User:SMcCandlish/RfC-replacement voting system

I participated in an academic survey of WP decision-making practices a while back. There were a bunch of questions about whether the usual RfC practices work well, and what I thought of alternatives including secret-ballot (direct democracy) ideas, among others. I have non-rigid opinions on this stuff, but have come to the following tentative conclusions:
 * Our open "!voting" system favors a meritocracy over mob-rule, because comments (which are not really supposed to be interpreted as votes) are generally tied to rationales, which everyone can see; "closers" (assessors and declarers of the consensus results) can discount pure votes without rationales, and comments with demonstrably faulty rationales.
 * It is seriously flawed in that the first few commenters who provide rationale have a undue effect on the outcome (e.g. editors like me who subscribe to a constant stream of RfCs via WP:FRS and look for RfCs that FRS doesn't tell us about, have a disproportionate effect on WP governance. It's also broken because of increasing failure by closers to discount no-rationale and bogus-rationale !votes thus turning them into actual votes (processes like RM and RfC are increasingly just vote-counting with less and less closer regard for policy- or source-based validity of arguments presented). A less-frequent problem is "super-voting", i.e. the closer discounting well-reasoned comments simply because the closer subjectively disagrees with them.
 * This could all be mitigated by a replacement system in which pro and con arguments (about a proposition, or one for each of several multiple-choice options) were presented in a table, and below each was a section for refutation arguments, and below that a space for rebuttal of the refutation. Each of these could be in successive collapse boxes so as not to text-wall people.  This is essentially the format used on paper by the Voter Information Pamphlet series put out in major cities by the League of Women Voters; it can make very complex propositions much easier to understand, though it is not totally immune to manipulation or oversimplification (nothing is).
 * The actual voting would then be done by [temporarily] secret ballot, closure of the argumentation-presentation period, above.
 * The first half of this could simply be done with templates, and a policy change (e.g. that anyone is empowered to revert as disruptive any vote like "Support" or "Oppose" wrongly placed in the presentational material, since it would not be an actual rationale but just exhortation/campaigning). The second half would require an actual secret ballot system be deployed. It would also require a "voting is open on X" notification system.  Finally, we would have to have a standard for what percentage of votes is needed for which kinds of proposals; this should include some kind of proportional/weighted voting system. It may sound complex, but our XfD processes are actually more difficult to use than this.
 * It will probably not happen within my lifetime, if ever, for the same change-phobic inertia reason that causes us to still have a terrible adminship system, a Mediation Committee that doesn't work, and various similar problems. It would likely have to be imposed by WMF the way the Arbitration Committee, the Copyright policy, and several other policies and procedures were.

Wikipedia's internal decision-making process has evolved quite markedly over time. All decisions were originally just open discussions (or, rarely, fiat declarations by Jimmy Wales), with a loose consensus emerging ... or failing to. By the late 2000s, potentially contentious decisions were conducted via explicit polls. RfCs also existed, but really were requests for comments, i.e. open discussions in which broader input was sought. Since the early 2010s, these merged, with RfCs becoming polls but (ideally) interpreted more like old-school consensus discussions.

Today, they are primarily just polls, despite denialism to the contrary. If we're going to proceed with a polling model (i.e. a vote even if we don't call it one) – and it looks like we are – then for long-term viability it should be done in a way that achieves all of the following: minimizes the early-commenter effect, cancels the mob-rule or "pure democracy" effect of always-visible votes (which induce pressure to vote with the majority), avoids the flaws inherent in winner-take-all voting systems (by us adapting one of the well-tested "distributed vote" systems, at least for anything with more than two options), and reduces the problem of the poster of the question/options often being able to shape the outcome by clever and biased wording. Accountability would have to be ensured by votes becoming public after poll closure, at least long enough for independent verification of the results (by anyone), e.g. to detect meatpuppetry.