note: Independent Democrats and Independent Conservatives are not included in the board version of the game. They are intended to represent a candidate who loses in the primaries and decides to run as an independent, or otherwise opposes his own party's nominee. Independent Democrats have the same state friendliness as Democrats and Independent Conservatives have the same state friendliness as Republicans. The PAC friendliness is a little different. (Lest you wonder, we called the one "Independent Conservative" instead of "Independent Republican" just so we could abbreviate Independent Democrat to I and Independent Conservative to C.)
You cannot have two candidates with the same party or the same personality. You can have any number of candidates have the same strength, and any number of humans or robos.
By default the game assigns a random Personality to each candidate. Players should agree before starting the game whether they want to use randomly assigned personalities or let each human player choose a personality. If playing solitaire you can, of course, do whatever you want. Personally, I usually accept the random personalities to avoid manipulating the results.
You can play solitaire or with multiple human players. If multiple humans are playing, you use a "hot seat" method: whoever's turn it is should sit at the computer. Sorry, no sophisticated networking system. Maybe we'll add that someday. We expect most games to be played against a "robocandidate", that is, a candidate controlled by the game's AI. In that case you have one human player and one or more robo players.
In general, robocandidates should go after human candidates. This prevents humans from taking advantage of the robo on the last turn or the last turn before a debate.
The program sets the start date of the campaign based on the number of candidates -- with more candidates it sets a later start date, so there are about the same number of players-turns per game. The game goes faster on the computer than as a board game, so you may want to set an earlier start date to make a longer game. (The earliest start in the board game is July 1. In the computer version we let you start as early as June 1.)
The game chooses semi-random dates for the "events": Donor Interest, Endorsements, Voter Attention, and 3 Debates. Election Day is always the first week of November. You should not normally change these dates. But if you want to mess with something -- if, say, you want to see how the game goes if the debates all come in September -- you can change them. If you change Election Day, Election Day ends the game, so any events schedule after that will never happen. (We put in the ability to change election day when we were developing the game, so we could set election day to turn 2 and test the election day code without going through a whole game. But once we put the feature in, no point taking it out.) See Behind the Scenes for details of how the game decides the schedule.
Candidates all start with no money, $0. The only actions you can do when you have no money are to try to raise money, so your first action on your first turn must be Solicit PAC or Fundraiser.
Note: Throughout this help page, we routinely say "the computer rolls a die". Of course the computer does not actually roll physical dice. This means that it generates a random number between 1 and 6.
Each turn, the game loops through all the candidates. The display is different depending whether this is human versus robocandidates.
A Major Party candidate can perform 4 Actions per turn. Minor Party candidates get 3, Spoilers get 2, and Super Party get 5. There is no reason to ever do fewer than the allowed number. If you really can’t think of anything else to do, do a Fundraise to have more money next turn. In practice you will regularly find yourself with far fewer actions than things you want to do.
For human players, the game displays a table with one row for each action you are allowed for the turn. For the first action, it displays a list of all possible actions, with radio buttons for you to pick one. Actions that you are not allowed to do right now -- we'll get back to why something might be barred later -- are disabled. Click on the action you want to perform. Then click the "Details" button. If the action requires additional information, the game will display your choices for you to pick one. For example, when you Campaign, you have to say which state. Such options are always multiple choice, either radio buttons and a "submit" button, or a set of command buttons. Select, and the game displays the results. If no additional information is required, we skip this step and the game immediately displays the results when you click "Details".
Beside each action it shows the cost. If you do not have enough money to perform an action, the game won't let you do it. Do something else, or raise some more money first. You raise money by doing Solicit PAC or Fundraise. If you have no money, these are the only actions you can do.
Once you complete an action, the display on that row "collapses" to the action chosen, and the next row now has the radio buttons
When you've completed all your actions for the turn, the Details button is disabled and the Next button is enabled. Click Next to go to the next player, or if you are the last player, to go to the next turn.
For robocandidates, there is a similar table with actions on the left and details on the right. But a robo's actions are determined by the cmoputer, so you are not given any opportunity to select an action. Instead, the computer selects the actions and displays the ones chosen. For each action, click "Details" to see the additional information and results. When the robo has finished all its actions, the Details button is disabled and Next is enabled, just like for a human. Click Next to go to the next player or the next turn.
The actions available, for both humans and robos, are listed below. There is a slight difference that for humans, "campaign" is broken into Small, Medium and High. For robos, it is broken into Any and High. But the end result is the same, the difference is just because of how the state is chosen.
Tour a state giving speeches, holding townhalls, and meeting voters
This is the main action in the game. Most of your time will likely be spent doing Campaign actions.
For humans, there are three entries for campaign: Small, for states with 10 electoral votes or less, costs $10; Medium, for states with 11 to 25 electoral votes, costs $20; and High, for states with more than 25 electoral votes, costs $50. When you pick one of these, the game displays a list of all the states of the selected size. Click on the one you want. For small states, you get two states for one action. The second state must be adjacent to the first. After you select the first state, the game shows a list of the adjacent states. (Note that the game displays the results from the first state before asking you to pick the second. You could change your mind based on the result.)
You can only campaign in each state once per turn. So states where you have already campaigned have their buttons disabled. Note that you can win voters in a state in the same turn in other ways, like Media or winning a debate. You just can't CAMPAIGN in the same state more than once per turn.
Your chance of succeeding at Campaign depends on the friendliness of the state to your party. Friendliness is given as a number from 1 to 6. The computer rolls a die. If the die roll is less than or equal to the friendliness, you get a "voter" in that state. If the die roll is greater than the friendliness, nothing happens.
Note that each state has a different friendliness for each party. Utah is very friendly to Republicans and not very friendly to Democrats. Hawaii is friendly to Democrats and not friendly to Republicans. But the minimum friendliness is 1, so there is always SOME chance of sucess.
In general, for purposes of selecting a second small state, "adjacent" means adjacent in the conventional geographical sense. However, there are a few special cases:
If there are no adjacent small states where you have not already campaigned, then you do not get a second state. If your first state was Alaska or Hawaii, there will never be an "available" adjacent state, so you will never get a second state.
Run a series of advertisements on television, billboards, the Internet, etc. Appear on TV shows or other media. Do something to get yourself in the news: If you’re an incumbent, you may propose an important bill or conduct an investigation. If you’re not an incumbent, you may lead a protest.
The computer rolls a die and takes the smaller of your Media rating and this die roll. It then picks this number of states at random. You get a voter in each of these states. Friendliness is irrelevant.
