Votes transparency

[marcoscaceres]

The election result announcement lacks a breakdown of who received how many votes were cast overall, and how many for each candidate etc. Some breakdown of how many people voted and how the votes went would be useful to gauge the health and engagement of the AC in the AB selection process.

The data might also expose things like if members are only voting for very few (or maybe even just one!) candidates, which would hopefully mean we switch to ranked voting or something better.

[chaals]

A certain amount of data has usually been provided to the members - how many votes, how many rounds of distribution (something that is meaningful in the ranked voting we use).

Before we used ranked voting, there were indeed a substantial number of people who only voted for a single candidate. That was one of the motivations to switch - ranked voting systems provide incentives to rank a lot of candidates instead of only expressing an opinion about a single candidate.

By the way I think the Advisory Committee, who end up deciding on any change, are pretty sick of this discussion. It's been frequent, and (partly my fault) gone into a lot of detail about different systems.

[dwsinger]

The AB usually gets a report prepared by Ralph; it takes a little while, and given how overloaded Ralph is right now, it may take longer. If I recall correctly, the last report indicated that in both the STV and approval experimental ballots, there were a significant number who voted for only one candidate, implicitly saying that no-one else was acceptable. This worries me.

[cwilso]

Although as @dwsinger says, the AB usually gets a report on the balloting, and the information is passed through in the AB minutes (e.g. https://www.w3.org/2022/06/16-ab-minutes#t03), I will point out that STV (any ranked-choice system, really) is pretty opaque to understand the results of without the individual ballots being publicized. Even in the pre-STV days, I believe the vote counts for each candidate were not shared (except on request, privately, with the candidate themself), but there's not as clear "how many votes did each candidate get" with STV (you can come up with such numbers, of course, but "how many 1st choice votes did each candidate get" is also a significant indicator.)

Personally, as someone who is (I believe) tied for losing the most elections, I believe transparency would be much preferable. Not knowing how the voting went, and effectively WHY (except at the most abstract level) I lost, was quite frustrating.

I think we should be sharing the ballots (the voter being anonymous, of course) in order to enable recreating the voting rounds, and understand how support works; I think there is a lot of additional data that would be interesting as well (e.g. I believe some of the hosts discuss voting amongst their members; it would be interesting to have host-affiliation distribution to understand if that is increasing voter engagement.)

[cwilso]

(I'd point out, BTW, that we already ARE on "ranked choice voting" - STV is a common type of ranked choice voting for multi-seat elections.)