You Should Know

The bigger problem in my opinion is more about the fact that all elections that select a single winner will always end up in stupid degenerate systems like this where flaws and imperfections exist.

The best thing to do (again, my opinion) is to abolish all single winner races and have multiple winners with proportional representation. Get rid of directly elected presidents and have a prime minister selected by a proportionally representative parliament instead. All presidential systems suck, and the larger the number of people voting, the harder and harder it sucks. It's not just a USA problem - you also see it in France and Turkey, where they also have an all-powerful president that is elected nationally and the election is a complete shit-show every time without fail. On the other hand, having a prime minister selected as the head of state from a proportionally elected parliament is a much fairer and more stable system in my opinion. It has downsides too of course, but nowhere near as bad as nationally elected presidential systems.

In any case, the example you pointed out is a potential flaw in approval voting, but I don't think it's very likely to happen. First of all, it would require all those voters in the second round to conspire a particular way, which isn't very likely. Secondly, there's the fact that the numbers would have to line up in a very particular way which has a very low probability of happening - tweak a few numbers here and there, and the spoiler effect vanishes. Sure, the scenario you point out is a hypothetical flaw in approval voting, but I think it's a much smaller effect and probability of actually influencing anything - definitely nowhere near as much of a strategic voting effect as in plurality voting systems.

Congrats, we have a glorified FPTP and spoiler effect yet again!

Not quite. As you've just observed, this kind of strategic voting is risky, and self destructive. Which means that many would recognize this, and not use this voting strategy. Its a game of chicken, and lots of people prefer not to play such a game and instead support the safe bet, which means supporting those you genuinely support.

And as [email protected] pointed out, it isn't possible to have a perfect voting system.

Then there is the fact that there is more to this than just voting strategies. There are the other effects to keep in mind. For example approval is far simpler to explain than RCV, especially when you explain how the counting works.

Another example is that approval is purely an additive process for counting, RCV is not. That means auditing results is significantly easier and quicker under approval than RCV. That leads to higher voter confidence in results than RCV audits.

Now, other election systems could also have strategic voting, but its less likely with, for example, RCV, since you can rank candidates.

RCV still can experience the spoiler effect just as FPTP, because it is in effect FPTP taking place over some number of instant rounds.

It's mathematically proven that no voting system is perfect, so you'll have to choose what you value most in a voting system. There is no clear best system.

I think the arguments for approval voting are strong. It's simple and easy to understand, no need for complex multiple rounds of counting. Since you can't rank candidates it doesn't suffer from the spoiler effect.

In your hypothetical scenario you're forgetting the Trump voters who will vote for Sanders again after they see Biden nearly beating him. RCV has a bigger problem which is called the spoiler effect where, without strategical voting a loser can influence the election results. And am I missing something or should the numbers in your RCV list add up to 100%?

https://electionscience.org/education/approval-voting-vs-rcv