Media runs win you more voters than an ordinary Campaign, and you can win voters in unfriendly states, but (a) a Media run is expensive, and (b) You don't get to pick the states. You may get voters in states where you are already well in the lead, or so far behind that it is pointless.
Example 1: Your Media rating is 4. The die roll is a 3. You get 3 states.
Example 2: Your Media rating is 4. The die roll is 5. You get 4 states.
Attack an opponent. You may criticize his policy positions or attack his character and try to embarrass him. We don’t consider whether your attack is truth or lies, only whether voters believe it.
Select an opponent. In a two player game, you only have one opponent so this step is skipped.
Either way, the computer rolls a die and takes the smaller of your Negative rating and this die roll. It then picks this number of states. For each of these states where your targetted opponent has one or more voters, he loses one.
If your target has no voters in a given state, nothing happens. There is no penalty for not having a voter to lose.
Opponents who were not the target are not affected.
Example: You have two opponents, A and B. You decide to target A. Your Negative rating is 3. The die roll is 2. The ame picks 2 states, Wyoming and Delaware. In Wyoming A has 2 voters and B has 1. A loses 1 of the 2 voters. In Delaware, A has no voters and B has 3. A cannot lose a voter because he has none to lose. B cannot lose a voter because he was not the target. So nothing happens in Delware.
Note that Negative Campaigning can be devastating, wiping out your opponent's support in many states. Or it could be totally ineffective, if your opponent had no support in any of the states the computer picked. Negative Campaigning is most likely to be effective against an opponent who has voters in many states.
Send out letters or emails, make phone calls, organize fundraising dinners, etc., seeking contributions to your campaign.
The game determines how much money you raise, if any. The amount you raise is affected by your Fundraise rating, and whether the Donor Interest event has happened yet.
See Behind the Scenes for details.
Attempt to get support from a PAC. PACs, or Political Action Committees, are organizations that attempt to influence the outcome of elections.
There are thousands of PACs, most with relatively small amounts of money. (From thousands of dollars to a few million. Big money to me and probably to you, but small compared to the total spent on a presidential campaign these days. When I was treasurer of a PAC, the president of the PAC once asked me if I thought we could “raise seven figures this year”. I said, “Sure … if there’s a decimal point in front of two of them.” Lest you wonder, we did actually raise over a million.) In the game, we group the PACs into categories, like “Business” and “Civil Rights” and treat them as a single entity. The names of the categories are terms that PACs use to describe themselves. (Need we say it? The fact that an organization with a certain name opposes a candidate does not necessarily mean that he is against the ideal expressed by the name. The fact that a Patriotic PAC supports a Republican doesn’t mean that Democrats are un-patriotic. The fact that a Women’s PAC supports a Democrat doesn’t mean that Republicans hate women. They may have different ideas about what these ideals mean. They may agree on the principle but disagree about methods. Etc.)
Most PAC activity is “independent expenditures”. Instead of giving money directly to a candidate, the PAC runs its own ads and holds its own events supporting and/or opposing candidates. In the game, it would be impractical to have hundreds of additional players whose only role is to control independent actions of one PAC, that is, to decide how to spend $100 one time. So instead, we give the PAC money to the candidate to spend on his own behalf. That is, SOME of the money a candidate receives from a PAC represents actual contributions, and some represents independent spending by the PAC to support that candidate. For simplicity we just lump it all together.
Each PAC has a friendliness to each party and a resource value. Resource is a measure of how much money they have. Note it is not the total amount of money they have, but just a number you plug in to a formula. You can see each PACs friendly and resource numbers by clicking the PACs button.
See Behind the Scenes for details.
You can only solicit each PAC once per game. Whether they give you anything or not, you can only ask once.
PACs generally give more money than Fundraising but may give nothing so it’s more of a gamble. And you can only solicit each PAC once but you can Fundraise any number of times.
In real life, PACs rarely give money to a candidate all at once like in the game. Contributions are spread out over time. Think of it as “pledged” rather than cash, and if you spend it faster than they give it to you, you are borrowing against these pledges. (We experimented with spreading out the PAC money and it increased complexity while adding little value.)
Surprises represent anything that might cause the vote on election day to differ from conventional wisdom and poll results. This includes flawed polls, differences in voter turnout, changing election laws to benefit your party in perhaps subtle ways, maybe manipulation and even fraud.
When you Plan a Surprise, the computer rolls two dice. For each die less than or equal to your Surprise rating, you get one Surprise. The game considers the two rolls separately: You might succeed at neither, one, or both.
You can get a maximum of 16 Surprises. After that, you are not allowed to do any more Plan Surprise actions. (If you have 15 Surprises, you can attempt another Plan Surprise but the most you can get is one.)
Example: Your Surprise rating is 4. The computer rolls a 4 and a 5. The 4 is a success and the 5 is a failure, so you get one Surprise.
Surprises affect who wins each state on Election Day.
Prepare for a debate. Study up on the issues, plan your responses and rehearse. Practice debating with aides standing in for your opponent. Attempt to sway public opinion in advance of the debate to your side.
The computer rolls two dice. For each die less than or equal to your Debate rating, you get 1 Debate point. The two rolls are considered separately: You might succeed at neither, one, or both.
Debate points are used during Debates.
You can't perform an action if you don't have enough cash on hand to do it. In that case, either do something cheaper or raise more money first with Solicit PAC or Fundraise. Note that you begin the game with no money, so the only legal actions for your first action on your first turn are Solicit PAC and Fundraise.
If you already have the maximum number of Surprises -- 16 -- you can't do Plan Surprise. You never lose or use up Surprises, so once you hit this total, you cannot do Plan Surprise any more for the rest of the game.
If you have the maximum number of Debate Prep points -- 8 -- you can't do Debate Prep. You spend debate points on each debate, so you could be at the limit this turn and then back under the limit next turn.
If you have already campaigned in all four big states this turn, you can't do Campaign/high. (This is only possible for a Super Party. You'd have to use your first four actions all as Campaign/high and then try to do a fifth.)
There is a set of button near the top of the screen that open various informational displays.
You can open and close the following displays freely:
There are also two displays which are opened automatically by the game when called for:
All the displays are shown on one long web page. A display you open is added to the page. You can open as many displays at the same time as you want. Scroll up and down between the displays. If the page gets clutterred, close some displays. Or, you can rip the display out to a separate browser tab. Click the Tab button to remove the display from the main page and put it on its own tab. You can then drag the tab into its own window if you want. If you have a large monitor or multiple monitors, you can arrange these windows however you like. You can also open a display in a tab by right-clicking the button rather than left-clicking.
