tea.mathoverflow.net - Discussion Feed (Signed voting)

Harry Gindi comments on "Signed voting" (3711)

2010-03-07T02:26:24-08:00

The person who has received downvotes does not have any information to "make this post better" unless somebody leaves a comment. Maybe an answer is to force commenting on downvotes but allow this to be anonymous. Anything that will encourage comments with votes down will suffice.

Ilya Nikokoshev comments on "Signed voting" (3709)

2010-03-07T02:18:23-08:00

I can't think of any legitimate reasons why you would want to downvote anonymously. Can you think of any?

Because if somebody upvotes or downvotes the post anonymously, the only way for a post owner to react is through his post, that is, s/he can say "Ok, I have receibed some upvotes and downvotes, and I think I have to make this post better".

If the post owner also has the information about who upvotes and who downvotes, this opens the possibility of going personal, starting with "How could you slap me in the face with downvote, I thought we were friends?", progressing to "Since you voted me down, I'm going to vote you down on the faculty meeting so you won't get tenure", or even "Thanks for your upvote, I'm going to upvote all of your posts now and send you this $100 gift certificate". I think you see the direction where this could be going, and I don't think it's something that should be encouraged.

Andrea comments on "Signed voting" (3702)

2010-03-06T10:37:34-08:00

I should also add that downvoting costs reputation, so I think most people use it only to mark "bad" questions, rather than as a general opinion.

Harry Gindi comments on "Signed voting" (3664)

2010-03-04T23:00:42-08:00

I don't think it's at all clear what a downvote should mean, and it doesn't seem like the moderators even all agree. Is there any way you could give voting suggestions of some sort on the FAQ? Maybe a section like "what does a downvote mean", or something along those lines.

Scott Morrison comments on "Signed voting" (3663)

2010-03-04T22:53:09-08:00

@fpqc,

re: "We both know that is not how voting down is actually used. If it were, we would see that most people had as many downvotes as upvotes. This is simply not the case. There's a saying that I find particularly appropriate, "Don't spit on me and tell me it's raining"."

Actually, you're wrong. I have given lots of downvotes, many of which I haven't left comments on, and nearly all of them fit what I said -- I'm just indicating to other readers that I don't think they should spend their time on this question. It is very much how I use downvotes. I understand you have an unusual perspective on the matter, as you've received spiteful downvotes, but please don't mistake what you've seen for what is actually happening on a larger scale.

Harry Gindi comments on "Signed voting" (3662)

2010-03-04T22:07:26-08:00

Almost every downvote I have ever received has not had an accompanying comment (nor a vote for somebody else's relevant comment). If you take issue with a post and want to vote it down (and nobody else has communicated your criticism), you should leave a comment. I don't see why people would object to this. I've always been of the opinion that votes down should be reserved for bad posts and votes up should be reserved for interesting posts. If you don't find a post interesting, then don't bother voting at all. As I noted above, I don't find most applied mathematics questions particularly interesting (not to say that there aren't any). However, I'm sure that most people here would agree that targeting applied math posts is not "in the spirit of the site".

Since there is no way to encourage this behavior, I think that publicly visible voting would at least allow the community to encourage and discourage certain approaches to voting. If we want people on MO to act more maturely than members of an average internet forum, I think that the cession of some amount of anonymity is required.

Ilya Grigoriev comments on "Signed voting" (3661)

2010-03-04T21:55:07-08:00

@fpqc from two posts up: One thing jumped at me: If you "either know the answer or could find it out with very little effort" to a question that is not a homework question, wouldn't it be nicer to just answer it?

But also (not to sidetrack the discussion), I think Scott is right about most downvotes not being malicious in any sense (and the reasons why sometimes people don't leave comments), and you are mostly wrong. I understand why it could be annoying if somebody downvotes a lot of your posts, but I think your best bet is just to ignore it. Even if there still are people systematically downvoting your posts (which I doubt, by now, but you can ask Anton), they'll get bored and go away.

As I said above, it might be good to make it harder/costlier to downvote in general (so that people have even less incentive to downvote many posts at the same time), and the system should certainly try to search for suspicious voting patters, but I think in general most people seem happy with this situation as it is.

