Essay talk:Voting

ListenerX's remarks

Most of this looks fairly good to me; small points of rebuttal are:

• I do not think that the trouble rising from the vote to ban MarcusCicero had very much to do with the margin by which it passed. It had to do with the fact that among those who voted against the ban were a small number of editors who were willing to cause trouble over it. If it had been the whole electorate against them in the vote, they would have used some other excuse to make trouble.
• In installing such a specific codification of policy, we would have to be careful not to let it become a sitting duck to be gamed by malefactors. ListenerX^TalkerX 05:22, 18 September 2011 (UTC)

"If it had been the whole electorate against them in the vote, they would have used some other excuse to make trouble." But that excuse probably wouldn't have been backed up by prominent wikilawyers, including mods. I agree that we might not want to get extremely specific; as it was pointed out below, we'll probably want no more than two divisions in each area of vote distinction. That means two types of thresholds, two types of durations, etc. Blue (is useful) 19:55, 18 September 2011 (UTC)

goat

If 'goat' is just a present vote then - for categories 1 to 3 when there are only two options - it is, at least when it comes to the formal result, a 'no' vote in every way but officially. Could there be some rule as to what should happen if the 'no' votes reach the numbers required for a 'yes' vote to pass? e.g. the subject should be dropped completely for a certain period. This would differentiate between the two options more. Peter ^{talk, or type, or whatever...} 05:59, 18 September 2011 (UTC)

I don't think it should be counted in a simple yes-no vote - take for example a vote on the banning of a user, if you don't want them banned, vote no. But if it's something like the poll on talk pages currently going on, it becomes a sort-of "none of the above" and so it should be counted - if no option can reach 75% then none of the options are good enough. Of course things could be cleared up by having an actual "none of the above" option on CS polls. Then again, looking at the poll results so far, 75% might be too idealistic. -- Nx / talk 10:34, 18 September 2011 (UTC)

A "goat" vote should be seen as a spoiled ballot. It's a way for somebody to say they oppose / don't understand / couldn't give a stuff about the current vote, but it should have no impact on the final result and shouldn't be taken into consideration in any way. --PsyGremlinSnakk! 11:53, 18 September 2011 (UTC)

Take this unlikely but possible scenario: a vote to adopt a measure to RW:CS results in 10 for, 2 against, and 25 goat. Would that vote really have resounding support? No. Blue (is useful) 19:58, 18 September 2011 (UTC)

sorry but a 10:2 Vote with 25 abstentions is a clear yes vote. That 25 people dont care does not impact the actual vote. Hamster (talk) 20:09, 18 September 2011 (UTC)

It's about semantics. Votes need to be supported; it doesn't matter if it "technically" passed if most people aren't going to care enough to enforce the result. Blue (is useful) 20:11, 18 September 2011 (UTC)

its not about fucking semantics , its about being a dick or acting responsibly. If a question is do we do a, or b , or c and someone votes D, goat or micky mouse they have pissed there chance to decide an outcome awa6y, and can just fuck off if they dont like the result. This attitude of "that was fun" , lets ignore it in future and go through it again is why this site is constantly arguing about shit that should have taken 5 minutes to decide , and stay decided. Hamster (talk) 20:25, 18 September 2011 (UTC)

If people here always acted responsibly and never acted liked dicks, we would have no need of rules at all, let alone voting thresholds or a standard way of counting votes. I'm with you in spirit, but in practice I personally think we can't ignore "goats," as much as you and I'd like to. Blue (is useful) 20:34, 18 September 2011 (UTC)

I always thought of goat as abstain, or not a vote one way or another. sterile^talk 21:26, 18 September 2011 (UTC)

I'm with Psy here, a goat vote should be interpreted as the equivalent of a spoiled ballot, and those are never counted into the tally. If people really cared about one of the options, they'd vote for it. If they don't, they'll abstain. "Goat" means "I've seen what's going on, but none of the options are acceptable to me and/or I'm unwilling to publicly take a stance on this issue." Fine, but that shouldn't count towards the threshold. Röstigraben (talk) 13:28, 19 September 2011 (UTC)

Mountain Blue's remarks

This essay was long overdue; it's a good thing someone finally took the time to write it. Blue is right; it would make things easier if there was a universal, binding "standard of voting" agreed upon once and for all. Blue is also right in that we should have this discussion while there is no policy debate and no drama already on the horizon again. The analysis is exactly as systematic as the topic needs and the answers proposed are a sound basis for discussion. Thank you, Blue.

