what's the word for when you elect multiple peeps (and thus can't elect a peep twice) vs when you're electing a party (and thus can have multiple winners)
-
Risottoreplied to deilann v -0.2 :neodog_hyper: :neodog_nom_verified: last edited by
@deilann that's two voting systems in parallel
each of the result outcomes is a demo, I don't want to combine the rows
-
deilann v -0.2 :neodog_hyper: :neodog_nom_verified:replied to Risotto last edited by
hmmm
the david can't win twice scenario sounds like a multi-round plurality situation where you have an instant-runoff with david removed but not as a loser
-
Risottoreplied to deilann v -0.2 :neodog_hyper: :neodog_nom_verified: last edited by
@deilann hmmm, is IRV the one for FPTP but multi?
-
deilann v -0.2 :neodog_hyper: :neodog_nom_verified:replied to Risotto last edited by
@risottobias IRV is one of those concepts that just depends on how it's being implemented
it can be in FPTP when there's no plurality, removing some candidates or for a secondary seat
-
Risottoreplied to deilann v -0.2 :neodog_hyper: :neodog_nom_verified: last edited by
@deilann hmmm, what would you pick for the 12 cells?
balancing ease of calculation vs other factors
-
deilann v -0.2 :neodog_hyper: :neodog_nom_verified:replied to Risotto last edited by [email protected]
not fully sure i get what you're asking but in this "multiple instant runoff" situation i'd only recalculate "david's" (obviously necessary, he's no longer running) votes and the votes of those who voted for "weaker" candidates to simulate voter-correction (feeling like their preferred candidate won't win the second vote)
-
Risottoreplied to deilann v -0.2 :neodog_hyper: :neodog_nom_verified: last edited by
@deilann hmmm
maybe there aren't good algorithms for different cells...
like I know STAR gets used for single winner stuff?
-
deilann v -0.2 :neodog_hyper: :neodog_nom_verified:replied to Risotto last edited by
i think the problem is that the second round alters preferences. so it's not just that we've lost a candidate but that the outcome changes how people vote.
-
Risottoreplied to deilann v -0.2 :neodog_hyper: :neodog_nom_verified: last edited by
@deilann well, some systems also make distinctions between voters "already being represented" by the first rep, vs carrying a harmonic series (or other fractional) amount of oomf to elect the next reps
-
@deilann this also complicates things in the case of surplus votes for a candidate,
e.g., which ballots count as spent and which count as still unrepresented? because their second choices might be different.
do you random shuffle the votes?
(this is also where some different methods shortcut it by assigning weights)
-
deilann v -0.2 :neodog_hyper: :neodog_nom_verified:replied to Risotto last edited by
thinking... eliminating the lowest candidates and then reducing the total number of votes by slightly less than those lost by dropping the poorest performing candidates might simulate overcorrection well
essentially, it accounts for those who still stand by their vote, but shifts preferences for the fraction that would
-
Risottoreplied to deilann v -0.2 :neodog_hyper: :neodog_nom_verified: last edited by
@deilann I think I should get the dataset, and compute any method people propose on it.
so, if people stop by my sponsor booth at pride and are like "aaaawww I like meek-STV" I'll be like "it's now on my todo list"
but otherwise I'll calculate ones that are easy to understand.
one of the librarians I met was also a voting nerd
-
Risottoreplied to Risotto last edited by [email protected]
@deilann (the work-around for this is also the same shuffles we've both discussed with Jefferson/whatever you called it and the other one - as /candidates/ get seats their fraction on the remaining rounds adjusts, so you don't need to fuss about with which /voter/ was rep'd)
-
deilann v -0.2 :neodog_hyper: :neodog_nom_verified:replied to Risotto last edited by
every time i ask a librarian for help there's that one librarian who is exactly what i need :neodog_heart: