Wikipedia:Arbitration Committee/Audit Subcommittee/October 2009 election/SecurePoll feedback and workshop

Add links
Source: Wikipedia, the free encyclopedia.
Wikipedia Arbitration
Open proceedings
Active sanctions
Arbitration Committee
Audit
Track related changes
This page is for gathering
Arbitration Committee elections
.

The intention of this page is to gather feedback from participants, observers, and anyone interested in the use of SecurePoll for elections on this project. Some starter questions have been included, but please add sections for observations and questions. Threaded discussion is encouraged, particularly in the workshop section, so that ideas can be developed and improved upon.

Please note that the purpose of this page is to discuss SecurePoll, not the result of the Audit Subcommittee election itself.

Feedback

Should SecurePoll be used to elect the new arbitrators at the upcoming
2009 ArbCom elections?
This has been the subject of considerable debate for several weeks now, with opinions sharply divided. If you have time, please review the arguments at: Public polling versus Secret voting
and perhaps add your own comments.


Was the SecurePoll ballot easy to use?

Did you encounter any difficulties when voting?

Were you able to vote?

Other observations

Please add a new section for each.

General comments

  • (out of sequence post) This all being a relatively new process, it may take a time or two for everyone to get a feel for "where to say what". Having been through it once now, I would indeed have made more use of the individual candidate's talk pages. Thank you for your reply. —
     ?  08:28, 12 November 2009 (UTC)[reply
    ]
  • I think SecurePoll should be used for the upcoming ArbCom election. Cla68 (talk) 08:08, 12 November 2009 (UTC)[reply]
  • Agree here with Cla68 (talk · contribs). Cirt (talk) 08:15, 12 November 2009 (UTC)[reply]
  • I certainly like the appeal of a secret ballot is this type of simple voting system. A preferential system would be nice too, but I won't hold my breath. \ Backslash Forwardslash / (talk) 08:20, 12 November 2009 (UTC)[reply]
    • For information, there's a lengthy discussion about preferential voting at the ArbCom RfC.  Roger Davies talk 09:38, 12 November 2009 (UTC)[reply]
      • And technically, preferential voting is already completely supported; it's used for the WMF board elections. Happy‑melon 22:17, 12 November 2009 (UTC)[reply]
  • Agree with Cla68. Ks0stm (T•C•G) 08:47, 12 November 2009 (UTC)[reply]
  • Agree with Cla68. Casliber (talk · contribs) 10:07, 12 November 2009 (UTC)[reply]
  • Also agree with Cla68 ~ Riana ⁂ 10:30, 12 November 2009 (UTC)[reply]
  • Agree with Cla68
  • It would be lovely if there could be a delayed tally of sorts - after each candidate gets 10 votes and after 24 hours or something similar, the tallies could be updated. But that's just my random musing of the day. NW (Talk) 12:46, 12 November 2009 (UTC)[reply]
  • A bad idea. Someone could scrape tally and correlate it with the timestamps on the votes, even if it was only updated at regular intervals. — Jake Wartenberg 23:44, 12 November 2009 (UTC)[reply]

Do you think SecurePoll would scale up well for the upcoming ArbCom elections?

  • Not being a technician I can see no reason for it to be otherwise. But others may know/think better. Saga City (talk) 09:32, 12 November 2009 (UTC)[reply]
  • If something better comes up...Until then my vote for this system..:-) AruNKumaRTalK 09:37, 12 November 2009 (UTC)[reply]
  • Yes --CrohnieGalTalk 09:49, 12 November 2009 (UTC)[reply]
  • It appears to be ideal for elections like this. Warofdreams talk 10:11, 12 November 2009 (UTC)[reply]
  • I have no idea why not. Casliber (talk · contribs) 10:14, 12 November 2009 (UTC)[reply]
  • Given the current absence of a better system, absolutely ~ Riana ⁂ 10:31, 12 November 2009 (UTC)[reply]
  • Yes. I think it scales better than the public voting where you need to edit 30+ pages and risk running into edit conflicts. Sjakkalle (Check!) 10:34, 12 November 2009 (UTC)[reply]
  • I see no reason why it wouldn't, it's been used before on bigger elections. -- Luk talk 10:44, 12 November 2009 (UTC)[reply]
  • I don't see why it wouldn't. Ucucha 11:52, 12 November 2009 (UTC)[reply]
  • Yes. Go for it.Cathar11 (talk) 12:21, 12 November 2009 (UTC)[reply]
  • There is no reason for it not to. — Coren (talk) 12:24, 12 November 2009 (UTC)[reply]
  • No, though I appear to be in the vast minority here. SecurePoll works excellently, but this is a wiki where transparency is crucial. I'm not sure I see the point of hiding something like an ArbCom election when a regular voting page worked for years. –Juliancolton | Talk 12:43, 12 November 2009 (UTC)[reply]
  • I agree with Julian, I prefer the open vote per last election...Modernist (talk) 13:06, 12 November 2009 (UTC)[reply]
  • I'm an editor, not a sysadmin, Jim!  Sandstein  13:18, 12 November 2009 (UTC)[reply]
  • Yes. Prodego talk 13:20, 12 November 2009 (UTC)[reply]
  • Yes, absolutely prefer this system. Geraldk (talk) 13:35, 12 November 2009 (UTC)[reply]
  • Yup, for reasons along Sjakkalle's line of thinking. ~ Amory (u • t • c) 14:04, 12 November 2009 (UTC)[reply]
  • Yes  Cargoking  talk  14:08, 12 November 2009 (UTC)[reply]
  • Yes, and Sjakkalle brought up some good reasons. Novickas (talk) 14:28, 12 November 2009 (UTC)[reply]
  • Most certainly. RP459 (talk) 14:38, 12 November 2009 (UTC)[reply]
  • Yes, especially with regard to the comment about e/c'ing in public polls with high participation. Bfigura (talk) 14:54, 12 November 2009 (UTC)[reply]
  • It would scale well, yes. I think it ought to be an open vote, though—if SecurePoll can be so configured, that is.
    AGK 15:06, 12 November 2009 (UTC)[reply
    ]
  • Why wouldn't it? — Jake Wartenberg 15:12, 12 November 2009 (UTC)[reply]
  • Absolutely. Tony (talk) 15:15, 12 November 2009 (UTC)[reply]
  • Certainly. (Com certeza)
    Off2riorob (talk) 15:36, 12 November 2009 (UTC)[reply
    ]
  • Echo Sandstein's remark, only he got the line wrong. It should be "Damn it, Jim, I'm an editor, not a sysadmin!" RayTalk 16:39, 12 November 2009 (UTC)[reply]
  • Yes, absolutely -- Александр Дмитрий (Alexandr Dmitri) (talk) 17:23, 12 November 2009 (UTC)[reply]
  • Without a doubt. BigDunc 17:39, 12 November 2009 (UTC)[reply]
  • Yes. — Becksguy (talk) 17:46, 12 November 2009 (UTC)[reply]
  • Yes, but it would be a good idea to also have links to question pages for candidates and comments made by users accessible from the poll. Captain panda 17:51, 12 November 2009 (UTC)[reply]
  • Yes, provided it scales technically. I think Captain panda has a point - the interface should feature >read the candidate's statement >ask the candidate a question >vote for your choice of candidate. @Juliancolton - every single democracy in the world has come to the conclusion that secret ballot is essential to avoid all manner of shenanigans. While Wikipedia largely works by consensus, which must be transparent, this is a true ballot, and there should be no possibility of calling an editor into question based on who they cast their vote for. Elen of the Roads (talk) 18:54, 12 November 2009 (UTC)[reply]

Workshop

Should SecurePoll be used to elect the new arbitrators at the upcoming
2009 ArbCom elections?
This has been the subject of considerable debate for several weeks now, with opinions sharply divided. If you have time, please review the arguments at: Public polling versus Secret voting
and perhaps add your own comments.


Brainstorm on potential improvements, ask questions about the functionality, or propose changes here.


Should the voter list eliminate blocked users?
  • How long, prior to the election, is the list run? If it's a comparatively short period of time, I'd imagine that's just something a previously blocked user would have to accept. It would simply be a further consequence of their actions, and it won't prevent them from participating next time around. I'd support excluding blocked users from the list. Regards, --—Cyclonenim | Chat  08:35, 12 November 2009 (UTC)[reply]
  • I agree with Cyclonenim - it depends on how long before the election the list is prepared. My feeling is that it is like users with fewer than 150 edits- if someone is at 148 mainspace edits, they'll be ineligible, even though by the time of the election they may have the edits - it's the same for blocked users. -- PhantomSteve (Contact Me, My Contribs) 08:51, 12 November 2009 (UTC)[reply]
  • There's a SecretPoll module that checks real-time whether a user is blocked at the time they go to vote. (I forgot to ask the devs to turn it on for the AUSC election so it wasn't activated until 3 November.) With this module, blocked voters are redirected to a page telling them they can't vote while blocked. Once the block expires, they can go vote. What SecurePoll doesn't (yet?) do is to check the electoral roll before tallying the vote to see whether someone was blocked after voting. This needs to be done manually and is important in order to strike votes of subsequently discvered socks. In the recent AUSC election, the scrutineers ran through the list of voters for subsequently blocked editors before certifying the results. I did it also: it was trivially easy because an alphabetical list of voters was available.
  • Can the developers set up something to check the voting log for illegal voters prior to tallying everything up at the end of the sessions? I don't know a lot myself about how to do this but I know that there is a way to easily (?) put some program together to gather illegal voters to remove them. Just a thought, --CrohnieGalTalk 10:27, 12 November 2009 (UTC)[reply]
  • Actively blocked users shouldn't automatically be allowed to vote, but sysops should grant polite requests from blocked users to be temporarily unblocked with the proviso that they edit nowhere but pages directly related to the election or its candidates and agree to being blocked for the remainder of the original block after a reasonable period (e.g., 1 hour). Blocks are supposedly preventive, not punitive, and refusing to unblock to allow voting seems highly punitive. We wouldn't want to disenfranchise anyone because a lapse in judgment led to their being blocked, would we? Of course, if they abuse the temporary unblock, then good faith need no longer be assumed and further preventive measures, such as longer blocking, could be imposed. Concur with above users who noted that confirmed socks and banned users shouldn't be allowed to vote. Rivertorch (talk) 02:13, 13 November 2009 (UTC)[reply]
  • Per Philosopher. It's not always clear exactly what a particular block is intended to achieve, but, whatever is intended, it's not disenfranchisement. MoreThings (talk) 13:52, 14 November 2009 (UTC)[reply]

What order should the voting buttons be in?

See the sample HTML/JS in
coṁrá) 23:39, 12 November 2009 (UTC)[reply
]
Yes, HTML allows having no selection, but that means the only way to get into this special fifth state is to clear all your selections. That's not usable and not worth it. And SecurePoll isn't just a pile of HTML. Let's trust Happy-melon on the fact that this would interact badly with the software. rspεεr (talk) 01:47, 13 November 2009 (UTC)[reply]
Updated with an
example of clearing part of a form
. It was the HTML/JS HappyMelon was referring to. Let's not dismiss a UI developer on this (me) or turn it into a wrestling point between techies. It's straightforward (and fairly old) HTML/JS.
The web 1.0 way of doing it (i.e. no JS, just old HTML) is to have a "No opinion" option that is selected by default - on first visit - and can be reverted to on later visits. --rannṗáirtí anaiṫnid (
coṁrá) 08:59, 13 November 2009 (UTC)[reply
]
So now you want to have five buttons, including "neutral", "abstain", and "no opinion"? Why? rspεεr (talk) 09:47, 13 November 2009 (UTC)[reply]
Huh? I'm answering a technical concern someone else had (regardless of our technical expertese, there are always trivial gaps in our knowledge). The answer to that question is independent of the number of choices on the ballot paper. --rannṗáirtí anaiṫnid (
coṁrá) 10:02, 13 November 2009 (UTC)[reply
]
A JS solution is not usable; the form is "web 1.0" style precisely because modern implementations require JavaScript bells and whistles. The vote has to be accessible, and equally accessible, to any supported browser or platform, with or without JS. Some of the site JS is actually disabled on the voting page (as with the login and change-password pages) because of the obvious security risks; it's entirely plausible that, for increased security, a user might want to disable the remaining JS even on a modern browser. There are also issues with the backend (the code that's called to process the submitted form) receiving a "required" field with a value that's not in the list of options that were presented; they're not insurmountable, but they are a little bit awkward.
The best option to register "no opinion" is, as you say, to have a radio option specifically for such. What you want to call that ("abstain" seems logical, there's absolutely no useful semantic difference between "abstain" and "no vote") is a separate issue. Happy‑melon 12:26, 14 November 2009 (UTC)[reply]
Good reasons. Replied on your talk. --rannṗáirtí anaiṫnid (
coṁrá) 00:02, 16 November 2009 (UTC)[reply
]
  • S/N/O/A, with "Abstain" selected by default. As Happy-melon says, "no selection" should not be a meaningful state, so it makes sense to use an "Abstain" state instead. The other use for "Abstain" is to strike a previously cast vote; neither use of this is quite the same thing as "Neutral". — Gavia immer (talk) 22:22, 12 November 2009 (UTC)[reply]
  • I don't care about the order, but a feature—perhaps called "Abstain"—that clears any existing selection should be available. If the technical implementation requires that this is a radio button (e.g. to avoid the need for javascript on the client) that's OK on the understanding that it is in fact an action button and not a countable vote. Where the ballot form asks for votes on multiple candidates or propositions, it is acceptable and meaningful to abstain from voting on one item and this isn't the same as deciding not to participate in the entire election/referendum. - Pointillist (talk) 22:42, 12 November 2009 (UTC
    • Why would choosing "neutral" not be an acceptable way to abstain? rspεεr (talk) 09:44, 13 November 2009 (UTC)[reply]
      • Historically, "neutral" has been used (e.g. in the RfA process) as a specific type of vote, whose meaning was made clear by an accompanying comment. I'm not a great supporter of "neutral", but in ballots where it is used, it clearly isn't the same thing as simply not voting. - Pointillist (talk) 10:18, 13 November 2009 (UTC)[reply]
        • In RfA (which, unlike ArbCom, was not intended to be an election), neutrals don't change the outcome. From the point of view of the bureaucrats, it's the same as if you didn't say anything. They're just there for the comments. In a situation where there can be no comments, how is voting "neutral" different from abstaining? rspεεr (talk) 10:22, 13 November 2009 (UTC)[reply]
          • I think we agree really. If voting "neutral" is the same as abstaining, then that option should be called "abstain", not "neutral" because then the meaning is clear. If they aren't the same, then the meaning of "neutral"—and how it will be counted—must be explained as part of the ballot procedure. "Abstain" already has a clear meaning (dictionary). - Pointillist (talk) 10:43, 13 November 2009 (UTC)[reply]
  • S/N/O/A/...Modernist (talk) 23:47, 12 November 2009 (UTC)[reply]
  • S/N/O/A, A=default. Rivertorch (talk) 02:14, 13 November 2009 (UTC)[reply]
  • Only one button that reads "Your vote is irrelevant; the Party has already decided who won." @harej 12:28, 13 November 2009 (UTC)[reply]
  • s/n/o/a, a=default. Ncmvocalist (talk) 07:17, 14 November 2009 (UTC)[reply]
  • Three buttons. Any order. The abstain/neutral button as default. MoreThings (talk) 13:57, 14 November 2009 (UTC)[reply]
  • Support/Neutral/Oppose/Abstain BejinhanTalk 11:13, 15 November 2009 (UTC)[reply]
  • Support / Abstain / Oppose - Having none selected by default is difficult to do per above, so having abstain selected by default seems sensible. Neutral/abstain do technically mean different things, so I am open minded on having them as separate options, though for a vote the difference would be purely symbolic.
    talk 12:50, 15 November 2009 (UTC)[reply
    ]
  • three buttons with the neutral/abstain button be the default. ~~ GB fan ~~ talk 23:15, 15 November 2009 (UTC)[reply]
  • Agree with above, three buttons with the neutral/abstain button as the default. —mattisse (Talk) 00:20, 16 November 2009 (UTC)[reply]

How long should the session cookie last?