Regarding some of the the specifics:

• I believe the voting standard should be as simple as possible. The standard will never have any great authority if it's convoluted and confusing. In particular, I believe there should be only one standard of eligibility. Different requirements for different types of vote may open the process to boring meta discussions about which class the current vote is an instance of. It would provide an attack to people interested in driving a wedge into the thing. Two months and thirty edits looks reasonable to me.
• I believe there should at most two different passing thresholds, one stricter threshold for Category 1 and 2 and one easier threshold for Category 3 and 4. I'd be OK with Category 3 getting the stricter threshold too; I'm picturing it classed with Category 4 because the essay basically defines it as "meh, everything else". Id' be happy with either a two-thirds or three-fourths requirement for Categories 1 and 2.
• I don't think a Goat vote should be treated as a Present because that would, well, make it a Nay in all but name, exactly as the essay points out. I've always seen the Goat as a way to express incomprehension as to why the vote is being held in the first place ("this is useless drama") or at least as a form of total abstentation ("whatever; who gives a shit"). But of course that was just my personal impression of why people were voting Goat.

Mountain Blue (talk) 11:28, 18 September 2011 (UTC)

I feel ignored. --★uːʤɱ socialist 13:22, 18 September 2011 (UTC)

Who are you? Have we met? Mountain Blue (talk) 16:35, 18 September 2011 (UTC)

• I agree that the standard should be as simple as possible; that doesn't mean we should sacrifice comprehensiveness for simplicity. Four separate franchise requirements would indeed be overkill, but one universal requirement might be stretching it. I would be comfortable with two classes of eligibility, one which could vote in all categories and another which could only vote in categories 3 and 4.
• Thinking practically, you're right. My estimates for different thresholds might be overly idealistic; two thresholds would probably get the job done. I would think two-thirds for categories 1-3, and the yeahs-4>=nays rule for category 4.
• The trouble with "goats" is that they can mean very different things to different people who use them. Because people are taking the time to participate in a vote, even if they're not taking a particular side, I think the easiest way to sort out "goats" is to count them toward the total but otherwise disregard them. As I said above, a vote that resulted in 10 for, 2 against, and 25 goat simply wouldn't have the support of the community it would need. Blue (is useful) 20:06, 18 September 2011 (UTC)

In such as case, what would the functional difference be between a goat and a nay vote? DickTurpis (talk) 21:47, 18 September 2011 (UTC)

There would be no functional difference... except goats would not be an expression of specific disapproval. A person voting "goat" might be persuaded to vote yea on a similar issue. The difference becomes more semantic than functional. Blue (is useful) 23:43, 18 September 2011 (UTC)

It really depends on the context. If the standard is simple majority, goat votes could keep a matter from passing, as would have been the case with say 6 goats in the MC railroad instead of 2. The margin would have been 1:1 for/not for instead of the slim 5:4. If it's plurality, you could have the goat option winning. So in my opinion goat is a bullshit vote. People should vote their conscience or abstain. You can vote a spoiled ballot in a private vote, but you're not going to get one with options for candidates or referenda that don't exist. 00:38, 19 September 2011 (UTC)

I think two tiers is better, as well. Serious Votes and Minor Votes.--^talk 01:09, 19 September 2011 (UTC)

There's no harm in that, I think. We might want to preserve article deletion as a special case, though. Blue (is useful) 01:40, 19 September 2011 (UTC)

Maybe just leave article votes out of this sort of scheme entirely. Voting on brainstars, frontpage, deletion, templates, etc. can all just keep being ad hoc and free.--^talk 02:24, 19 September 2011 (UTC)

I included them to be as inclusive as possible, so people couldn't say "oh but you forgot about x vote" or "what about votes to do y?" I don't think we'd suffer from leaving what I call category 3 and 4 votes ad-hoc and free, or at least keep the existing Community Standards for them. Blue (is useful) 02:28, 19 September 2011 (UTC)

Good thinking. So in a simplified version, then, it's Standards change: Two months + 30 edits, 75% of the vote, two weeks duration; Ban or Promotion change: two months + 30 edits, 2/3 vote, one week duration? I'm combining your subtypes of 2.--^talk 03:19, 19 September 2011 (UTC)

Who wrote this?