At the top of each display is a Close button. (Note the button closes what's below it, not what's above it.) For events, this button is labeled "Continue" to remind you that you have to close it to continue the game.
While a display is open, it does not change, even if the data shown changes. To update the data, click the Update button or click the display button again. Exception: The map is updated automatically at the start of each turn, if it was open. It reflects any events that happened that turn. The election map is updated as states are resolved.
The Event display is opened automatically when there is an event. You cannot proceed with the game until you close it.
The Opinion Poll display opens automatically every time a new opinion poll is taken, i.e. the end of each turn. You can close it and re-open it if you wish.
The Election Day display is opened automatically at the end of the game. Once this display opens, you have to follow Election Day to completion, and then the game is then over, so there's no going back to other displays.
In many displays, the party names are abbreviated to one letter: D for Democrat, R for Republican, L for Libertarian, and G for Green.
A list of all the candidates, with each one's party, strength, personality, robo and debater flags, how much money he has on hand, and the number of Surprises and Debate Prep points he has accumulated.
The ratings for Personality are abbreviated to two letters to make the display a little more compact. Hover over the heading to see the full name.
Shows each state's abbreviation, name, number of electoral votes, its friendliness to each of the candidates, and the number of voters each candidate has. At the bottom it shows the count of the number of states in which each candidate has any voters, and his total number of voters.
A list of all the PACs. For each it shows the name, friendliness to each candidate, resources, and how much money it has given to each candidate, if any.T Thee is also a line for how much each candidate has obtained by Fundraising.
Friendliness determines how likely a PAC is to give to a candidate. Resources determines how much they can give. See Behind the Scenes for more details.
Shows the event schedule. It lists each event and the date when that event will happen. There are:
This displays each state with its two-letter postal abbreviation and number of electoral votes. (We don't display the name because to fit on the screen the text would have to be too small to read.) If you're not familiar with state postal abbreviations, see the States Display. Or check a list from the post office. Or search on the Internet. Or look at a US map and figure out which state is which by its shape and location.
Each state is colored to show who is leading in that state. The states are colored blue for Democrats, green for Greens (duh), yellow for Libertarians, and red for Republicans. If the candidate is ahead by just one voter, it displays a lighter hue. If he is ahead by two or more voters, it displays a darker hue. If a state is tied, it displays gray. If no one has any voters there it is light gray; if there are voters for two or more players (must be at least two to have a tie) it is dark gray. Note that in a 3 or 4 player game, a state is tied if any two candidates are tied. So for example if the Republican and the Democrat both have 3 voters and the Libertarians has 1, the state will show as tied, i.e. gray, even though for the Libertarian he is 2 voters behind. To see the exact number of voters for each candidate, look at the States display.
For the smaller states on the east coast, there is a set of boxes below the main map to help you see the colors.
This shows which states are friendly to which candidates. It shows each state in the color of the candidate that that state is friendliest to: blue for Democrats, red for Republicans, green for Greens, and yellow for Libertarians. If a state is equally friendly to two or more candidates, it shows gray. Special case: If it is equally friendly to Democrats and Republicans and no one else, it is purple. (We considered having different colors for every possible combination of parties, but there are 15 possible combinations (D, G, L, R, DG, DL, DR, GL, GR, LR, DGL, DGR, DLR, GLR, and DGLR) so it would be too hard for players to remember them all and the display would likely be confusing. If playing with more than 2 candidates, you'll have to consult the States display for details.)
If a candidate who is among the leaders currently has the most voters in the state, the color is shown in a light hue. If some other candidate has the most voters, it is shown in a dark hue. So states in your color in a dark hue are good candidates for campaigning.
Note that this map resembles the map from the board game.
US elections use the “electoral college”. Each state gets a number of votes that depend on its population. Each state’s law sets how they will decide who gets these votes. In the past some states had the state legislature decide. Today all states decide by popular vote. 48 states and DC give all their electoral votes to the candidate who got the most popular votes. If A gets 51% of the popular vote and B gets 49%, A gets 100% of the electoral votes and B gets none. Two states, Nebraska and Maine, may divide their electoral votes. In the game, we ignore this fact and Nebraska and Maine give all their votes to one candidate like the other states. Nebraska gets 5 votes and Maine gets 4, so in practice this might make a difference of 2 or 3 votes. This is so unlikely to change the outcome of an election that the complexity of having special rules for these two states wasn’t worth it. If you play a game where a candidate wins by 1 vote, and he just narrowly won Nebraska, it’s possible that a rule that provided for Nebraska splitting its vote would have made a difference. (No offense to the people of Nebraska and Maine intended.)
When we reach Election Day, we determine who wins each state and thus who wins the election and the game.
There are four parts to the Election Day display:
Under the state list is a "Next" button. Click this button to show the winner of the next state. Keep clicking until all states are shown.
The states are listed in order of poll closing time, which basically means from east to west. For states with the same poll closing time, they are in random order.
For each state, the game picks a Surprise. Whoever has the most votes, including any votes from the Surprise, wins the state. If it's a tie, the game picks more Surprises until a winner is determined. For more details of the rules, see Behind the Scenes.
To win the election a candidate must have a majority of the votes, not just the most votes. In a race with 3 or more candidates, they may split the votes so that no one has a majority. In a 2 candidate race, it is -- barely -- possible for the candidates to exactly split the vote, 269 to 269. In either case, according to the Constitution the election is decided by the House of Representatives.
In the House, each state gets one vote, regardless of the population of the state. If no candidate gets a majority of the states on the first ballot they have another ballot, for as many times as it takes for someone to win. The two times that an election has gone to the House, 1800 and 1824, there was a lot of negotiating behind the scenes. So much that opponents called John Quincy Adams’s victory in 1824 the result of a “Corrupt Bargain” with capital “C”.
All of this is difficult to represent in this game without having a whole separate set of rules to elect members of the House and more to recreate the deal-making. So instead, we resolve the House vote using the following rules.
According to the Constitution, the House can only consider the top 3 candidates from the electoral vote. So in a 4-candidate race, #4 is eliminated.
The game then resolves each state. This is similar to how we decide the electoral vote, but using a different process. We do not draw a Surprise for each state, but go strictly by voters on the map. For details about breaking ties, see Behind the Scenes.
In a 3 candidate race (or 4 candidates with one eliminated for the House vote), the candidates may split the states so that no one has a majority. In real life there would then be a lot of politicking to convince states to change their votes, and if that doesn't work, ultimately someone drops out and supports another candidate -- probably after demanding some promises in return. In the game, whoever has the fewest states drops out, and we then re-evaluate each state that supported him to see which of the remaining two candidates they now vote for.