Harry Gindi comments on "Signed voting" (3660)

2010-03-04T19:32:32-08:00

If voting were public, I would gladly use my real name.

With regard to your last paragraph, I would just like to note that the majority of my non-community-wiki downvotes are signed with an explanation or a vote up on a comment that offers an explanation (this is as good as signing as well). I was admitting to falling prey to that temptation, as I'm sure many people have.

Anyway, my original proposal was not to make all voting public. I'll summarize it in one line:

People who are willing to hold themselves to a higher standard with voting should be able to expect it in return.

Douglas Zare comments on "Signed voting" (3659)

2010-03-04T19:12:04-08:00

How can you talk about how voting should be signed when you are trying to remove references to your real name? How can you say, "If I'm going to vote someone down, I'm willing to put my name on it and own it?"

I choose to provide my real name here, but I can see reasons why people don't, and I have been anonymous elsewhere. I can also see reasons why honest people vote more often and more informatively for others when they don't have to sign their votes, up or down.

A negative vote with no explanation is going to be interpreted as much more objectionable by many people such as myself. This contradicts past suggestions from some moderators. Maybe they should rethink their position, or at least not treat it as obvious (when they even contradict each other). However, sometimes it's ok to take an action which the author finds objectionable, and sometimes someone else has already written a response.

By the way, excessive grammatical or spelling errors from a native speaker who should know better, or horrendous/misleading notation, shows a lack of respect for the potential readers. Some people argue that the internet is just too fast for punctuation or capitalization, but this means they waste the time and energy of 50 readers, in contrast with 1 person who might read a text message. It's reasonable for that to be a contributing factor in a vote down or up, or lack thereof. However, even if I agreed with everything on your list, I would still not want every vote to be open.

I'm surprised that you say all of your unsigned downvotes are "totally out of spite." None of mine are. Maybe you should work on your own behavior before imposing rules on others. (If people give you lots of etiquette suggestions, that doesn't make you an expert. It suggests the opposite.) There is a balance between preventing abuses by malicious users and making the system work well for those who are trying to help others. Even if your focus is (incorrectly) on malicious users alone, think how much more fun a troll has if everyone has to sign votes.

Harry Gindi comments on "Signed voting" (3658)

2010-03-04T17:12:00-08:00

Certainly, there are good and bad reasons to vote a topic up and down. Keeping voting public gives more of an incentive to vote for "good reasons".

What are good reasons for voting a topic down?

The question does not make sense.
The question is too elementary
- You are familiar enough with the subject to make this judgement
- You either know the answer or could find it out with very little effort
The question is not well-researched
The question is written like a homework question
The question provides insufficient background.
The question is wrong on definitions and technical points.
Of course there are more.

Bad reasons to vote down:

You dislike the questioner.
You think that the question may be too easy, but you have no idea how to solve it.
The question is related to a field that you actively dislike.
Grammar
Spite
Typographical errors
You don't personally find the question interesting

Many of the above could be useful when deciding not to vote a post up, but are not sufficient to vote a post down. If you agree with this list, then certainly any valid reason to vote down is also a reason to leave a comment.

@Anton: If I voted down every topic on applied mathematics because I categorically don't like applied math, is that okay? I would say no, but at least using your criterion, I would be totally justified in doing so.

Anton Geraschenko comments on "Signed voting" (3657)

2010-03-04T17:10:09-08:00

I think that is how voting is typically used. It makes sense that people are more on the lookout for things they like than things they don't like, so I expect people cast more upvotes than downvotes. Aside from asking themselves, "do I like/dislike this enough to vote on it?" (and effectively "are other people likely to want to read this?") how do you think people vote?

Harry Gindi comments on "Signed voting" (3656)

2010-03-04T16:44:09-08:00

We both know that is not how voting down is actually used. If it were, we would see that most people had as many downvotes as upvotes. This is simply not the case. There's a saying that I find particularly appropriate, "Don't spit on me and tell me it's raining".