I am fully in agreement with the thrust of this essay. I was going to add a template to credit the author when I found that two users seem to have been have been involved. To whom does this this essay belong?--Bob^{Spring is sprung!} 13:48, 18 September 2011 (UTC)

Blue. Doctor Dark edited it but then reverted. -- Nx / talk 16:20, 18 September 2011 (UTC)

Yeah, I thought it was like WP where essays in project space are open for anyone to edit. Sorry about that. Just pretend it never happened. Doctor Dark (talk) 22:58, 18 September 2011 (UTC)

Tiny, obvious thing.

There are pages for which a five-day waiting period to delete would be undesirable: spam, racist attacks, pages revealing IRL information or consisting of nothing but a personal attack, another useless essay or article courtesy of <insert name here>; these should be deleted sooner rather than later. B♭maj7 "If two men say they're Jesus, one of them must be wrong." 16:52, 18 September 2011 (UTC)

Good point, I've added this to the "duration" section of the essay. Blue (is useful) 19:03, 18 September 2011 (UTC)

Might a week be better? Just avoids the problem of weekend? Just a thought. sterile^talk 20:00, 18 September 2011 (UTC)

I'm not sure what you mean... Blue (is useful) 20:08, 18 September 2011 (UTC)

Some people don't edit on weekends, hence if you have a vote over the weekend, they have less time to vote. Other editors are much more active on weekends, and hence if you had a M-F vote, they'd miss it. sterile^talk 21:32, 18 September 2011 (UTC)

That makes sense. Seven days might be the best timeframe after all. Blue (is useful) 23:52, 18 September 2011 (UTC)

Speedy/obvious deletes could be shorter. sterile^talk 02:14, 19 September 2011 (UTC)

Flagging likely socks

I'd like to see a policy in place wherein users can flag votes that are likely to be sockpuppets of other users, though they may just meet eligibility requirements. Of course, this could stir up quite a bit of controversy, and the decision of whether flagged votes will be discounted would have to be taken very seriously, and clear a high bar. Let's face it, there are various socks and joke accounts out there which have been around a while and have enough edits to technically qualify, but really have no business voting. This proposal discusses known socks, but the unknown ones, and the ones in which the "real" editor is unknown, can potentially turn an election. DickTurpis (talk) 21:57, 18 September 2011 (UTC)

This is why we need to implement a "real first name and least initial" policy. And night mode. B♭maj7 "If two men say they're Jesus, one of them must be wrong." 21:58, 18 September 2011 (UTC)

Ha ha. Seriously though, wouldn't it behoove us to have something in place to thwart transparently obvious attempts to game the system? I'd like to think no one is registering multiple accounts, logging just enough minor edits to enfranchise themselves, and then trying to tilt an election in their favor, but if they do it would suck to have to count blatant chicanery such as that because they technically meet the qualifications, and we're effectively powerless to do anything about it. Hell, I've got at least one joke account that meets voting eligibility, but I'd like to think if I ever tried to use it people would rightly cry foul. DickTurpis (talk) 22:37, 18 September 2011 (UTC)

I don't think we should install CheckUser, but I don't think there's an all-around good way of solving the sock voting problem without it. I think you've proposed the best way of dealing with unknown socks, which is a definite problem, but it won't be pretty. Blue (is useful) 23:51, 18 September 2011 (UTC)

Not a problem that can be solved. If someone is committed enough, he can game the system through socks, and that's just the way it is. Social pressure and custom is the only good restraint on this, and when people fail to abide by custom then they should just be castigated with our contempt, like when Human used his sock in the recent talk page vote.--^talk 01:07, 19 September 2011 (UTC)

Digesting

The page might be more digestable if the voter eligibility, threshholds & runoffs and duration were listed under each category. I can't remember that long down the page. sterile^talk 02:06, 19 September 2011 (UTC)