If we are down to 2 candidates and they split the House vote exactly 25 to 25, so neither has a majority, then any states that were decided by a Surprise, plus 4 more states, are decided again. If it turns out that there is again a 25/25 tie, it does it again, until someone wins a majority.
A candidate cannot have fewer than zero voters in a state. If a candidate has no voters and anything calls for him to lose a voter, nothing happens. There is no penalty for not having a voter to lose.
If a candidate gains or loses voters in more than one state, the states have no effect on each other. For example, if he must lose one voter each in states A and B, but he has no voters in A, he still loses one voter in B.
You cannot have more than 8 Debate Prep points at one time. You cannot do Debate Prep if you already have 8. If you do Debate Prep and you already have 7, the most you can gain is one, even if you roll two successes.Note that if there is only one debate left, there is no point accumulating more than 4. If there are no debates left, there is no point accumulating any.
You cannot have more than 16 Surprises. Once you reach 16 you can no longer do Plan Surprise actions.
Some actions call for the game to pick one or more random states, such as Media and Negative. The game will not pick the same state twice in one player turn. Every state is equally likely.
In the rare event that all 51 states are picked for one reason or another, the game resets the list and starts over. (Rare as in, it never happened in our testing, but it's theoretically possible.)
Participating in a debate costs one action, so each candidate gets one fewer other actions that turn. All other events take no actions.
Events come before any player takes any actions in a turn. The game interrupts the normal turn sequence to resolve the event. Once the event is complete, resume a normal turn sequence. Events happen after opinion polls, so the last opinion poll does not show the results of the event.
Donor Interest affects Fundraise actions from the turn it occurs, including that turn, and for the rest of the game. All other events are resolved as they happen and have no “ongoing effect” (other than that voters placed as a result of the event continue in play as normal).
In real life, most of the events do not happen all at once for all candidates at the same time. They are spread over a period of time and that time period is different for different candidates. But for simplicity in the game, we consolidate this into 3 “big events” rather than 10 or 20 small ones. We experimented with many small events, but this was more trouble and made little difference.
The winner of a debate and the size of his victory are determined by comparing the number of debate points that each candidate has. Each candidate “spends” all the debate points he has accumulated, up to a maximum of 4, on the debate. Any left over are saved for the next debate. If you have debate points left over after the last debate, they are worthless.
Each candidate gets one free Debate Prep before each debate. This uses your normal Debate rating, but it is not an action and does not cost any money. Everyone will then have between zero and 6 debate points: zero to 4 accumulated plus zero to 2 from the free Debate Prep.
In the following rules, “win a voter” means the game picks a random state and you gain a voter in that state. “Lose a voter” means the game picks a random state and, if you have any voters in that state, you lose one. If you do not have a voter in that state to lose, nothing happens.
In real life, the winner of a debate gets a small increase in support in many states. In the game, the winner gets voters in a few states. You can think of this as, In a few states, the increase in support is enough to get rounded up to a whole voter, while in most states it rounds down.
Example: A candidate loses 2 voters. The game picks 2 states, Kentucky and Maryland. He has no voters in Kentucky and 3 voters in Maryland. So he loses 1 voter in Maryland, leaving 2. He has nothing to lose in Kentucky so nothing happens there.
If there are two candidates in the debate:
If the two candidates have the same number of debate points, the debate is a tie. Each candidate wins one voter.
Otherwise, whoever had the higher total wins the debate. Find the difference between the two numbers. The winner wins that number of voters and the loser loses that same number of voters.
Example 1: Candidate A has accumulated 5 debate points. He can only spend 4 on each debate so he will have 1 left over for the next debate. Candidate B has 3 debate points. On the free debate prep, A gets 1 more debate point while B gets 2. A’s final total is 5 and B’s is 5. The debate is a tie. They each win one voter.
Example 2: A has accumulated 4 debate points and B has 2. On the free debate prep roll, A succeeds on both rolls and gets 2 more debate points, for a total of 6. B fails on both dice and so gets zero. A wins the debate, 6 to 2. 6 - 2 = 4 so A wins 4 voters and B loses 4 voters.
If there are 3 candidates in the debate:
The game determines each candidate's debate points as for a 2-candidate debate. Whoever has the most is the winner, whoever has the least is the loser, and the one in the middle is second place. Resolve the debate between the winner and the loser as for a 2-candidate debate. Second place wins one voter. That is, the winner wins a number of voters equal to the difference between his debate point total and the loser’s, the loser loses that same number of voters, and second place wins 1 voter.
Example: In a 3-candidate debate, A’s final total (including the free roll) is 4, B’s is 5, and C’s is 2. B wins the debate, A is second, and C loses. B’s total minus C’s total is 5-2=3. Thus B gains 3 voters, A gains 1 voter, and C loses 3 voters.
If the two top performing debaters get a tie but the third candidate gets some lower number, calculate the difference between one of the top 2 (as they're the same it doesn't matter which) and the loser. Both the top 2 gain that number of voters and the loser loses that number of voters.
If there is a clear winner but a tie for loser, the winner wins the difference and the losers both lose difference.
If all candidates tie, everyone wins 1 voter.
If there are 4 candidates in a debate:
There is a winner, a loser, a second place, and a third place. Resolve similar to a 3-candidate debate: Winner wins the difference, loser loses the difference. Second place wins one voter. Third place loses one voter.
The Debate display lists all the candidates in the debate, whether each won, lost, or came in second or third, and a list of states where a voter was gained or lost. For lost voters, the game puts a check mark beside states where the candidate had a voter to lose and so really lost one, and an X beside states where the candidate had no voter to lose and thus lost nothing. (As gains are always successful, there is no check or X by these.)
When you are finished reviewing the display, click "Continue" to continue the game.
As election day approaches, donors become more interested in the campaign and become more generous.
From this turn and for the rest of the game, when rolling for Fundraising, we don’t subtract 2 from the die roll. That is, each Fundraise action gives a little more money. This has no effect on Solicit PAC.
The game displays a message saying that this event has happened. Click "Continue" to continue the game.
Notable politicians, other famous people, and media outlets endorse their preferred candidates. This can increase a candidate’s support in a state where he is doing well, or generate support in places where he would have difficulty winning it on his own.
When this event happens, before any players get any actions this turn, the game resolves endorsements for all players.
