[dragonwriter]

> I still don't think Condorcet is complicated.

Conceptually, no. But we’re talking about, mostly, tallying complexity. Condorcet itself, and therefore some but not all Condorcet methods, is “simpler” than IRV in that the number of values that need to be tracked is smaller (at least where the ratio of the size of the electorate to the number of candidates is large, as it typically is in public election of officials), but that's not true of all Condorcet methods.

> One of the major disservices to Condorcet is that the vote-theorist communities insist on talking about the loop-breaking algorithms as being Condorcet methods, rather than as separate tie-breaking (loop-breaking) processes that are bolted on after identifying the (usually 1-member) Smith Set. Some of this is because the algorithmic implementations do both steps at once, but it's not necessary.

That's not a “disservice to Condorcet”, its a necessity for the evaluation of real-world election methods.

[tunesmith]

I mean it's a disservice to communicating and messaging about Condorcet by making it seem more complicated than it is. There are countless Condorcet variants, of various complexity. To look at the complexity of those variants and paint with a broad brush by saying that "Condorcet is complicated" does a disservice to the aim of seeking broader acceptance of Condorcet.

An example is that the Condorcet Method (identifying the usually single-winner Smith Set) is not subject to the same various "voting flaws" that the various tie breaking mechanisms are. But then the community is happy to say that Condorcet is flawed along those lines when it's not.