The voters
Hover anything for its numbers. Click a news source to see who listens to it.
1Pick a country
Tuned to real survey data (World Values Survey)
Made up, to show the idea
2Pick how to rig it
Combine as many as you like.
One good candidate. One bad one. 1,001 voters. Someone wants the bad one in, and will pay for votes or buy the news. Every election is counted twice: one person one vote, and educated votes worth 5×. Which one is harder to steal?
Hover anything for its numbers. Click a news source to see who listens to it.
Tuned to real survey data (World Values Survey)
Made up, to show the idea
Combine as many as you like.
Every election is counted two ways. Equal votes: one person, one vote, like now. Educated ×5: an educated person's vote counts as five. Same voters, same ballots, two results.
Voters pick by weighing the two candidates. The good one is only slightly better, so other things easily tip people: the news they follow (every source leans a random way each election, even when nobody pays it), their party, and plain misjudgement. A whole village or a whole channel's audience can be pushed the same way at once, and that is what makes a group wrong together.
Every election is also replayed with no rigging from the same random draws, so the page can tell you exactly what the rigging changed.
One election can be luck. Run a thousand with the same country and the same rigging, and count how many the rigging stole under each way of counting.
Each one is used in every election. Its source says whether it comes from a survey, a reported figure, or is our assumption. The country and rigging buttons above just set these.
Not in this model: turnout, many seats at once, fake degrees, and parties picking candidates to suit the rule. The research repo tests those separately.