The game picks 8 states for each candidate. (Or fewer: if the candidate is from a Minor Party he gets 6, a Spoiler gets 4.) For each state, it rolls a die. If the die roll is less than or equal to the friendliness to that candidate, he gets a voter in that state. Essentially, he gets a free campaign roll in each state picked.
This costs nothing and is not an action.
Click "Continue" to continue the game.
Voters start to pay attention to the campaign. They read or watch the news and start to form opinions. Candidates may pick up some “free” support from voters seeking out information themselves.
When this event happens, before any players get any actions this turn, the game resolves voter attention.
The game picks 12 states for each candidate. (Or fewer: if the candidate is from a Minor Party he gets 9, a Spoiler gets 6.) For each state that has a friendliness to the candidate of 4 or more, he gets a voter in that state. If the friendliness is 3 or less, he gets nothing. The friendliness to other candidates does not matter.
This costs nothing, is not an action, and does not involve any die rolls.
Example: For the Republican, one of the states the game picks is Alabama. The friendliness of Alabama to Republicans is 5. The computer rolls a 4. 4<=5 so he gets a voter.
A robocandidate is a candidate who is controlled by the computer instead of by a human player. A robocandidate is subject to all the same rules as a human player except where otherwise noted here.
During setup, you make a candidate a robocandidate by not checking the "human" box. If he's not human, he's a robo.
To play solitaire, play against a robocandidate.
You can add a robocandidate to a 2 or 3 player game. For example, you could have two humans playing the Democrat and Republican and then include one of the third parties as a robocandidate. You may make such a candidate a “Minor Party”. Players must agree to include such a candidate.
The easiest way to assign a Personality to a robo is to accept the Personality assigned by the game. If the human players agree -- and if you are playing solitaire you just have to agree with yourself -- you can pick any Personality you want from the dropdown.
The game decides what actions the robo will attempt each turn. The first action is always "Campaign/first". After that the game picks actions from an "action pool". See Behind the Scenes for details.
In general, the robo resolves actions just like a human. ...
Just like for a human.
Just like for a human.
The robo only solicits PACs that have a friendliness to him of 4 or more. Just like a human, the robo can only solicit each PAC once per game. So "available PACs" are PACs the robo has not yet solicited that have a friendliness of at least 4.
The computer rolls a die. On 1 to 3 it picks the available PAC that is first on the list. On 4 to 6, it picks the friendliest available PAC.
From there, it is resolved just like for a human.
The robo decides which state to campaign in. See Behind the Scenes for details on how it picks. The small print summarizes how the robo picked the state. You can ignore this if you don't care.
Robos have two kinds of Campaign: Campaign/any and Campaign/high. Campaign/first works just like Campaign/any. As the name implies, Campaign/any may pick any state. Campaign/high always picks a high-vote state (California, Texas, Florida, or New York). The computer than pays the appropriate cost depending on the size of the state. (Campaign/any shows the cost as $50, but this is just the "worst case".) If the robo picks a small state, it gets a second state, just like a human.
The two types of Campaign are just to control the probabilies: we bias the robo slightly to choosing high-vote states. Early versions of the game didn't do this and the robo ended up not picking high-vote states enough.
The campaign is then resolved just as for a human player, i.e. the computer rolls a die and compares it to the state's friendliness to the robo's party.
If the robocandidate already has the maximum number of Surprises, i.e. 16, he picks an alternate action.
Otherwise, just like for a human player.
If the robocandidate already has the maximum number of debate points, i.e. 8, he picks an alternate action. If there is only one debate left in the campaign and the robo already has 4 debate points, he picks an alternate action. If there are no debates left in the campaign, the robo picks an alternate action.
Otherwise, just like for a human player.
The robo must select a target for the negative attack. If this is a 2 candidate race, there is only one possible target. If there are more than 2 candidates, the robo picks one. Generally, but not always, the one with voters in the most states. See Behind the Scenes for details.
Then the robo picks states just like for a human player.
In a few cases the robo ignores the action token picked and does something different. This decision is made when we get to the action, because previous actions might change what is legal. For example, a candidate might have plenty of money at the beginning of the turn but spend it all before he gets to the action in question.
When the robo decides it must choose an alternate action, the game crosses out the original action on the display and shows the new action below it. The old action is now irrelevant. (In early versions we just changed the action in the display but this was confusing sometimes: an action just mysteriously changed with no explanation, and the player was often left wondering, Wait, I thought it said ...)
Some actions cannot be played because they would be illegal. The same rules apply to robos as for human players. In these cases, the robo does an alternate action.
On the first turn of the game, for his first action the robocandidate does Solicit PAC instead of Campaign/first.
If the robocandidate gets to an action and he does not have sufficient money to do it, he does Fundraise instead.
The cost of Campaigning varies depending on the size of the state, which we don’t know until after we’ve gone through the priority routine, which is superfluous if we’re not going to do the Campaign. So if the next action is Campaign, the robo will Fundraise instead if he has less than $50. If he does the Campaign, he still pays the normal price; this rule just means that he assumes the “worst case” when deciding if he must Fundraise instead.
If the next action is Fundraise or Solicit PAC and the robo has more than $500 million on hand, he picks an alternate action.
There’s no point devoting time to raising money when you already have plenty.
On the last turn of the game, normally Oct 23, a robo ignores any Fundraise or PAC tokens and picks an alternate action.
This rule prevents the robocandidate from spending actions on the last turn raising money that he will never get a chance to spend. Note that if the robocandidate is broke, he may end up having to do a Fundraise. Thus this could result in replacing a Fundraise and then doing a Fundraise anyway.
A robo is affected by events just like a human.
When a robocandidate participates in a debate, he loses one action just like a human player. The action lost is the Campaign/first. He draws the remaining actions from the pool as usual.
Other events do not take an action for the robo, just like for a human player.
You can play with more than one robocandidate. In that case, each robo gets his own action pool, based on his own ratings.
At the end of each turn, the game conducts an opinion poll. The results of this poll are then displayed at the top of the page. It list all poll results since the beginning of the game, so you can see the trends. The poll shows the total of electoral votes in all states where a candidate is leading, and his share of the popular vote.
The electoral vote count only includes states where one candidate is in the lead. If a state is tied, it doesn't show up. So the total won't add up to the total for all states.
The popular vote always includes all states, so it should always add up to 100%. (The usual caveat: May not total 100% because of rounding.)
The electoral vote count has a column for each candidate. The popular vote total also has a column for each candidate, plus one additional column, labeled "?", for "other/undecided". The number of undecideds tends to shrink as the campaign goes on.
