Now that everybody has digested somewhat their voting trauma, it is time to reopen the wounds. 😱
The voting procedure for the BVWSC consisted up to now in me simply winging it, but last month, I didn’t wing it correctly, and dukkha ensued. First, I got trigger happy and announced a winner when some important members (ahem @[email protected] ahem) had not yet voted. In order to prevent this from ever happening again, what I would like to propose is that we formalize a bit more the process, in the sense that there should be some fixed amount of time (48 hours?) for each of the soap nomination thread and the voting thread, and that a winner can be announced only after this time.
Second, there is potentially an issue with the voting procedure and count. What I have done so far is to allot a number of points which is inverse to the rank you proposed. That is, assuming three proposed soaps, the soap ranked first gets 3 points, the soap ranked 2nd get 2 points, and the soap ranked 3rd gets 1 point. Then, I simply add up these points, and announce the winner. This is similar to a Borda count (which I didn’t know, but you can read up here and more in detail here. Borda counts are unproblematic as long as every voter ranks every soap - which has not been the case in the last vote, and which may have cost Westman Shaving’s Noir the victory. @[email protected] suggested that someone who did not rank all soaps should be counted as attributing some points to the soaps he did not vote for. That is, assuming our example with 3 soaps, someone who ranked only one soap (which therefore gets 3 points) should also be counted as attributing in the end some points to the other two soaps. I see Dave’s point, and am sympathetic to it, but I don’t know exactly what would be the best formula.
The easiest method in that spirit would be that the unranked soaps get to share the points that could have been, but have not been, attributed (which would make 1.5 points for each remaining soap - which could also be rounded down to 1 point?). Alternatively, we could also require that any voter provide a full ranking. Any opinion on these proposals, or idea or other proposals how to better implement something similar, but better?
I don’t have any intrinsic preference, other than I would like to have a method that can be easily implemented in a spreadsheet (by me, without much knowledge of spreadsheets).
Thank you in advance for your input, and if you have any other complaints/proposals/questions about the process or BVWSC more in general, please feel free to do so.
I was going to suggest instant runoff ranked voting, but it looks like @[email protected]’s linked thingy does that.
The advantage of that is that we don’t have work, the downside is that if someone wants shenanigans, they can participate and skew the vote, even impersonate us.
We can probably consider this my professional paranoia going amok and give it a try until we run into actual problems?
There’s a setting to require inputting a username, so that would help a lot.
But what’s holding me back from voting using your username?
I thought of a solution which I’ll post under the main thread.
There’s a requirement for signing in with your google or facebook id - which should be safe enough, but will require new mail addresses (unless you’re happy to doxx yourself)
Google or Facebook logins? Let’s not
I would be fine with creating a burner gmail address for this purpose.
Though this is perhaps more friction than we want for the voting process?
It also wouldn’t be too difficult to self-host this and throw it on an authenticated route. I’m not very familiar with the architecture of lemmy, but in theory we could leverage our existing username/password to access the RCV app.
After thinking about this for a few minutes, this is almost certainly over-complicating things.
Why don’t we try disabling anonymous voting and have folks input their usernames? I think we can trust that people won’t try to game the system.