[EM] Condorcet meeting

[Kristofer Munsterhjelm]

On 2023-08-24 15:55, Colin Champion wrote:
> Also, I'm not sure if cardinal voting (such as Star) solves anything. 
> For most voters, sorting a list of candidates has quadratic work, and 
> has to be avoided if the field is large. But is cardinal voting any 
> better? If the scores had an intrinsic meaning, you could work through 
> the list and assign the correct value to each candidate. But they don't, 
> so you'll start by assigning tentative scores until you find an 
> inconsistency - eg. two candidates have the same score but vastly 
> different acceptability - and then you have to adjust the numbers you've 
> already given. You'll probably end up sorting and grouping into quintiles.

My idea with approval was just that voters would think (at least I 
would) "I'll just approve anyone who I think is reasonably good, we'll 
sort it out in the general". While having to use approval to elect a 
single candidate would be very hard (what's the other voters' scalings? 
what should mine be? how do I deal with the Burr dilemma? etc), it's 
easier when the particular ballot is much more accommodating of noise.

This would fail if we get to a multiparty/candidate system that's 
vibrant enough that there are more than six viable candidates. At that 
point we'd be kind of out of luck; at least the overhead of being 
careful with one's vote would return.

-km