[EM] Instant Runoff Voting 3-candidate elections - pathologies considerably more common than you may have thought

[Jonathan Lundell]

On Aug 27, 2010, at 3:39 PM, robert bristow-johnson wrote:
> 
> 
> On Aug 27, 2010, at 4:15 PM, Jonathan Lundell wrote:
> 
>> On Aug 27, 2010, at 12:48 PM, robert bristow-johnson wrote:
>>> 
>>> with Score and Approval, it's easy to mark your favorite candidate (Score=99 or Approval=1).  and we know that Satan gets Score=0 or Approval=0.  then what do you do with other candidates that you might think are better than Satan?  that question has never been answered by Clay.  and any answer must be of a strategic nature.
>> 
>> That is, to my mind, a fundamental problem with cardinal-rating systems--approval, borda, range, etc. Except for some trivial cases (notably two-candidate elections), the voting act is necessarily a strategic exercise.
> 
> in my opinion, if you can break down a multi-candidate election into two-candidate elections where the end result would not be different (the Condorcet winner would be elected) then i think you've gotten past the burden of strategic voting.
> 
> in my opinion, when Candidate A is preferred by a majority of voters to Candidate B, then, if Candidate B is elected, some kind of anomaly or pathology (that might encourage one to vote strategically in the future).
> 
>> With an ordinal method (IRV, Condorcet (though one could, I suppose, specify a non-ordinal cycle breaker), Bucklin) it's at least possible (and usually a good idea) to cast a sincere ballot.
> 
> i just think that the strategy of swinging an election from the Condorcet winner into a cycle is just a risky strategy.  you never know who will come out on top; your worst enemy might just as well as the centrist may or as well as your candidate.  with something like Schulz (or Ranked Pairs which does not result in a different candidate with a Smith set of 3, and bigger than 3 seems to me even more unlikely than that of getting a cycle anyway), you are emphasizing more decisive elections in settling the ambiguity of a cycle.
> 
> say Ranked Pairs was the law, what kind of realistic strategizing can a party or group of candidate supporters do? that example of tossing a close election between RC (for radical center) and M (the centrist candidate who is also the Condorcet winner if sincere ballots are cast) is, in my opinion, too contrived to be a secure strategic guidance.  any strategy that can just as well backfire, is no strategy.

Let me be a little more clear. A voter, or group of voters, can always strategize (except in the case of a two-candidate plurality election, or perhaps a probability-based rule). Many voters are constitutionally inclined to strategize, and they're not necessarily good at it or rational about it. So I doubt we'll eliminate strategy with any of the methods that get discussed here.

My objection to cardinal-rating methods is that they don't just encourage strategic voting: strategy is required. If I top-rank my favorite, and bottom-rank my least favorite, then the question of what to do with any candidates in between is a strategy question pure & not so simple.

>> Why the latter is a good thing is a separate discussion...
> 
> why casting a sincere ballot is a good thing or not?  i would be interested in reading such a discussion.  and, maybe even, piping in.

By "sincere" here I mean "strategy-free" (the ballot, not the counting method). I'm sure we don't all mean the same thing by this. Here I mean something like: a ballot informed only by the voter's preferences over candidates, and independent of knowledge or conjecture about the preferences, behavior or possible strategies of other voters.