OT
Not true at all. Tell me, in any realistic scenario with multiple options, how are you going to know in advance how the relative comparisons and any loops that form are going to play out? You can't. It's impossible. Polls are never going to break down how each voting block views each candidate relative to every other one. What percent of the electorate prefers A to B? A to C? A to D? A to E? B to C? B to D? B to E? C to D? C to E? D to E? And it gets worse the more options you have. All of that information is never going to be captured by polls - and even if they did, the margin of error would be too great for that information to be useful.
With approval voting, it's so simple that even a child can tell without being informed that if you want to bury someone who poses a threat to your preferred candidate, you simply don't approve them. That's it. It's that simple. That obvious. And that effective.
If you look at the literature, Condorcet methods are *far more* suspectible to "burying" than approval voting, and they're susceptible to *insincere* burying, while approval voting is only susceptible to *sincere* burying.
(Every system is susceptible to sincere burying.)
Polling data directly translates. If A is at 30% and B is at 29% and the remainder is split by third parties, and you like both A and B but you'd prefer B, are you going to approve A? Of course you're not. Because you're not an idiot.
If polling data is at 30+-3% / 29%+-3%, and C is polling with a chance of winning, yes, you approve A. Because yes, the strategy is obvious.
The discussions of strategic voting in approval (where strategic voters generally can't have much effect on the voting of non-strategic voters) vs. Condorcet (extremely susceptible to strategic voters reducing the effective voting power of non-strategic voters, though not as bad as e.g. the Borda count) are all over the literature now.
Approval voting inherently will be voted strategically. Pretty much every time. In practice, it's effectively "First-past-the-post, plus pity votes for the people I feel have no chance of winning but I like regardless"
And in practice, this means what you call the "pity vote" candidates win pretty often.
Which is the whole point.
As for being easy to understand... hey, the easiest system to understand is dictatorship, but that doesn't make it the best! Next easiest is first-past-the-post, but most people don't like that one either! While loop resolution can be tough to explain, the basic Condorcet principle is very easy to grasp: "If any choice would beat all other choices, it's the winner." Simple. Fair.
But not.
It doesn't find a winner frequently, and when it does, *it may not even be the most popular choice*. It sounds at first glance as if "would beat each other option head to head" is a good criterion, *but it isn't*, if you're trying to find the option which the most people tolerate. Satisfying Condorcet means violating consistency, violating participation, and causing favorite-betrayal to work particularly if you have *less* information.
Your assumption that strategy in Schulze is "too complicated" to use is essentially security through obscurity -- there are high incentives to use insincere tactical voting, and even if everyone votes honestly, you still get IMO undesirable results.
Bottom line for me:
I consider the majority criterion undesirable for government. It basically discourages more-popular compromises and leads to organizational breakups. Most of the time when there's a Condorcet winner, approval voting will in practice select the Condorcet winner; but when it doesn't, approval is doing the right thing and Condorcet is doing the wrong thing, IMO.
Here's a sadly-not-nearly-hypothetical-enough scenario.
30% of the population is hardcore evangelicals, with extremist opinions, who will always bullet-vote for their theocratic candidate and be angry if anyone else wins.
21% is softcore evangelicals, who think the theocratic candidate is the best and feel some sort of moral duty to vote for them *and* rank them in first place, but are open to compromises and want to get along with everyone so will vote for / rank other candidates.
49% will vote for anyone but the theocratic candidate, and are terrified of the theocratic candidate and will be angry if he wins.
Work out the result of approval voting vs. Condorcet. See which you think creates a more peaceful society.
If you still like Condorcet in that scenario (which is all too real right now), I will not burden you with further comment.