Calculation of electoral votes is straightforward: Identify all the states where the candidate has the most voters, and add up their electoral votes.
Calculation of popular vote is more complicated. See Behind the Scenes for details.
Below the chart are two graphs showing the trend. Note the scale on the Y-axis goes high enough to include most numbers that are likely to happen during the game, but if someone exceeds the top of the scale, the game just changes the scale to make it fit and redraws the trend line to the new scale.
If there is an event this turn, the poll is done BEFORE the event, so the poll results do not include the event, and the poll is shown above the event.
You don't have to know this information to play the game. But if you're wondering how some of the calculations work, here it is.
The game first schedules the three Debates, then Voter Attention, then Endorsements, then Donor Interest. The order is important because of the conflict rules (which we'll get to).
For each event, the computer rolls 2 dice and selects the date from the following table:
| Dice | Debate | Voter Attention | Endorsements | Donor Interest |
|---|---|---|---|---|
| 2-3 | Oct 23 | debate-2 | start+3 | start+3 |
| 4 | Sep 8 | debate-1 | start+4 | Aug 23 |
| 5 | Sep 15 | Aug 23 | start+5 | Sep 1 |
| 6 | Sep 23 | Sep 1 | Aug 23 | Sep 8 |
| 7 | Oct 1 | Sep 8 | Sep 1 | Sep 15 |
| 8 | Oct 8 | Sep 15 | Sep 8 | Sep 23 |
| 9 | Oct 1 | Sep 23 | Sep 15 | Oct 1 |
| 10 | Oct 15 | Oct 1 | Sep 23 | Oct 8 |
| 11-12 | Oct 15 | Oct 8 | debate-1 | Oct 15 |
The table comes straight from the board game. If we had designed this as a computer game from scratch, we probably would have had a formula rather than a table.
Two events cannot be scheduled on the same turn. If the date rolled for an event is the same as the date rolled for a previous event, we move the second event forward one turn. If that date is also already taken, we move forward another turn, etc. If this would result in an event being on or before the first turn, then we put it on the earliest open date.
The computer rolls a die. It adds this die to your Fundraise rating. If Donor Interest has not yet happened, it subtracts 2. It then multiplies this result by $20.
The difference between "before Donor Interest" and "After Donor Interest" is whether we subtract 2.
Example: Your Fundraise is 3. It is before Donor Interest. The computer rolls a 4. You receive (3+4-2)x20=5x20=$100. If it was after Donor Interest, you would receive (3+4)x20=7x20=$140.
The computer rolls a die and adds it to the PAC’s friendliness to you. If the result is 6 or less, you get nothing, Otherwise, it subtracts 6 and multiplies the result by the PAC’s resource value. That is, the formula is (Friendliness+D-6) x Resource.
Example 1: The PAC’s friendliness to you is 5 and their resource is $50. You roll a 3. You collect (5+3-6) x $50 = 2 x $50 = $100.
Example 2: The PAC’s friendliness to you is 4 and their resource is $50. You roll a 1. 4+1=5 which is <= 6. You get nothing.
If a PAC gives money to multiple candidates, these are all treated independently. That is, the PAC does not “run out of money” giving to candidate A and thus have nothing left to give to candidate B.
According to the Constitution, each state gets a number of electoral votes equal to the number of members it has in the House of Representatives plus the number in the Senate. Each state gets a number of representatives proportional to its population. Each state gets 2 senators regardless of population. So the number of electoral votes is proportional to the population PLUS TWO. Furthermore, every state gets at least one representative, no matter how small its population. The number of representatives must be rounded off to a whole number, but if it would round to less than 1 they still get 1. Because many states only get 1 or 2 representatives, the rule for how we round off makes a lot of difference. You might think this would be simple: just apply the usual rounding rule: If the fraction is 1/2 or more round up, otherwise round down. But there's a law that says that there must be exactly 435 members of the House, and rounding this way might result in a few more or less. So Congress passed a very complicated rounding rule. But we need not get into the details of that here.
This means that states with very small populations, like Wyoming, get more votes per voter than states with large populations, like California. Wyoming has about 300,000 voters and gets 3 electoral votes, or 1 electoral vote per 100,000 voters. California has 18,000,000 voters and 54 electoral votes, or one electoral vote per 333,000 voters. This was the result of a compromise when the Constitution was being written. The big states said states should get a number of votes proportional to their population, so if a state had twice as many people, it should get twice as many votes. The small states were afraid they'd be steamrolled by the big states, so they said every state should get the same number of votes. In the end they came up with this compromise where every state gets proportional to population plus 2, and every state gets at least 3. This means that small states have a little more influence on the outcome of the election than a strict proportional vote. It means that every now and then a candidate wins the election with a majority of the electoral vote, even though his opponent got more popular votes. You can debate the wisdom and fairness of this endlessly, but as a practical matter, without this compromise the Constitution would probably not have been approved.
The game determines the winner of each state as follows:
First, it counts the number of voters that each candidate has in the state. Then it picks one Surprise from the "Surprise deck".
The Surprise Deck is created during the course of the campaign. Each time a candidate succeeds at Plan Surprise, he adds zero, 1 or 2 Surprises to the Surprise Deck. A surprise counts as extra voters. The game assigns a number of voters to each Surprise: 1/2, 1, or 2. (About half of Surprises will be 1/2, one-third will be 1, and 1/6 will be 2.) There are also 25 "No Surprise"s in the deck. "No surprise" is zero votes. (In the board game, this is a literal deck of cards. In the computer game, we basically simulate a deck of cards.)
So to determine the winner of a state, the game randomly selects one Surprise. It adds the number value of the Surprise to that candidate's vote total. Whoever has the most votes, including any Surprise, wins the state.
Example: There are three candidates in the race. Candidate A has 2 voters, B has 1, and C has none. The surprise is B 2. The game adds 2 to B's total, giving final values of A 2, B 3, and C 0. B wins the state.
If the total, after adding in the surprise, is a tie, then the game picks another Surprise. If this is for one of the candidates who were tied, then that candidate wins the state. Otherwise, if it's "no surprise" or for some other candidate, the game picks another surprise, etc, until it finds a Surprise for one of the tying candidates. That candidate then wins the state. Note that the number on this second Surprise doesn't matter: it must have been a tie for us to be looking for another Surprise, so any number means that candidate now has the most votes.
Each surprise is only used once, until all the surprises, including all players and all the "no surprises", are used. Then we start over. In the board game, it's a deck of cards, and when they're all used, we shuffle the discard pile and start over.
