View on GitHub


By Jameson Quinn, statistician/theorist

Voter Satisfaction Efficiency, or VSE, is a measure of the quality of a election method. A VSE of 100% would mean an impossibly perfect method; 0% or lower would mean that the society would be better off picking a winner at random.

To calculate VSE, you simulate thousands of elections, using voters who cluster on issues in a realistic way. Since the voters are simulated, you can know exactly how satisfied they would be by each candidate; that is, how close the candidate is on the issues they care about.

I’ve run VSE simulations for a number of methods. Here’s a chart with only the best and most well-known methods:


Some other methods I’ve tested, but (aside from IRV) left out of the figure above, include:

Other assumptions about approval voting:

If you want to see a figure which includes all of the above methods, see here.