Not a bad idea. I was thinking of creating a little chart to make it easier to read whatever organized proposal we come up with. Blue (is useful) 02:08, 19 September 2011 (UTC)

Notification

Don't forget notification standards. These go hand in hand with vote durations. You don't want the other kind of secret ballots. Also, the notification should reach all eligible voters as a minumum and by default (no need for anyone to go looking or "opt in" for notifications). LowKey (talk) 02:57, 19 September 2011 (UTC)

Great point. I think we should address when and how to notify the electorate at some point, perhaps not at the same time as coming up with the voting standard proper. Blue (is useful) 04:44, 19 September 2011 (UTC)

Urge more co-ordination

Since UHM already started one of these here I propose that these be worked together, or at the very least cross referenced or something to tie them together. It seems like you're trying to reinvent the invented bits of the wheel that was only half invented in the first place. Why keep re-starting and ignoring what has gone before? LowKey (talk) 03:09, 19 September 2011 (UTC)

I don't know, his is in userspace? I didn't know about UHM's bit when I wrote this essay, sorry. Blue (is useful) 04:18, 19 September 2011 (UTC)

No problem. I put mine in userspace because I wanted input on it all. Let's have one RationalWiki: space page where some input can be made and bigger proposals can be linked to. (with tl;dr versions). I think that would fit the coordinating bill better, than putting everything on one giantic page. --★uːʤɱ netlabelist 14:59, 19 September 2011 (UTC)

That's a very good idea. Perhaps a project space page or a forum page, like we did with the May reform debate (Forum:Consolidated site rules discussion). Blue (is useful) 19:26, 19 September 2011 (UTC)

Tabular representation of proposals

This table represents one version of a final proposal, dividing votes into "major" and "minor" types:

	Major votes	Minor votes
Encompasses	Community Standards changes User penalization	Editing policy changes Article deletion
Franchise	30 edits/two months	30 edits/one month
Threshold	two-thirds (with runoff)	plurality
Duration	1-2 weeks (user penalization) 2-4 weeks (RW:CS changes)	at least 1 week

This table represents a final proposal which distinguishes between "policy" and "penalty" votes, leaving my categories 3 and 4 alone:

	Policy votes	Penalty votes
Encompasses	Community Standards changes	Banning a user Penalizing a user less drastically
Franchise	30 edits/two months
Threshold	three-fourths (with runoff)	two-thirds (banning) majority (other)
Duration	2-4 weeks	1-2 weeks

These have both been approximated from people's comments over the last day or so. I hope this helps. Blue (is useful) 04:38, 19 September 2011 (UTC)

Crank up the edit requirements, by at least double, is my thought.--^talk 04:48, 19 September 2011 (UTC)

60 edits/two months? Makes sense, especially as a way to deal with socks. Blue (is useful) 04:53, 19 September 2011 (UTC)

Actually, now that I think about it, maybe we should go through and look at some various socks and see how many edits they have. That can give us a good idea of where we should be on the edit counts. Imma go look now.--^talk 05:06, 19 September 2011 (UTC)

Yeah, I'm going to go with sixty and two months.--^talk 05:09, 19 September 2011 (UTC)

I'd say at least 60, maybe more like 100 for the major votes. Also, I assume that when we say "account" we do not include unlogged-in ips, correct? DickTurpis (talk) 12:52, 19 September 2011 (UTC)

I'm also in favor of a higher number. Even 100 is still rather low, especially when you set it against the other requirement, two months of membership. Really, how many non-sockpuppet editors are there who jump into a vote on community issues and are able to make an informed decision after spending so little time here? Getting the feel of how the place works, internalizing its priciples and getting to know the other people takes a little more effort than that. Sure, every requirement can be subverted by a sufficiently dedicated puppeteer, but there's no reason not to make it a little harder. Röstigraben (talk) 13:20, 19 September 2011 (UTC)