A Surprise might break a tie. For example, if A and B have 2 voters each and the Surprise is A 1, then the final total is A 3 and B 2 and A wins.
A Surprise can create a tie. If A has 2 voters and B has 1, and the Surprise is B 1, then the total is now 2 to 2 and it’s a tie. (And so the game will have to pick additional Surprises to break the tie.)
A 2-point Surprise could result in a come-from-behind upset. If A has 2 voters and B has 1, and the Surprise is B 2, then the final total is A 2 and B 3 and B wins.
Note you can also get an upset victory if a Surprise creates a tie, and then a subsequent Surprise gives the victory to the candidate who had fewer voters on the map. Example: A has 2 voters and B has 1. The Surprise is B 1. So the total is now A 2 to B 2, a tie. We draw another Surprise and it’s B 1. B wins the state, even though A had more voters.
A Surprise may be futile. If A has 3 voters and B has 1 and the Surprise is B 1, then the final total is 3 to 2 and A still wins. Nice try, B, but no cigar.
Finally, a Surprise may be superfluous. If A has 3 voters and B has 1, and the Surprise is A 2, then A now wins 5 to 1 instead of 3 to 1. All that work and he would have won anyway.
A ½ vote can break a tie but cannot create a tie. If A and B have 2 voters each and the Surprise is A ½, then the final total is 2½ to 2 and A wins. But if A has 1 and B has 2 and the Surprise is A ½, then A has 1½ and B has 2 and B still wins.
The display shows what surprises were picked. For example, if a Republican 1-vote surprise was picked, it displays "R1". It displays "N" for "no surprise". (No surprise is always zero so there's no point showing a number.) If it picks multiple no surprises in a row, it shows an "x" and the number picked, like "Nx4".
Example: The vote is Republican 1, Democrat 2, Libertarian 1. The first surprise is R1. This gives totals of R 2, D 2, L 1, a tie between the Republican and the Democrat. The game must pick additional surprises to resolve the tie. The next pick is a Libertarian 2. The libertarians was not part of the tie so this does not affect the outcome. Then the game picks 3 "No surprise", before finally picking a "Deomcrat ½". This breaks the tie and the Democrat wins the state. For surprises the game displays "R1 L2 Nx3 D½".
If no one wins a majority of the Electoral Votes, the election goes to the House.
The game counts the number of voters in a state for each candidate as it did for the Electoral Vote. But there is no Surprise pick. If a candidate has the most votes, he wins the state.
If there's a tie, then whichever candidate has the highest friendliness (from those who tied for the most votes) wins the state.
This rule is intended to represent the fact that the members of the House would be likely to vote for the candidate from their own party, regardless of how the popular vote came out. We assume that the House members reflect the friendliness of the state.
If there's still a tie, the game picks a Surprise, and the owner of the Surprise wins the state. It only uses each Surprise once, until all Surprises are used, when it "reshuffles" and starts over.
Whoever gets a majority of votes in the House wins the election.
If no one wins a majority ... In real life there is wheeling and dealing and additional votes. Usually someone will drop out and throw his support to another candidate. So whoever got the fewest votes drops out. We then have another round of voting, where his states are "up for grabs" by the remaining candidates.
In real life, a candidate who was way behind would likely drop out and throw his support to another candidate -- after demanding some promises in return. But if the count was close, they'd probably try convincing some states to change their vote first. We considered a rule that said that if the race was close, we'd have another round where states decided by a Surprise card, maybe a few others, are reconsidered. But the number of such states is usually small and odds are they'd split among multiple candidates and just result in another no-win. So having extra rounds before someone dropped out would likely be unproductive and boring.
As there are 50 states, an even number, it is possible for the vote in the final round to be a dead tie, 25 to 25. In that case, we have another round of voting. For any states decided by a Surprise, we draw a new Surprise. In addition, the game chooses 4 states at random to also be decided by a Surprise. (The extra 4 is so just in case there were no states decided by a Surprise, there will always be some states that might change their vote and break the tie.)
To decide what actions a robocandidate does each turn, the game creates an "action pool". In the board game, the action pool is a set of markers that are put in a bowl. Each marker has the name of an action on it. The players then draw markers from this pool to decide what actions the robo attempts this turn. In the computer game, we simulate this pool with entries in a table, and the computer randomly selects entries.
The game builds the pool based on the robo's ratings. There are more tokens in the pool for actions the robo is good at, so that he is more likely to do these things and less likely to waste his time doing things he is unlikely to succeed at. But we still include SOME tokens for things he's bad at, so he does try now and then in the hope of getting lucky. And making the robo a little less predictable.
The pool always includes a Campaign/first, Campaign/Hi vote, 2 Campaign/any, a Solicit PAC, and a Fundraise.
If the robo's Surprise rating is 1, it gets one Plan Surprise token. If its rating is 2 it gets 2, and 3 or more and it gets 3. If the start date of the game is August 1 or later AND the robo has a Surprise rating of 3 or more, he gets one more Plan Surprise token. (By default, the start date is August 1 if there are 3 players and August 15 if therer are 4.) That is, in a shorter game the robo will have a bigger chance of planning a Surprise each turn, so that over the course of the game he still gets in a fair number.
Similarly, if the robo's Debate rating is 1, it gets one Debate Prep. If its rating is 2 it gets 2, and 3 or more it gets 3. If the start date is August 1 or later AND the robo has a Debate rating of 3 or more, he gets one more Debate Prep token.
If the robocandidate’s Media rating is 2 he gets 1 Media token, rating of 3 he gets 2 tokens, 4 he gets 3 tokens, and 5 or more and he gets 4 tokens. If the robo's Media rating is only 1 he gets no Media tokens, that is, it will not do any Media actions this game. Because with a rating of only 1, it's not worth the time and cost.
Similarly, if the robocandidate’s Negative rating is 2 he gets 1 Negative token, rating of 3 he gets 2 tokens, 4 he gets 3 tokens, and 5 or more and he gets 4 tokens.
Some tokens are included in the pool from the start of the game. Other tokens are added as the game progresses.
Negative tokens are added on turns 3, 5, 6, and 7.
This prevents the robo from wasting time on Negatives when no one has a lot of voters on the board yet so the chance of getting "hits" would be small.
The Fundraise token is added on the turn of the Donor Interest event.
The first Debate token is in the pool from start. If there's a second it's added 6 turns before the first debate. The third is added 4 turns before the first debate. If there's a fourth it's in the pool from start.
When a robo doesn't have as many tokens as the list above, the question of when they are added to the pool is simply moot. So for example, if a robo gets only one Negative token, it is added on turn 3.