  • If voters open their "ballot" and then go back and forth between the ballot and the candidate pages, their cookie might expire before they have finished voting on the full slate of candidates. Thoughts? Risker (talk) 06:41, 12 November 2009 (UTC)[reply]
  • (If possible) Until they log out? Ks0stm (T•C•G) 08:51, 12 November 2009 (UTC)[reply]
  • Cookies! Oh, wait... yeah, I agree with the above. –Juliancolton | Talk 12:45, 12 November 2009 (UTC)[reply]
  • Until they log out as per above RP459 (talk) 14:39, 12 November 2009 (UTC)[reply]
  • Page session. Once they move away from the SecurePoll extension the cookie should expire. A person may take a long time to vote but navigating away from the SecurePoll page is a clear indication that they have "left". Signing out is a less certain guarantee. The user may never log out of their account. --rannṗáirtí anaiṫnid (
    coṁrá) 20:42, 12 November 2009 (UTC)[reply
    ]
  • Although the technicals remain unclear, I like the only-as-long-as-necessary approach. We should clearly state the expiration scheme to users when they log in. HereToHelp (talk to me) 20:54, 12 November 2009 (UTC)[reply]
  • As rannṗáirtí anaiṫnid says, the cookie should expire upon navigating away from the interface page. Anything else is carries an unneeded risk of having one's votes changed afterward. This also means that every link present on the interface page should open in a new tab or window; preferably this would be achieved automatically rather than relying on the page creator to get it right. — Gavia immer (talk) 22:36, 12 November 2009 (UTC)[reply]
  • Page session is good. Please do not use sodium benzoate to achieve this. @harej 12:28, 13 November 2009 (UTC)[reply]

How do we encourage people to comment?

Fully support the above - can't say it better! Saga City (talk) 09:36, 12 November 2009 (UTC)[reply]
Why would we comment/debate during the election? Am I misunderstanding this? Should the comments/debates be done prior to election? I would think that all this would do at election time is cause heated debates and disclose a voters secret vote. I didn't use the comment pages because I researched the editors in advance to voting so there was no need to go and comment. --CrohnieGalTalk 10:32, 12 November 2009 (UTC)[reply]
Fully endorse that. The whole point of using Secure Poll is for anonymous voting. If I then go on to praise or damn a candidate, that rather defeats the purpose. -- Александр Дмитрий (Alexandr Dmitri) (talk) 17:30, 12 November 2009 (UTC)[reply]
  • I believe the lack of idle debate is fantastic - streamlines the process considerably, removes the need to trawl through pages of argument and encourages users to make their own decisions. This mania for prolonged commentary that we've propagated in places like RfA is unnecessary, wearying and time-consuming - I'm glad that it's been curtailed here. Hopefully an indicator of future trends. And if people feel especially strongly, they will say something and they will call others' attention to the problem. ~ Riana ⁂ 10:36, 12 November 2009 (UTC)[reply]
  • Just to add to that - I enjoy the discussions when they actually go somewhere and mean something. My dislike is more of commentary for its own sake - feeling you have to justify everything to others because unless you've written an essay on an 'opppose', there's no way you mean it! ~ Riana ⁂ 00:42, 13 November 2009 (UTC)[reply]
  • Prominently link and mention the comment page, and give it a section with a format like RFCs or RfA voting. I think some people really enjoy the freewheeling discussions we have in things like RfA, whereas others differ. RayTalk 16:36, 12 November 2009 (UTC)[reply]
  • I think the lack of discussion is simply due to people not being used to this process yet. As soon as someone starts a discussion, others will follow. I considered starting a discussion but since the discussion sections were all empty, I was too hesitant (wasn't sure if it was the right thing to do). Perhaps a short explanation underneath the "Discussion" section header like "Comments and discussion of the candidate is encouraged here." would be enough to kick-start it. -kotra (talk) 19:14, 12 November 2009 (UTC)[reply]
  • Commenting areas should be completely separate to the voting area. People should be able to vote in peace without the comments of other bearing undue influence on their decisions. --rannṗáirtí anaiṫnid (
    coṁrá) 20:46, 12 November 2009 (UTC)[reply
    ]
  • Prominent linking to (discuss) links within the voting page. @harej 12:28, 13 November 2009 (UTC)[reply]
  • I agree with harej; more work needs to be put into linking. Ncmvocalist (talk) 07:22, 14 November 2009 (UTC)[reply]

Review and edit votes

Is it possible to review and edit votes prior to the final tally in the case where additional information comes to light? - 2/0 (cont.) 09:27, 12 November 2009 (UTC)[reply]

Yes. You can review and change your votes at any time.  Roger Davies talk 09:28, 12 November 2009 (UTC)[reply]
I couldn't see my earlier vote. After I saved my vote, the screen showed an empty vote card - with a change in the order of the nominees, but no visible record of my vote. SilkTork *YES! 10:08, 12 November 2009 (UTC)[reply]
It doesn't currently how the votes were earlier cast. It probably should.  Roger Davies talk 10:09, 12 November 2009 (UTC)[reply]

When you are reviewing your votes, SecurePoll doesn't show how your votes were previously cast

At present, when you review your votes, SecurePoll shows a fresh ballot paper. In other words, your earlier voting pattern is not displayed. In big elections, this is likely to be a significant time-sink as all votes need casting again. Shold this be changed?  Roger Davies talk 10:14, 12 November 2009 (UTC)[reply]