Scott Morrison comments on "Signed voting" (3653)

2010-03-04T16:08:03-08:00

A slightly different characterisation of what a downvote might mean: "You probably don't want to bother reading this post, but I don't have the energy to explain why."

It's obviously suboptimal, but in an attention-scarce world, I think it's still useful.

Scott Morrison comments on "Signed voting" (3652)

2010-03-04T16:07:02-08:00

@fpqc, I really strongly disagree with your characterisation of a downvote without a comment. I think it is much more useful to interpret it not as a message to the poster at all, but to the community at large, as an inexpensive and not necessarily particular serious indicator that the post is less worth reading than others.

I have no intention of reading all posts (although I'll admit to an rss habit which means I do skim every question) in detail, and voting for me is very helpful -- the combination of an alluring title and the indication via voting of community interest is helpful in deciding where to spend my time. If user A downvotes user Bs post, then that is useful to me (neither A or B) in much the same way an upvote is, and this is the case regardless of whether there was a corresponding comment.

Now I'm not saying that you're totally wrong in your thinking about downvotes, just that I'd be happier if you also acknowledged this reason to like them!

Harry Gindi comments on "Signed voting" (3640)

2010-03-03T23:31:36-08:00

What I meant by "a slap in the face" was precisely a vote down without a comment. I'd even be happy if there were a way to leave comments anonymously, as long as they were mandatory. What kills me is receiving downvotes with no indication that anything is wrong. I'm sure that other people feel the same way.

A vote down with no indication that something is wrong is equivalent to saying, "I don't like your post, but you don't deserve to hear why."

Ilya Grigoriev comments on "Signed voting" (3639)

2010-03-03T23:27:07-08:00

I think one thing that could be good is making it hard to make many downvotes. For example, the system could enforce an at most 1:10 ratio of down/up votes.

Anton Geraschenko comments on "Signed voting" (3637)

2010-03-03T23:18:01-08:00

Like I said, I can see where you're coming from, but I remain unconvinced. I think the problem is that you're interpreting a downvote as a slap in the face; it's not, and (I believe) a large part of the reason that people don't want to identify themselves when voting on your posts is that you think it is. I think there are legitimate reasons to vote anonymously, even on things that don't matter, and I think that a purely signed voting system would result in people bickering and angsting over things that really don't matter. I can kind of see the argument for an opt-in system, but that's pretty much what we have now since you can leave a comment.

"I would be interested in seeing what fraction of votes (down) cast are accompanied by a comment or a vote up on a comment." If I get the chance, I'll see if I can easily extract this from the (non-public) database dump.

Harry Gindi comments on "Signed voting" (3635)

2010-03-03T23:05:59-08:00

I don't know that said people have voted down my posts for legitimate reasons, mainly because they never left comments.

The difference between anonymous voting in politics and anonymous voting on a website is that in politics, each person has one vote about things that really matter. People don't have incentive to vote for a candidate out of spite given the amount that is at stake. On a website like MO, every person has 20 or 25 votes that are replenished daily to vote on things that are not important.

Concerning the honesty of the vote: I am willing to vote honestly and openly. I think that my request to be dealt with in kind is completely legitimate, but I do not see how that will ever happen with so many votes happening "in the dark", so to speak. This is why I proposed it as an opt-in type of thing.

There's no reason why you should be able to slap a person in the face without taking responsibility, at least if we want MO to be a professional site. If we want anonymity and nonsense, we can go anywhere else on the internet.

Anton Geraschenko comments on "Signed voting" (3633)

2010-03-03T22:30:56-08:00

Yes. Here's one example. Suppose I don't think a question is very good. It's poorly constructed and poorly thought-out, but say it got three votes. Let's say I can think of a pretty good answer. So I answer the question, but I don't vote it up, and I don't have the time or energy to explain why it's not a very good question. It would be (potentially) annoying to me for the asker to know that I didn't vote up his question. I'd rather he not have that information, so I'd like upvotes to be anonymous.

