OldSchoolNoBull wrote:AEnigma wrote:I read your point in that seven people gave votes to Jokic in some capacity and seven people gave votes to Barkley in some capacity (assuming there were no Barkley/Jokic ballots), but a) it is not some injustice that Jokic won, b) Jokic did have more first place votes, and c) Barkley is probably going to win this next vote with his main competitor being the guy to whom you gave your lead vote. I am sure it is annoying, but you had time to count votes and reach out to non-voters (or non-alternates) to create a clearer sense that Barkley’s support was even with Jokic’s support.
I will wait for Doc to make the call.
Like I said, I didn't realize that in this type of case - 5-4-4 - that no secondary votes from anyone who voted from those three would be considered. I'm not sure if we've actually had this situation so far in the project, so I was unaware, otherwise I would've changed my vote. It just strikes me as anti-democratic to not let it go to a runoff given the actual vote-counts and I hope Doc sees it that way.
It is an injustice if a runoff would potentially give Barkley a win at #26 instead of #27.
EDIT: Again, I don't think we've been in this situation so far, and there's nothing in the OP about how votes are to be counted in that situation.
Continuing the conversation here. First, let me just lay out the process I've been using:
1. Tally all #1 votes. If majority, end.
2. Go to #2 vote for all who don't finish in Top 2 of previous count. Add to prior #1 votes. If majority, end.
3. Anyone previously qualifying in Top 2, who now don't? Go to #2 for their #1 votes. If majority, end. Repeat if necessary & possible.
4. Extend deadline one day, quote voters who didn't vote.
5. Create runoff thread with poll on PC Board. Whoever's ahead after one day wins.
I believe we had a #3 situation previously, but I could be mistaken. #4 has obviously happened. #5 never has.
/+-
It's understandable that you're frustrated that this process isn't specifically stated in either OP. The reality is that this is something that evolved over the lead up to the project as we added things like #2 votes.
From what you've written, I believe you were expecting something more like:
1. Tally all #1 votes. If majority, end.
2. For all candidates who finish Top 2, perform a one vs one runoff taking the #2 votes from all other candidates not in that particular one vs one. If victor, end.
3. Extend deadline one day, quote voters who didn't vote.
4. Create runoff thread with poll on PC Board. Whoever's ahead after one day wins.
I can understand why you'd think this superior and not much more work in practice. If people advocate for it, I'll consider it, but:
I made clear from the beginning that I wasn't looking to long vote lists for Instant Runoff Voting, and that also precludes an exhaustive Condorcet voting mechanism. This was a statement rather than an argument, but to give some reasons:
1. Too much voting complexity is bad for the project runner simply because of the greater work involved.
2. Too much voting complexity is problematic for the voter because it makes them feel like they need to have pre-made lists available before the actual discussion, which makes it easy for voter disengagement.
3. Too much voting complexity is problematic for the reader because even if it doesn't cause there to be less discussion, it tends to separate the actual live argument about a given player from the thread in which he is Inducted.
4. Not even Condorcet is fool-proof. You can get yourself caught in rock-paper-scissors situations. Sure we can break any tie from any method by soliciting the opinions of more people, but we shouldn't kid ourselves that there's a perfect way to weight votes in order to always get something that represents the most valid solution. Imperfection is an inherent part of voting systems, and once we accept that, it's just a question of what our priorities are.
5. I've said that my priority is discussion. We can arguably get just as good of a list by soliciting Top 100 lists from people on thee board and averaging them. We could potentially achieve this in hours/days, rather than a 10 month process. So then, why bother with something so inefficient? Because that "inefficiency" is caused by communication and rumination which is then publicly available years after the fact.
I'll say again that I'm not looking to tell people to "shut up and let me drive". By all means people can make cases for tweaks to the approach I'm using, and I may make said tweaks just as I've done several times already. But in the end I will be the judge, because I've taken on the responsibility of carrying this out, and in general, close races that might have ended up differently with a different process are not something that are a surprise to me. I see them as something inevitable, and in such cases, the concept of the ranked list gets exposed for its superficial nature. No one-dimensional scale can possibly capture the shape of these players, let alone one that implies an equal gap between all increments of it.