@Krazer How do you feel about crowdsourcing the mascot design? We could make a meta post asking for character designs, with some minimal guidance as to what characteristics we want. Then let everyone work on designing something for a week or two.
I think the guidelines we should put forward are basically what you said earlier, Krazer. "The idea [is] to have something simple, represent a concept, [be] cute but not overly [cute], and easily distinguishable"
We could also put forward things about the -tan families, and such things. But if we're totally putting it out there, we don't want to be too limiting.
@LoganM Is that a personal preference, though, or a collective one? Because if it's personal, obviously the human ones are more likely to be voted up anyway.
We could just crowdsource the whole thing. Make a meta post just asking for mascot characters. Let people comment their suggestions on each of the posts. In the end, if there's one which has a lot of upvotes, go with it, otherwise back to the drawing board.
@Eric We'd have to tell people to be conservative with voting. We don't want something ugly or terrible being our mascot. It should be clear when there's a community consensus. I think it's better to leave it undetermined for now, and we'll decide if there's a good enough choice when it appears.
Also we should encourage downvoting submissions you don't want to be the mascot.
@Eric Because otherwise the mascot could be determined entirely from one subgroup of the community which really likes a nomination without any input from another that doesn't like it.
I suppose that's true, but I think upvoting would suffer a similar fate if that were the case.
It basically just means every person gets twice as many votes. Downvote candidate A and upvote candidate B is the same as B getting +2 and everyone else (except A) getting +1.
@Eric Not quite. In terms of the vote totals, you're right, but newcomers are at a disadvantage if we had +2/+1/+0 since they won't get as much time to gather votes. With +1/+0/-1 it's fairer. We don't want the FGITW problem with our mascot.
@Eric In the +1/0/-1 model, I don't think what matters is the vote total. The vote count is important. If something is at +16/-5, I think it's a worse choice than a +10/-0 candidate.
@Krazer I think the contingency is to abandon the idea entirely for now. I don't think it's that crucial, and if we need to wait a while to get a good mascot then so-be-it. Even if we come up with ideas ourselves, we should be vetting them with the community in the same way as the other ideas.
@Eric I disagree with this. A downvote isn't conveying the same information as a lack of an upvote. A downvote is saying "I have a problem with this" while an upvote is saying "I am in favor of this". There is plenty of room in the middle.
+1/0 also has the same FGITW problems as +2/+1/0. +1/0/-1 alleviates all of them if we do a Bayesian parameter estimation for a binomial distribution to determine which is the highest (though if it's close we're better off doing a runoff).
I think the point is that we want to check not only that there is significant support for the nomination, but that there isn't significant opposition to it either. This isn't democratic; we're aiming for a consensus, not a majority. We can't do that with +2/+1/0 or +1/0 since we don't get a count of how many people have seriously considered the nomination. +1/0/-1 solves the problem IMO.