Another example: I know (because of email exchanges) that some people have downvoted your (fpqc's) posts for legitimate reasons, but not left a comment because they thought (on good evidence) that you would react negatively and/or hold a grudge. I think this is a legitimate use of anonymity in voting.

Basically, I think we should have the option of anonymous voting for the same reason we have it in politics: people vote differently (and less honestly) if everybody knows how they voted.

Harry Gindi comments on "Signed voting" (3631)

2010-03-03T21:47:53-08:00

I think the question I'm asking is: why would you want to vote anonymously? Every time I've voted down anonymously, it's been totally out of spite. Conversely (contrapositively?), I can't think of a single upvote I've made that I would want to keep secret.

I can't think of any legitimate reasons why you would want to downvote anonymously. Can you think of any?

Edit: I would be interested in seeing what fraction of votes (down) cast are accompanied by a comment or a vote up on a comment.

Anton Geraschenko comments on "Signed voting" (3629)

2010-03-03T21:30:59-08:00

People who are willing to own their votes are rewarded, as we discussed in the vote campaigning thread. They are rewarded by other people agreeing with them. If you think something is wrong with a post and you leave a comment explaining your downvote, you get benefits that just downvoting doesn't provide:

Other people may also downvote (or at least refrain from upvoting) based on your comment, so your vote is effectively heavier.
Other people will apply your reasoning to other posts when deciding how to vote, so you end up on a site with "better citizens".
Other people (including the original poster) will react to your critique. If it is valid, the original poster is likely to fix up the post or at least fix up future posts.

All the same benefits come with commenting on upvotes.

That said, I can understand where you're coming from. It does kinda ruin your day when somebody downvotes you and doesn't give you any indication of why. To some extent, this frustration is the cost of "living in a free society". I think any kind of enforced signed voting would break the system more than it would fix it.

Harry Gindi comments on "Signed voting" (3620)

2010-03-03T20:43:38-08:00

The whole point is that people who are willing to own their votes should be rewarded. People who want to keep their votes secret would then be warned when voting on such a user's posts that "Votes on this user's posts will be visible to this user".

In cases where there is real criticism, the people here tend to leave comments. The only time that voters would want their votes kept secret is if they aren't voting based on substance. It takes a while to write a decent question, but it takes just a second to "ruin someone's day" by voting down anonymously without comment.

I agree with your statement on votes up. Voting up should be subject to the same rules.

Ben Webster comments on "Signed voting" (3617)

2010-03-03T20:31:31-08:00

fpqc-

I think I would suggest such a system being opt-in on the part of the voter; it's a little ambiguous from your post exactly which opt-in points you're envisioning. If that were the case, the only objection I could see would be technical (I will note that whatever discussion we have here is almost surely moot for quite a while, but still worth having so we can make requests on meta.SE).

Of course, some times we just have to accept that other people on the site consider things that annoy us to be perfectly reasonable. There are plenty of questions I really wish people wouldn't vote up, but they do, and I'm not going to try to stop them. Though let me be clear, I'm not condoning malicious systematic downvoting (which I know has happened to fpqc); that's not acceptable behavior, but that's separate from an isolated downvote without a comment.

Harry Gindi comments on "Signed voting" (3612)

2010-03-03T19:56:46-08:00

There's no way for the community to frown on something that's completely invisible. The current system seems to be designed to invite this kind of behavior.

Qiaochu Yuan comments on "Signed voting" (3611)

2010-03-03T19:52:56-08:00

The current system isn't ideal, but I think signed voting would on the whole invite too much bickering. The solution to the problem of people not commenting when they downvote should come from a change in norms, not a change in the software.

Harry Gindi comments on "Signed voting" (3610)

2010-03-03T19:39:02-08:00

I don't see how anonymous voting is helpful or useful on a site like MO. If I'm going to vote someone down, I am willing to put my name on it and own it. However, some people are not willing to reveal their identities in this way.

I propose an option called "signed voting", which signs a user's name on a list of votes visible to the questioner, but in return, that user can see who has voted on his or her question. The way the current system works, there is no incentive to be courteous and post a comment of any sort. I know that this is controversial, so I'd be interested in hearing counterarguments.