Yes. ++Lar: t/c 10:42, 12 November 2009 (UTC)[reply]
That would be a significant improvement. CWC 10:48, 12 November 2009 (UTC)[reply]
Definitely. –blurpeace (talk) 11:55, 12 November 2009 (UTC)[reply]
That would be best: set the buttons to what you voted previously as their defaults. This way, someone who wants to change one vote needs not revote on everything. (This was manageable with a handful of candidates, but could be a nightmare with dozens). — Coren (talk) 12:34, 12 November 2009 (UTC)[reply]
Yep. Might also be worth considering letting users view their own voting details. –Juliancolton | Talk 12:48, 12 November 2009 (UTC)[reply]
Show the previous votes as defaults. Novickas (talk) 14:38, 12 November 2009 (UTC)[reply]
That would be a big improvement. RP459 (talk) 14:40, 12 November 2009 (UTC)[reply]
Things should stay as they are. The change would represent an unneeded security risk. — Jake Wartenberg 15:18, 12 November 2009 (UTC)[reply]
Yes, because with more candidates I might not decide to cast all of my votes the first time. -- Александр Дмитрий (Alexandr Dmitri) (talk) 17:32, 12 November 2009 (UTC)[reply]
Yes. Captain panda 17:52, 12 November 2009 (UTC)[reply]
Yes. I had to vote a second time because of a technical glitch, and I'm there thinking "who did I vote for.....???" (I have a dreadful memory, and I only knew one of the candidates more than slightly, so all other votes were cast based on statement and a bit of research - that's why I hesitated) Elen of the Roads (talk) 19:14, 12 November 2009 (UTC) who hastens to explain that she didn't just vote at random.[reply]
Yes. Completely. --rannṗáirtí anaiṫnid (
coṁrá) 20:47, 12 November 2009 (UTC)[reply
]
Yes please. @harej 12:28, 13 November 2009 (UTC)[reply]
Yes, per Coren. - Jredmond (talk) 18:09, 13 November 2009 (UTC)[reply]
Yes, per my comment in the earlier section. I can't remember how many times I ended up voting on certain candidates. Ncmvocalist (talk) 07:26, 14 November 2009 (UTC)[reply]
Yes, absolutely, per the conversation above.--
talk • contribs) 09:26, 15 November 2009 (UTC)[reply
]
If we use SecurePoll, yes, please. Heimstern Läufer (talk) 10:36, 15 November 2009 (UTC)[reply]
Yes, this would be helpful if it could be implemented.
talk 12:54, 15 November 2009 (UTC)[reply
]
Yes, per many of the yes comments above, especially by Coren. — Becksguy (talk) 15:15, 15 November 2009 (UTC)[reply]
Yes, please ~~ GB fan ~~ talk 23:18, 15 November 2009 (UTC)[reply]

What's supposed to be the difference between "neutral", "abstain", and no selection?

I see many people advocating having separate buttons for "neutral" and "abstain", and some of them advocate yet another state of the interface where no option -- not even abstain -- is selected by default. Why are those all different states? What does that even mean?

Suppose two candidates have the same number of supports and opposes, but candidate A has more "neutrals" while candidate B has more "abstains". Also, a few people didn't click either "neutral" or "abstain" on candidate B, so candidate A has more recorded votes overall. Who won? Is this really a situation you want to contemplate? If there's any sanity to the system, then all non-votes will have the same effect on the outcome, which is nothing, so they shouldn't be called two or three different things.

I urge people to change their recommendations to keep it simple by having three possible states. Not four, and five is right out. rspεεr (talk) 10:39, 12 November 2009 (UTC)[reply]

I'd say there is a difference between abstain and neutral, but it's a semantic one, I'll admit. If I abstain from voting, I don't care whether the candidate wins or not and don't wish to register a preference. If I vote neutral, then I've consciously said that I don't care either way and have registered a preference to say that. It's a very minor difference: abstentions don't count in any way; neutrality does count, but in which way is up to the community to decide, if at all. To be honest, the ideal way to do this is to have it black and white, like with real voting in real life - yes or no. But the community isn't going to buy that. Even better, now I think of it, is the
❞ 11:09, 12 November 2009 (UTC)[reply
]
Indeed. I'd imagine there's a large difference between the two terms. I class abstain as the option to choose if I were in some way biased to that candidate, be it negatively or positively. I class neutral as a "I've never seen you before, you're good, but I have too many issues to overcome" sort of vote. Regards, --—Cyclonenim | Chat  12:12, 12 November 2009 (UTC)[reply]
Abstain may also mean that "I am too close to a specific situation, and choose not to vote" - so it also might be a conscious decision. In that situation, abstain can be a positive or a negative. Neutral is a conscious "I don't care". (
BWilkins ←track) 13:15, 12 November 2009 (UTC)[reply
]
Netural means you have no opinion toward a specific candidate and voted as such. Abstain means you will not issue a vote toward a specific candidate. No selection means you did not vote at all on any of the selection options toward a specific candidate. Chris (talk) 13:32, 12 November 2009 (UTC)[reply]
Abstain indicates lack of willingness to commit to any of Support, Neutral and Oppose - this may, amongst a host of reasons, be indicative of a flawed electoral system. Neutral indicates that having weighed up all the issues and being happy with the level of information supplied there is no commitment to either Support or Oppose. They are quite different so this is why neither should be a default. Saga City (talk)

While I agree that there is a semantic difference in the meaning of abstain and neutral, no one has really commented on how this should be interpreted in the context of a vote. If we're going to advocate for separate 'abstain' and 'neutral' buttons, we should have a firm definition of what exactly they mean. Because right now, if voting is done by a simple tally, things could get rather confusing --Bfigura (talk) 15:02, 12 November 2009 (UTC)[reply]

I don't see the purpose of having an 'abstain' button - abstaining is equivalent to just not voting. If you don't want to register your opinion on a particular candidate, then you should just not vote on that one. Robofish (talk) 15:40, 12 November 2009 (UTC)[reply]

In some electoral contexts, non-voting and abstention are two different things. For example, for personnel matters in my workplace (which has shared governance), we have four options--for, against, abstain, no vote. "No vote" is taken to mean, "I don't have enough knowledge to state a position," whereas abstention is generally seen as more negative, since the voter acknowledges the candidate and chooses not to vote in favor. In borderline cases, the difference is taken into account. This only matters in our case if Jimbo is going to take subjective matters into account when judging close cases, which I know is a controversial issue.
Chick Bowen 16:06, 12 November 2009 (UTC)[reply
]
Isn't abstain, the way you put it, simply being against in that situation? I'd see abstention in that case as being "I refuse to make a decision on this candidate", whereas no vote would mean "I don't have enough information to make the decision". Regards, --—Cyclonenim | Chat  17:16, 12 November 2009 (UTC)[reply]
Well, an abstention certainly carries more weight in a system, like the one I'm talking about, in which opposes are rare. Opposes don't seem to be rare in Wikipedia elections, so that's a difference. I'm just pointing out that there are plenty of electoral traditions in which a significant distinction between these four options can be made.
Chick Bowen 17:24, 12 November 2009 (UTC)[reply
]