This is the first I've seen people explicitly stating a preference for excluding all "sock puppets." What if I hypothetically have an alternate good faith account with something like oh say 1,000+ edits who isn't a dick like Nutty Roux is, that I've hypothetically been lying about having for years and have never previously voted with because it hypothetically felt weird and I hypothetically didn't want Nx to use his server access to out me, why should I not be able to hypothetically vote? I'm aware of others who may hypothetically have accounts with more edits who hypothetically never voted on anything. Should these hypothetical accounts, which are hypothetically run in complete good faith, be able to vote as a general principal? Coming up with an arbitrary criterion for excluding other accounts would nonetheless indicate you all believe each physical person behind the keyboard should only vote once. I might hypothetically like to start another account to edit differently than the other hypothetical account. I might then hypothetically like to vote 3 times. What moral, ethical, prudential, or other concern underlies your preferences that even these kinds of (hypothetical) accounts shouldn't vote? 13:46, 19 September 2011 (UTC)

"One person, one vote" is an almost universally recognized principle, and I don't see any need to reinvent the wheel here. You seem to be suggesting this hypothetical other account has a different personality, and is therefore a separate person (who might hypothetically have different views and might vote contrarily to how Nutty Roux would)? Very intriguing, but I'm not sure we are subject to a sort of "Citizens United" rule for accounts. In any case, I'm just talking about obvious attempts to game the system. Your hypothetical example wouldn't seem to raise any red flags, and therefore, however we may like it or not, would have no trouble voting. DickTurpis (talk) 13:57, 19 September 2011 (UTC)

I'm not talking about red flags or gaming the system. I'm asking the basic question of whether it really is one man, one vote, or as they can be read to imply the franchise rules would permit a single person to vote for every (hypothetical) account he has that qualifies.

There would need to be rules surrounding known socks, but since this is a wiki (and a wiki without CheckUser or other verification features at that), there is simply no way to sort out unknown socks. The best solution is to set a high bar for voting eligibility, even if it's a very imperfect system. But there is no way to avoid users creating socks and getting them into the franchise as you described. Not without a voting extension that checks IPs, and even that wouldn't stop dedicated sockmasters.

We can condemn that practice and create harsh penalties for it, but we are just unable to avoid it. Blue (is useful) 19:00, 19 September 2011 (UTC)

I'm assuming one man, one vote, but I'm open to hearing others' opinions on this.DickTurpis (talk) 17:55, 19 September 2011 (UTC)

I would imagine one man one vote too. But if we are counting edits there is another possible way which would both make socks irrelevant and give a greater say to those who have more interest in the site.

What we could do is give people a number of votes based upon their number of edits. For example for every one thousand edits (or whatever) a user would get one vote. In such a case it wouldn't matter if such a voting right had been obtained by a sock or not. I'm not saying that we should do this - just pointing out that it is an option.--Bob^{Spring is sprung!} 18:12, 19 September 2011 (UTC)

One word: Ken. One person might need three edits to add a wigo entry because they screw up the formatting. Another might only need one edit to post a new article they've been writing in notepad. Are those edits worth the same? -- Nx / talk 18:16, 19 September 2011 (UTC)

Exactly. Edit counts are notoriously bad indicators of "community value," first of all, and we would be faced with a ridiculously easy way to game the system by making thousands of tiny, bad edits. Blue (is useful) 18:52, 19 September 2011 (UTC)

Yeah, that would only lead to pissing contests and endless bickering about the relative value of edits. One editor, one vote is the way to go, coupled with a reasonable threshold to keep at least the neglected socks and overeager newbies away from the polls. The solution doesn't have to be perfect, just something to reduce the chance of voting-related HCMs. Röstigraben (talk) 19:12, 19 September 2011 (UTC)

I agree. 20:17, 19 September 2011 (UTC)

Getting back on the number of edits I would make the number as big as possible. It might be an idea to only count edits to article pages, but that would probably be going too far.--Bob^{Spring is sprung!} 15:24, 19 September 2011 (UTC)

At 1000 edits, that would leave about 120 or so accounts with voteing privileges, and several are bots and admitted socks. 15:29, 19 September 2011 (UTC)

The last time this was discussed, I suggested a mainspace quota, maybe something like 10% of the total threshold. I'm not so convinced by that idea anymore, though - there are people who almost never touch mainspace except to revert vandalism, but are nevertheless very active regulars who care about the site. Going with a higher total edit count is better than setting a mainspace quota that would have to be rather low so as not to be too exclusive. Röstigraben (talk) 19:12, 19 September 2011 (UTC)

