Is before, with Strong Condorcet vs Symmetrical ICT, I'm going to list some disadvantages of Strong Condorcet in comparison to Approval and Score. Then I'll ask what redeeming advantages Strong Condorcet has, to outweigh those disadvantages.
But, in this case, I'll supply an answer for you, though you might have other answers too. I'll discuss the answer that i supply. Again, I'll re-state what Strong Condorcet is: It refers to some versions of unimproved Condorcet that are popular with EM members. It includes Beatpath, Ranked-Pairs, River, Goldfish, Kemeny, VoteFair, and maybe others too. Kristofer defined it as methods that meet the (illegitimately-defined) Condorcet Criterion, and a few weakenings of IIAC. Maybe Kristofer included another requirement too, but I'm mostly defining Strong Condorcet by naming some methods that the term includes. Some disadvantages of Strong Condorcet, in comparison to Approval and Score: 1. Favorite-Betrayal Criterion (FBC):. No need to go into detail. You know that Strong Condorcet fails FBC, and what the societal consequences of that are. Strong Condorcet advocates speculate that people's voting won't be affected by FBC failure. The purpose here is not to adjudicate that question. 2. Later-No-Help (LNHe): That has been discussed here too. We agree that, when making out your ballot, it's better if you don't need to vote for additional candidates, in order to fully help those for whom you've already voted. Later-No-Help compliance considerably simplifies strategy, especially u/a strategy. Strong Condorcet gives voters a strategic need to rank unacceptable candidates in reverse order of winnability, due to Strong Condorcet's LNHe failure. 3. Participation: That's an embarrassment criterion. It's about a ridiculous thing that Strong Condorcet will do. When opponents of an enactment proposal tell everyone about Strong Condorcet's Participation failure, that will be fatal to the enactment proposal. 4. Consistency: Same as above, though Participation is probably worse. Strong Condorcet fails FBC, LNHe, Participation, and Consistency. Approval and Score meet FBC, LNHe, Participation, and Consistency. 5. Count-fraud vulnerability: I won't repeat about this again, because I've already thoroughly discussed it here on EM. Suffice it to say again that all Condorcet methods are incomparably more count-fraud-vulnerable than Approval and Score are. Approval is by far the least count-fraud-vulnerable of the voting system reform proposals. 6. Arbitrariness, innumerable versions: There are innumerable ways to count rankings. The question will be asked "Why your way instead of one of the infinitely many other ways?" Opponents will make good use of that infinity of rank counts, to obfuscate the issue. Rank-balloting advocates haven't, can't, and won't agree on a proposal. Approval, in contrast, is the obvious, natural, minimal improvement on Plurality--the simple and obvious result of repealing Plurality's forced-falsification rule that is the cause of Plurality's inadequacy. No arbitrariness. No issue, debate or controversy about how to count the ballots. So here is my question to Strong Condorcet advocates: What advantages do you claim for Strong Condorcet, advantages not possessed by Approval and Score, that outweigh the Strong Condorcet disadvantages that I've listed above? As I promises, I'll supply an answer for you: The Condorcet Criterion (CC). Allow me to comment on that: CC is a valuable and desirable strategy improvement--when it's genuinely available. That's why I like Symmetrical ICT, because it's designed to make CC available, by eliminating the worst strategy needs. But we're talking about Strong Condorcet, not Symmetrical ICT. Strong Condorcet doesn't share Symmetrical ICT's properties of eliminating those strategy needs. There are several things that will make CC unavailable with Strong Condorcet: 1. FBC failure. (Again, Strong FBC advocates speculate that it might not be a problem, and this isn't the place to adjudicate that claim) 2. Chicken dilemma. No, I don't claim that Approval and Score don't have it too (and deal with it nicely). I merely mean that the chicken dilemma, when it occurs, causes a non-sincere-voting need that would wipe out the availability benefit from CC. 3. CC isn't available if the CC-complying method can't be enacted. Above, I told of some reasons why Strong Condorcet would be difficult or impossible to enact. Having already discussed them, I'll now just list them: Participation-failure. Consistency-failure. More complicated and arbitrary definition and count rule. 4. Count fraud: CC benefit isn't available if the winner is being decided by count fraud. -------------------------------------------------------------------------------------------------------------------- Those four problems to the availability of CC benefit, I claim, wipe out the claim that CC outweighs the Strong Condorcet methods' disadvantages with respect to Approval and Score. But maybe Strong Condorcet advocates have some other Strong Condorcet advantages, in comparison to Approval and Score. Or maybe they disagree with one or more of the four CC availability problems that I listed directly above. Strong Condorcet advocates are invited to answer my question about what mitigating advantages Strong Condorcet has, to outweigh the above-listed 6 disadvantages, or to tell why every one of the 4 CC availability problems that I listed directly above won't be a problem. Mike Ossipoff But ---- Election-Methods mailing list - see http://electorama.com/em for list info