Abstain is not the same as neutral. Neutral implies a vote in the middle between Support and Oppose. Maybe there are valid arguments on both sides, with neither side being sufficiently compelling as to overpower the other. However, to abstain is to announce a decision to not vote (a non-vote, as it were), and is effectively to recuse one's self publically, as compared to just not participating at all. One might abstain because of potential COI or prejudice, or because one doesn't understand all the issues, or have enough information, and can't fairly vote. Abstention is very different from just not participating—which could mean not noticing, not caring, or refusal to participate in the system—in that abstention validates the system as a non-vote. There is a big difference among abstain, neutral, and not participating (or absence). — Becksguy (talk) 17:34, 12 November 2009 (UTC)[reply]

Please clarify. Given that no human will see the ballots, what do you expect the computer to do about this "big difference"? rspεεr (talk) 01:53, 13 November 2009 (UTC)[reply]
I'd expect "no selection" to be the default state before a voter has recorded their vote for a proposition/candidate. "Abstain" should be a button that clears any existing selection, like striking out an existing vote. For Wikipedia purposes it should mean the same as not participating in that part of the ballot, it isn't a vote and it isn't meaningful to count it (except for psephological analysis). That's why—AFAIK—there's never been an Abstain section in any election here at en.wikipedia. Re: "Neutral", I'm not a fan of it for elections, but it could be useful for surveys of the love-like-neutral-dislike-hate variety. - Pointillist (talk) 18:13, 12 November 2009 (UTC)[reply]

Sorry for posting a similar section on the discussion page before getting down this far, but it seems clear that there is a semantic difference between "neutral" and "abstain", but no effective difference as far as this election is concerned. If we were using a system whereby a support vote gave 2 points, a neutral vote gave 1 point and an oppose gave 0 points, with winners selected by how many points they had gotten, it would make sense to have an option "between" support and oppose votes. However, for RfAs and it seems SecurePoll elections, we simply express support votes as a percentage of the sum of support and oppose votes, and ignore the rest. As such, it is misleading to have a neutral option under this system.

barbarian  18:19, 12 November 2009 (UTC)[reply
]

What I believe the difference to be is:
  • "neutral" means that the voter has no preference for either "support" or "oppose"
  • "abstain" means that for some reason you do not feel that you can vote on this issue - perhaps (a hypothetical) you know that it is your brother, and so you don't want to vote S, O or N, but 'abstain' from voting as you have a CoI?
At least, that is the general meaning of abstention - for example, in a Supreme Court situation, a judge who feels that they are too involved with a case will abstain, rather than vote, to avoid their CoI. -- PhantomSteve (Contact Me, My Contribs) 18:29, 12 November 2009 (UTC)[reply]
Just to clarify my last: IMO the only reason to have an "Abstain" button in a Wikipedia election is if the ballot interface doesn't allow a voter to clear their existing selection in any other obvious way. -
Rannpháirtí anaithnid (coṁrá)'s 20:55 comment below. - Pointillist (talk) 21:56, 12 November 2009 (UTC)[reply
]
In addition to the remarks above in favour of an "abstain" button, an "abstain" button is also a clear marker to the user that they don't have to cast a vote. One problem with conducting any business through a computer interface is that people are not immediately sure of what is "allowed". Without an "abstain" button people may presume that they have to fill in a vote for each candidate - or not feel certain that they have abstained without the definitive feeling of checking a button that says so. An "abstain" button could also come by way of a button that clears a set of radio buttons so long as it was clear to the user that they did not have to vote for all or any candidate. --rannṗáirtí anaiṫnid (
coṁrá) 20:55, 12 November 2009 (UTC)[reply
]
We could have an "Abstain/neutral" button at the top of every list, selected by default. Users would be required to vote in every poll, but that first option would count as zero. After subtracting the number of disapproval from the number of approvals, compare the results. (What happens if the net results are the same? Leave it to Jimbo, or do a run-off.) HereToHelp (talk to me) 21:05, 12 November 2009 (UTC)[reply]
Aren't we trying to find a form-based approach that is most similar to our current wiki-based ballots? For example, in last year's ArbCom Elections there were only three options for each candidate: Support, Oppose or (by implication) Abstain. You would be abstaining if you didn't vote for a candidate, or if—having voted—you subsequently struck out your vote. "Neutral" wasn't an option, and users were not "required to vote in every poll" as perhaps you were suggesting. If the community wants to offer the option of a protest vote, well, that can be implemented as a specific option in a ballot. But abstention should still have its usual meaning of "not voting", without anything being inferred from it. - Pointillist (talk) 22:21, 12 November 2009 (UTC)[reply]
Sure, a common argument for an abstention option (in real life) is to allow the voter to participate in the process but to note that they have not made a mistake or spoilt their ballot - simply not marking your ballot paper is normally considered a spoilt vote, equal to if you had made an error or purposefully defaced it. So long as people can participate fully but choose not to vote the outcome is the same. Their intentions for doing so can be as many as there are intentions for voting for or against a candidate: the voter may believe they have a conflict of interest, they may not feel as if they know enough, they may not have an opinion. The important thing is that they can participate, while still not actually voting, either in full or part (i.e. abstain from voting for a particular option or candidate).
This is very distinct from "netural". Which is an option that I believe can or cannot be on a ballot. "Abstain" is a must (particularly if there are multiple votes as part of the same ballot e.g. voting for/against several candidates). --rannṗáirtí anaiṫnid (
coṁrá) 23:01, 12 November 2009 (UTC)[reply
]
I agree that deliberately spoiling a ballot paper shouldn't mean the same as attending an election but not casting any votes, but in general politicians have been nervous about offering "A plague on both your houses" as a voting option. Wikipedia can do better: Secure Poll can display options for protest vote, "don't know" and even "hey, I turned up, did due diligence and still can't decide". Call me an existentialist if you will, but I just don't think any of those is the same as abstaining, and given that we currently allow editors to review/amend their votes during the ballot period, we have to allow some way for them to "undo" their vote(s) if that's what they want to do. "Neutral" is an entirely different matter: in the current RfA process some people find it a useful way to contribute comments about the candidate without explicitly voting, which wouldn't make sense if elections are secret. - Pointillist (talk) 23:30, 12 November 2009 (UTC)[reply]
My own, completely personal, interpretation of abstain - especially if clearly marked as the default on the ballot - is as a statement that "I don't know enough to vote in this instance". If clearly marked as default it also works as a visible reminder that you did not vote there - an improvement over real life, where you have only a few minutes to go over the ballot and you may have inadvertantly omitted some votes. (Ballot design is, um, difficult sometimes.) Those seem like good reasons for the abstain option. I have the impression that the abstain and neutral options in various votes and !votes have their own historical purposes here on WP, but am not likely to do enough research to get a good feeling for them. Neutral is widely used in RFAs and abstains so in Arbcom decisions, so I'm supposing they stem from some early WP discussions. Novickas (talk) 01:44, 13 November 2009 (UTC)[reply]
Can we agree that the lowest-common-denominator meaning of Abstain is just "I am not voting on this proposition/candidate"? The reason is probably personal and anyway I think it'll be difficult to agree a more subtle definition, given the wide variety of en.wikipedia editors' political systems and first languages. Pointillist (talk) 01:58, 13 November 2009 (UTC)[reply]