True. It seems like people are leaning toward something that might look like this: For any vote, the franchise consists of users with at least 100 edits and a registration date at least two months prior to the conclusion of the vote. Two franchises for two types of votes does not appear to be popular. Blue (is useful) 21:03, 19 September 2011 (UTC)

It should be noted that the criteria for voting in moderator elections and elections to the RWF Board are three months, 75 edits. Blue (is useful) 21:23, 19 September 2011 (UTC)

That actually makes sense - longevity is of slightly more value in understanding the issues than participation (as measured by edit count). I'd go with 3 months, 100 edits to split the difference.--^talk 04:11, 21 September 2011 (UTC)

I agree. I don't think 3 months/100 edits sets an unreasonably high bar, but it might help with casual sockmasters. Blue (is useful) 04:33, 21 September 2011 (UTC)

Actual mechanism to record votes

What is the best way to record votes? This question applies to yay/nay votes, pick 1 from a list votes, and pick a number from a bigger number votes. I am not offering answers; just saying that the question is important. LowKey (talk) 06:54, 19 September 2011 (UTC)

Listing signatures and timestamps under each option seems to present the fewest complications. Blue (is useful) 19:34, 19 September 2011 (UTC)

Aspect of the franchise

I think it would be beneficial, if we're leaning strongly toward one single franchise standard, to keep that consistent with the eligible usergroup. That way, it would be simple to shift votes between secret and public ballots, and we could create an intercom group to notify eligible users of votes without throwing up notices that even everyday visitors can see. Blue (is useful) 20:50, 20 September 2011 (UTC)

Never mind, apparently the intercom thing is not possible. Blue (is useful) 03:45, 21 September 2011 (UTC)

Another correction - apparently it could be possible, according to Nx. Blue (is useful) 00:20, 22 September 2011 (UTC)

An Alternative

Since it looks like yet another vote (Mod-Squad mode 2) has resulted in madly attempting to find the ducks so that someone can look to thinking about beginning to get them in a row, I feel obliged to suggest a simple alternative which has the merits of enacting community will almost as consistently as the recent methods, as well as at least being a simple and clear procedure. It is also equitable, as everyone should be equally unhappy with it. LowKey (talk) 03:36, 21 September 2011 (UTC)

Nonsense, if anything the recent debacle has made it more imperative that we get the standards sorted out. Blue (is useful) 03:48, 21 September 2011 (UTC)

I acknowledge that the above is an over-reaction, but it was intentionally so - high sardony, if you will. Sometimes Usually even when being earnest I find much humour. Here, I was being humourous but also making a point. <harrangue> The point is this -> while many are here seriously discussing all of this because they feel that it is important that voting is handled "properly" (and all that that implies) there is vote after vote after vote going on where the "proper" things are mere afterthoughts, maybe even unthoughts or even worse. Sure, formal standards are not yet, erm, formalised but the underlying considerations are receiving short shrift indeed. The values that you are trying to uphold here do not actually appear to be held in any regard where it actually matters (i.e. in practice). Why would that change when the values are formalised? I am having very little faith that anything that gets decided here will actually be enacted. Most community decisions at RW are almost academic. RW has a strange pathology. </harrangue> LowKey (talk) 04:52, 21 September 2011 (UTC)

p.s. pardon the pun. LowKey (talk) 04:52, 21 September 2011 (UTC)

If and when the voting standard is accepted into the community standards, I believe there are many people who will make every effort to see it enforced. That tends to happen around issues that gestate for a while here. Blue (is useful) 04:58, 21 September 2011 (UTC)

Moving forward

I think the issues presented here have gestated well enough. A good next step will be to create a forum in which specific proposals can be discussed:

• Factoring in "goat" votes
• The franchise
• Types of votes
• Thresholds and runoffs
• Durations
• The role of mods in creating vote rules
• Dealing with socks

I'll be gathering the salient points made so far and creating Forum:Voting standards soon. Blue (is useful) 00:28, 22 September 2011 (UTC)