@jack-waugh then perhaps the solution, though it would somewhat complicate the tallying process, is to allow each voter to select a baseline score for unknown candidates. I think there is a way to do it with just one count of the ballots and an algorithm that encodes baselines for unknown candidates by tallying up to a set representing known candidates and then adding that score to any candidate not in that set as a second (and much quicker) summation round.

@gregw there is a cyber security technique called "fuzzing" in which attacks are simulated with random data. The VSE simulations seem to provide a framework for fuzzing, where in this case the random data would be some kind of strategy. Developing a genetic algorithm to evolve a strategy that breaks a system would be an interesting side project. When I get the spare time and energy, I'll see if I can cook one up and set up a computer to just chug away at it until I have some results. (I wrote and published a library called bluegenes in case anyone wants to try stapling libraries together before I get around to it.)

@bmjacobs An open source solution using cryptography would go a long way toward making the system robust against accusations of rigging. You could do something like the following: 1) take a cryptographic commitment of each ballot using sha256; 2) sort the list of ballot hashes; 3) concatenate and hash them together or progressively hash them into a single sha256 hash; 4) use the resulting hash as the seed for a CSPRNG (cryptographically secure pseudorandom number generator). This procedure is deterministic and thus impossible to rig, but the output will be impossible to predict and functionally random.

@toby-pereira very interesting. Many ancient cultures employed sortition for allocating responsibilities or making decisions, though perhaps this will be a more difficult sell without the appeal to "the will of the gods" becoming rhetorically effective again.

@bmjacobs An open source solution using cryptography would go a long way toward making the system robust against accusations of rigging. You could do something like the following: 1) take a cryptographic commitment of each ballot using sha256; 2) sort the list of ballot hashes; 3) concatenate and hash them together or progressively hash them into a single sha256 hash; 4) use the resulting hash as the seed for a CSPRNG (cryptographically secure pseudorandom number generator). This procedure is deterministic and thus impossible to rig, but the output will be impossible to predict and functionally random.

@gregw there is a cyber security technique called "fuzzing" in which attacks are simulated with random data. The VSE simulations seem to provide a framework for fuzzing, where in this case the random data would be some kind of strategy. Developing a genetic algorithm to evolve a strategy that breaks a system would be an interesting side project. When I get the spare time and energy, I'll see if I can cook one up and set up a computer to just chug away at it until I have some results. (I wrote and published a library called bluegenes in case anyone wants to try stapling libraries together before I get around to it.)

@toby-pereira very interesting. Many ancient cultures employed sortition for allocating responsibilities or making decisions, though perhaps this will be a more difficult sell without the appeal to "the will of the gods" becoming rhetorically effective again.

If we're going to put candidates into a tier list, we ought to include an S tier worth >1 point. I am curious about what effect that would have. Is the code used to run the ABC simulations open source?

@cfrank the main issue with STV is that it is fairly complex, making it somewhat challenging to implement and also to follow the algorithmic logic with any real detail. I read through the ballot tallying report for an Australian Senate election a few years back, and it was awful and tedious -- iirc it was over 60 pages long. By comparison, a cumulative vote tallying report would just be one page of numbers.

It seems that MMP is a much simpler and easier method than STV that gives reasonable results. (Whether the official inclusion of parties is a problem or not is philosophical speculation considering that political parties exist in reality, but that is a separate matter.) Are there any other methods for proportional representation that are simple enough to be both practical and easily comprehensible to concerned citizens?

@kaptain5 I think that your first point only holds true in an election for which all known candidates are hated by the majority and thus rejected. I suppose I could test these theoretical issues under various conditions by forking the voter simulation project code and running some customized simulations. (Iirc, last time I looked at the code, it was incompatible with negative voting and secret/write-in candidates.) If I do, I'll make a new forum post to discuss the results.

To change the subject somewhat, I wonder if negative score voting with a default neutral score and no write-in feature would be valid as part of a legislative process. For example, it could be used to select sections to include in a draft piece of legislation, with every section with average or better positive score being included; this process could then also be used for addenda or maybe amendments. This might be a bit beyond the scope of what is typically discussed in this forum, but it is still interesting to contemplate.

It appears that cumulative voting is more popular for proportional representation in corporate elections to fill seats in boards of directors but not particularly popular for political use or among voting reform advocates. Why is that? In particular, cumulative voting for boards of directors is mandated by some state laws, yet it looks like those same states do not require cumulative voting for their own legislative branch. For example, Alaska makes cumulative voting the default for boards of directors, but its own legislative branch uses single-winner districts; as another example, Arizona requires cumulative voting for boards of directors but uses multi-winner plurality for its house of representatives.

In the US, cumulative voting has been used historically in some municipalities and in Illinois for 120 years until its repeal, and it has been mandated by courts as a remediation for National Voting Rights Act issues, but it seems unpopular among voting reform advocates. Why do voting reform advocates pass it by for more complex systems? Could this not be a simple, workable compromise similar to how approval voting is often seen as a workable compromise for single-seat election reform?

Are the problems with cumulative voting too great for it to be effective in practice? From what I have read, some places repealed it due to gridlock caused by minority representation, but representation of minority interests to oppose the majority seems to be rather a point in its favor.

@cfrank $50+ for an ebook is excessive. Is this a college text book?

@chocopi said in Approval Voting as a Workable Compromise:

$0.

I think we all know this figure is missing a few zeroes. /s

More importantly, this still leaves the issue of the Electoral College. How would we translate the EC district votes into EC votes under this system? Approval voting would be an immediate improvement for electing Congress Critters and state/local officials, but at some point we'll have to think through how to reform the Electoral College. The way I see it, abolishing the EC is too radical a proposition to have associated with any advocacy for an Approval Voting reform, so a plausible idea for how the EC could work is worth having. Simply leaving the EC the way it is while selecting Electors using Approval Voting would probably lead to more problems.

@cfrank It was just a random philosophical musing. People are very sensitive because modern politics are very polarized, but I wonder if anyone would care if the slightly less supported consensus candidate won over the slightly more supported consensus candidate -- if the level of consensus is within a typical margin of error, then statistically we cannot say that one is a better candidate than the other with reasonable certainty. I have seen some example elections and how various tallying methods could cause the less supported consensus candidate to win, but I am not convinced that it much matters if there is a sufficiently broad consensus for both candidates.