What's the point of all this? Neutral, abstain, and "no selection" are going to be counted exactly the same unless users are in for a nasty surprise. While there may be a semantic difference between neutrality and abstention -- and we've seen a few dozen different explanations of what that difference might be -- the computer is the only one that sees the votes, and it doesn't care about semantics. Do you want to explain this fine semantic difference, which just gets thrown out anyway, to every user wondering if they should vote "neutral" or "abstain"? rspεεr (talk) 01:50, 13 November 2009 (UTC)[reply]

Also, why would we want the ability to "publicly declare your non-involvement" on a ballot that isn't at all public? rspεεr (talk) 01:58, 13 November 2009 (UTC)[reply]

Rpseer, 1) I don't think it's decided whether the ballots will be published after the election ends; 2) even if the ballots are not published, the numbers of abstentions/neutrals/no-votes for any candidate will presumably be published, and they don't all mean the same thing. 69.228.171.150 (talk) 03:53, 13 November 2009 (UTC)[reply]

"Neutral, abstain, and 'no selection'" - "Abstain" and 'no selection' would be the counted same (i.e. not counted). "Neutral" could potentially be counted. In any case "Abstain"/'no selection' are semantically very different from "Neutral". It would be inappropriate to say, "If you're neutral don't vote" or "If you're neutral your vote doesn't count." The reasons why a person may abstain/make no selection are multiplicity. --rannṗáirtí anaiṫnid (
coṁrá) 09:06, 13 November 2009 (UTC)[reply
]
Add: if "neutral" is included as an option in any electino then should have an exact and useful meaning towards how the vote will be counted i.e. "Neutral" should be an option that is 'counted' in some way that can meaningfully influence the vote. "Abstain"/'no selection', on the other hand, should be included ni all votes with the meaning that that particular vote or part of a vote won't be counted. --rannṗáirtí anaiṫnid (
coṁrá) 09:09, 13 November 2009 (UTC)[reply
]
In which direction would you like "neutral" to influence the vote? Do you think people casting a neutral vote would expect it to meaningfully influence the vote? rspεεr (talk) 09:32, 13 November 2009 (UTC)[reply]
Typically when there is a requirement to have the support of more than half the votes cast. This, I'm afraid, is beyond the topic of this discussion as it's much more to do with the rules of the election that the mechanism of voting but it is meant to stop examples such as 35% Support / 35% Neutral / 30% Oppose victories. (NB Abstains don't count here which is why they are needed in the voting mechanics - which is what I thought we were discussing). Saga City (talk) 14:10, 13 November 2009 (UTC)[reply]
Percentage support has always been calculated from just the support and oppose votes before. The "neutral" vote you're describing here is not very neutral, because it helps to defeat a candidate. I certainly don't think there should be an option to place a vote that's called "neutral" but actually is slightly negative, because that's not what people expect it to mean. There should be one neutral option and it should have no effect on that election whatsoever. rspεεr (talk) 07:20, 14 November 2009 (UTC)[reply]
"In which direction would you like 'neutral' to influence the vote?" That depends on how the vote is to be counted (but as Saga City says, this kind of discussion is beyond the scope of this discussion). If "Neutral" is not a meaningful option then it should not be included. In much the same way, if the only meaningful choices in an vote are "Yes" or "No" then we should not include an option to vote "Maybe" (or "Pineapple" or "Dog" or any other random gesture that is meaningless in terms of the vote).
We should, however, always include the option to abstain (though that is obviously not a 'vote' and would have no influence over the outcome of the poll). --rannṗáirtí anaiṫnid (
coṁrá) 18:23, 13 November 2009 (UTC)[reply
]
It actually sounds like we agree now that there are only three options that make sense, and two of them are called "support" and "oppose". It's just that I'm calling the third option "neutral" while you're calling it "abstain". The word we use for it doesn't really matter. rspεεr (talk) 07:22, 14 November 2009 (UTC)[reply]
Ah, I getcha. For me "Neutral" is an actual 'vote'. A real option that counts in some way. 'Abstain' means, 'I'm not voting' for one reason or another. I prefer the word 'Abstain' because the reason for abstaining may be because the person is not 'Neutral' or not in a position to decide if they are. --rannṗáirtí anaiṫnid (
coṁrá) 10:39, 14 November 2009 (UTC)[reply
]
I don't know whether this helps, but for what it's worth here are the online OED (US view) definitions of neutral and abstain. I've also been trying to imagine how this would work for voters who have English as a second language. If they look up "abstain" in an English dictionary the voting meaning is clear. But if they consult an English-to-Mother-tongue dictionary, there's no guarantee that the "choose not to vote" meaning will be covered. Is that a problem? - Pointillist (talk) 14:20, 14 November 2009 (UTC)[reply]