There are two types of Campaign: Campaign/any and Campaign/Hi vote. Campaign/first is resolved just like Campaign/any.
For Campaign/any, the game picks 3 random states. The robo will then campaign in one of these 3 states. For Campaign/Hi vote, the available states are the 4 states that get the most electoral votes: California, Texas, Florida, and New York. In either case, the robo then picks from these states using a set of priorities.
First condition we call "triage". For each available state, the robo determines if he is leading in that state, tied for the lead, or someone else is leading and he is behind. He then gives first priority to states where he is tied, second to states where he is trailing, and last to states where he is leading. If no candidate has any voters in a state, this is zero to zero, and counts as a tie.
If two or more states tie on that criterion, the robo compares its friendliness to that of other candidates. First priority goes to states where it has the highest friendliness, second to where it's tied, and third to where it is trailing. (Note that the choice of top priority here is different than for triage.)
If still tied, it picks the state with the most electoral votes.
If still tied, it picks one at random (from among those tying on all previous criteria).
The small print summarizes this process. It displays the initial list of states. Then it says "popular" if it picked one by the triage criterion, "friendly" if that was a tie and it picked based on friendliness, "votes" if it was a tie after the first two and it picked based on electoral votes, and "random" if there was still a tie and it picked one at random.
Whether it wins a voter in the selected state is then determined according to the same rules as for humans. That is, roll a die and if it's less than or equal to the friendliness, he gets a voter.
If this is a small state, the robo then gets a second small state. It must be adjacent to the first state. The robo chooses from among the adjacent states according to the same priorities as above.
Example 1: The robocandidate is a Republican. This race is between a Republican and a Democrat – no Libertarian or Green. The robo draws Alabama, Arizona, and California. No one has any voters in any of these states. Thus they are all considered ties and therefore equal triage. Alabama has friendliness 4665, Arizona is 4654, and California is 6643. There are only Democrats and Republicans in this race so the only friendliness that matter are AL 45, AZ 44, and CA 63. AL is friendlier to Republicans than to Democrats so that is first priority. AZ is tied so that is #2. CA is friendlier to Democrats so that is #3. The robo campaigns in Alabama.
Example 2: Again, the robocandidate is a Republican and this race is between a Republican and a Democrat. The robo draws North Dakota, Colorado, and New Mexico. In North Dakota the robocandidate has 2 voters and his opponent has 1. In Colorado both have 1. In New Mexico no one has any voters. Colorado and New Mexico are both ties while the robo leads in North Dakota, so we narrow the choice to Colorado or New Mexico. Both are friendlier to Democrats than Republicans, so this criterion does not narrow the choice. Colorado has 10 electoral votes while New Mexico has 5, so the robo chooses Colorado.
If the first state was Alaska or Hawaii, then there are no adjacent small states. These two are the only states for which this is true. In that case the robo does not campaign in a second state.
Like a human player, the robo cannot campaign in the same state twice in one turn. If the initial pick of states includes one or more states where the robo has already campaigned, these are ignored. If it had initially campaigned in all 3 states, it picks another three. If doing a small state, for the second state it ignores any states where it has already campaigned. If it has already campaigned in all adjacent states, than it doesn't get a second state.
In a two player game, the robo only has one opponent and so any Negative attack must be against that opponent.
If there are 3 or more players, the robo must pick one. It picks a random state. Whichever opponent has the most voters in that state is then the target of the attack. Any voters for the robo itself are ignored. If opponents in the selected state are tied for the lead, including zero-all, the robo picks another state.
If there is no state on the map where any opponent has a lead, the robo picks an opponent at random. This is very unlikely and in fact we've NEVER seen it happen in testing, but it could happen if, say, you did a Negative very early in the game and most states were zero-all.
Example: The robo has a Negative rating of 3. It is a 3 player game. He picks a state to decide who to attack. He draws Wyoming. He himself has 3 voters there, opponent A has 2, and opponent B has 1. Ignoring his own votes, A has the most votes, so he attacks A. He now draws 3 more cards. Wyoming is NOT one of the states attacked.
For each state we know the total number of registered voters (as of 2020). On the state screen this is labeled "People", to avoid confusing it with "Voters", which always means the voter points you accumulate during the game.
To calculate the popular vote, the game examines the number of voters each candidate has in each state. It takes the number of voters plus 1 for each candidate, and factors in 1/2 voter for undecided and "other". It then computes each candidate's percentage of the total, and multiplies this by the number of People to get the candidate's popular vote in that state. It adds these together for all states, and then computes each candidate's percentage of the total.
Example 1: There are two candidates in the race, A and B. Neither has any voters in a state. The state has 2,000,000 registered voters. The game adds 1 to each of A's and B's totals and adds in 1/2 for undecided. That makes 1 to 1 to ½. A then has 1 ÷ 2 ½ = 40%. B is the same. Undecided is ½ ÷ 2 ½ = 20%. So A has 40% of 2,000,000 = 800,000, B also has 800,000, and undecided is 400,000.
Example 2: Again, there are two candidates in the race, A and B, and the state has 2,000,000 registered voters. A has 3 voters and B has 1. Add one to each and factor in ½ for undecided and you have 4 to 2 to ½. Then A's share is 4 ÷ 6½ = 62% (rounding to nearest integer), B is 2 ÷ 6½ = 31%, and undecided is ½ ÷ 6½ = 8%. (Note this totals 101% because of rounding.)
There is no maximum number of voters per state. For the board game, we limited each candidate to 4 so we didn't have to deal with voter markers that could potentially be unlimited. On the computer, it's easier to just let the numbers keep going up.
In the board game, when campaigning in a small state you have the option to campaign in another adjacent state, or to pay more to campaign in a non-adjacent state. In the computer version, you must campaign in an adjacent state.
In the computer version, there is automatically a poll at the end of every turn. There is no bonus for winning the poll. Popular vote is calculated in a much more complex way, because the computer can do arithmetic much faster than a human.
In the board game, the final tie-breaker for a robo in picking a state to campaign in is "which card was drawn first". As the computer is not drawing cards, it picks a state from among those tying at random.
In the board game, when the election goes to the House, the final tie-breaker is to roll dice. In the computerversion, it picks a Surprise from the Surprise deck.
If the House vote is a tie, we have another round. In the board game, states up for grabs in the next round are any states that were decided by a die roll, plus two random states from each candidate. In the computer game, states up for grabs are any states that were decided by a Surprise pick, plus four random states chosen from all available states.