Glossary¶
- benchmark
A suite of multiple election simulations for a specific voter model, in which particular parameters are varied.
- voting system
An voting method such as plurality, instant-runoff-voting, scored voting, or otherwise.
- plurality ratio
The ratio of a plurality winner’s honest votes over the total votes cast in the election.
- plurality winner ratio
See plurality ratio.
- Condorcet winner
The candidate which conforms to the Condorcet Criterion
- majority winner
The candidate who recieves greater than 50% of 1st choice votes.
- plurality winner
The candidate who recieves the most honest 1st choice votes.
- utility winner
The candidate that maximizes the average utiltiy of the voter population.
- C scenario
An election scenario in which the Condorcet winner is not the utility winner.
- CPU scenario
An election scenario in which a Condorcet-plurality-utility winner is found; where the the Condorcet/plurality/utility winner are coincidental.
- CU scenario
An election scenario in which the utility winner is also the Condorcet winner.
- CP scenario
An election scenario in which the plurality winner is also the Condorcet winner. However, the Condorcet-plurality winner is not the utility winner.
- MU scenario
An election scenario in which a majority winner exists and is also the utility winner.
- M scenario
An election scenario in which a majority winner exists but is not the utility winner.
- nc scenario
An election scenario with a Condorcet cycle. In other words No Condorcet winner is found.
- Voter Satisfaction Efficiency (VSE)
Measure of voter average utility in which at 100% VSE, the candidate that maximizes utility is elected. At less than 0% VSE, a candidate with utility worse than the average candidate is selected. This metric was proposed by Jameson Quinn.
- Voter Regret (VR)
Measure of voter utility proposed by John Huang. At 0% voter regret, the candidate that maximizes voter utility is elected. When VR is greater than 100%, a candidate is elected whose regret exceeds the difference between electing a candidate with preferences of a random, average voter and an ideal candidate located in the median centroid of the voter population.
- voter tolerance
The maximum difference in utility between the voter and the candidate in which the voter will reward the candidate with a rating greater than zero. At utility differences greater than the voter tolerance, candidates recieve zero score or recieve no ranking.