"The word we use for it doesn't really matter." Yes, it does, since the two words mean entirely different things. Neutral is a vote in the middle between oppose or support, and is an actual vote with either support or oppose being acceptable as an end result to that voter. To abstain is, in the words of the COED, "formally choose not to vote". It's a rejection of the other choices and is a selection that says: Don't count this vote. Neutral != Abstain under any stretch of the definitions or usage. So no, we do not agree that there are three options. There are four: S/N/O/A. — Becksguy (talk) 22:26, 14 November 2009 (UTC)[reply]

There are other options as well. Recuse is slightly different from abstain. My interpretation of abstain is that one doesn't vote, but still takes part in discussions on the subject. Sometimes the proposer of a motion abstains because they wish it to pass without their formal vote, but as proposer they would still explain the motion and take part in discussions. A formal recuse goes further, as it (in theory) removes the recused person entirely from the discussion and any attempt to influence the outcome. Other options are "reopen nominations" and "none of the above", but those don't really count here, but would in different voting systems. Carcharoth (talk) 19:30, 15 November 2009 (UTC)[reply]
I can see the difference, and I accept that in the pre-SecurePoll world it was possible to make finely-nuanced statements like Comment I recuse due to conflict of interest, which would be visible during the ballot. When votes are secret recusing would need to be implemented as a way of publicly removing yourself from the electorate, so that you cannot vote – or so that your vote(s) are ignored. This might be implemented automatically (similar to the way that other potential voters might be disenfranchised by e.g. lack of mainspace contributions) or it might be handled manually by overseers. Either way, the overseers would confirm that the announced results exclude votes by recused or otherwise disenfranchised voters. Does that work for you? - Pointillist (talk) 22:15, 15 November 2009 (UTC)[reply]
Having just spent a day at jury selection, there are only two votes that count: FOR or AGAINST. No other types of votes (or complete lack thereof) change the results. That we have jumped down an enormous rat hole debating, among other finer points, what the potential nuanced meanings of "abstain" might indicate is proof enough that we should just fill the rat hole and move on. Two buttons, FOR and AGAINST. Three including "NO VOTE" as a control to confirm someone didn't accidentally skip marking FOR or AGAINST a candidate, although if the voter is not immediately prompted to correct upon submit, there's no point. (I have not used SecurePoll and so can't comment.)  PЄTЄRS VЄСRUМВА ►talk  19:56, 17 November 2009 (UTC)[reply]

Options for ensuring accuracy in tallying

In the two RFCs, there have been a number of voices objecting to SecurePoll due to (perceived or real) unaccountability or uncertainty in the poll results. What are all the options available that address this issue? As a convenience, this section attempts to catalog all the options presented so far. -kotra (talk) 02:02, 14 November 2009 (UTC)[reply]

  1. Standard SecurePoll setup: overseer(s) bear sole responsibility to not intentionally or unintentionally get the results wrong.
  2. Wikipedia:Secret ballot process, a number of proposed technical measures, including unique voter IDs, intended to prevent fraud by overseers and voters alike.
  3. "New compromise solution", which offers standard SecurePoll secrecy until the election is over, when all votes are made public.
As the writer if option 2, I would far prefer option 3. Transparency is not only about "trust" or "confidence" (which is a better word), it is also about legitimacy. Casting a ballot in secret is a good idea - it stop the "pile on" effect or the undue influence of comments when one is casting their vote - but I don't that that there is any real suggestion of post-vote intimidation or reprisals (TBH if anyone would use an ArbCom seat to exact reprisals then they should not be on the committee). Opening the list of ballot afterwards would be a natural way of demonstrating transparency and legitimacy in the poll and securing confidence in the process. It would, I think, do a great distance to satisfying both "sides" of the open vs. closed debate. --rannṗáirtí anaiṫnid (
coṁrá) 10:36, 14 November 2009 (UTC)[reply
]
I disagree, i.e., I prefer #2, not #3. In areas of contention, editors will seek to see how certain perceived "blocs" of editors voted. That will lead to "the friend of my friend is my friend" and "the friend of my enemy is my enemy." There is nothing to be gained by revealing who voted for whom: only presumptions of alliances, for or against, which will only taint the elected and hamper their ability to be effective.  PЄTЄRS VЄСRUМВА ►talk  20:03, 17 November 2009 (UTC)[reply]

Workflows, screenshots, test election etc.

It would be very helpful if someone could post a workflow (and preferable screenshots) of the user experience of voting using SecurePoll. A series of user scenarios would also be helpful. Most people I presume are like myself and have only a one or two experiences of using SecurePoll. And when we did, we didn't do so with a user review of the system in mind.

Are there any such example out there? Or additionally, would it be possible to set up a test election in the system so that we can return to it and give a proper review. --rannṗáirtí anaiṫnid (

coṁrá) 21:00, 12 November 2009 (UTC)[reply
]

Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Arbitration_Committee/Audit_Subcommittee/October_2009_election/SecurePoll_feedback_and_workshop&oldid=1081884220"