3 ms·
> IRV may not be perfect, but it is heaps better than the current system. This is probably not true. If you look at Voter Satisfaction Efficiency [1] if peopl
by jefftk 7y ago
> IRV may not be perfect, but it is heaps better than the current system.
This is probably not true. If you look at Voter Satisfaction Efficiency [1] if people are voting tactically then IRV actually does worse than the current system.
Approval voting is much more robust, and does very well even when people vote tactically.
[1] http://electionscience.github.io/vse-sim/VSE/ http://electionscience.github.io/vse-sim/VSE/
- oehpr 7y agoThat result is a rather strange anomaly. As far as I can tell it's not explained. Given that the system of ranking this is based on a simulation of agents, it may be unwise to jump to any conclusions about what "strategic voting" looks like here.