5 ms·

Any voting system that incorporates mean scores will in no way be immune to strategic deviations. (imagine a tie between your second ranked candidate and some o

by ulucs 6y ago

Any voting system that incorporates mean scores will in no way be immune to strategic deviations. (imagine a tie between your second ranked candidate and some other one, you simply inflate your score of your second ranked candidate by one) Also isn't achieving the lowest regret trivial, as the other voting systems never report anything cardinally?

sova 6y ago
In regards to your first statement, that's absolutely the point of Score Voting, to have a clear leaderboard with 1st, 2nd, 3rd and so forth. With regards to your second statement, "because we cannot access data means the data never existed?" Bayesian regret exists in all elections with all candidates, just because it's not "reported" doesn't mean it's not statistically there. The point is that Score Voting is a superset of all voting systems, and everything else is a blurrying of the detail that is available only in Score Voting. Bayesian Regret is just used to demonstrate that it will yield the highest rated candidate overall, which is logical given how it works. Nevertheless, examine the science for yourself and I believe you will soon find that all voting systems are some sort of blind or ignorant form of Score Voting. Means or averages are convenient, but you could simply take the greatest total score as well.