If an answer is downvoted, the person who wrote it gets only -2 rep, while an upvote gives +10 rep, which means that in order to negate the upvotes and encorage the poster to delete/edit their answer we need 5 downvotes for every upvotes, this creates a situation where answers that get alot of downvotes are still "profitable" for their posters, here are some examples:

Is Barack Obama a natural born US citizen? an answer with +4 and -21, the most downvoted answer in the site, and the answer with the lowest score on the site, gives its writer only -2 rep.

Do Israeli soldiers kill Palestinian kids? an anwer with +5 and -15, the second most downvoted answer in the site gives the writer +20 rep

Is CO₂ the cause for Global Warming? answer with +2 -12, gives the writer only -4 rep

Is there anything inexplicable about how the WTC Twin Towers collapsed? an nswer with +3 and -12 gives the write +6 rep

Was McCarthy directly involved with the House Un-American Activities Committee (HUAC)? an answer with +3 and -10 gives the write +20 rep.

This is a ridiculous situation in my view, beyond the fact that answers which are clearly shown by the community they are no good give their writers rep, which suppose to indicate trust.


After reading Should the weight of downvotes be increased? I tend to agree with the reasons for not increasing the penalty for all downvotes, but I still think that the examples I've brought and some others are ridiculous. So I have two propositions for this:

  • increase the penalty differential, where from a certain amount of downvotes each downvotes reduces more rep, this addresses both issues raised by Jeff.

    a. It'll not change the cost for a downvote, so no "nerfing" the tool.

    b. This way it won't be used to bludgeon a user, unless a lot of people agree that an answer should be bludgeoned, and will not support indevidual retaliation against people.

  • Since we are (still) a fairly small site, we have only 1 user that is not a mod and can nominate answers for deletion, so we can't really trust on the community to do its own, because in this matter there is no community yet. What I suggest is that answer that get enough downvotes would be offered for closure to the community as a question in meta, and if enough people support it (ether by answers, or by votes) then the modes will delete them with their godly powers.


I myself have answers with few downvotes, so If such a measure is taken I might loose some rep. (Which I'm OK with).

2nd Disclaimer:

The new proposed methods will probably not affect me, but even if they do, I probably deserve it.

  • I agree with this, especially since voting down an answer COSTS the user 1 rep. The effect of a down vote should at least be doubled. If it's doubled, the ratio of Rep gained/lost per vote is same for questions and answers.
    – Wertilq
    Commented May 27, 2013 at 18:01
  • This was already proposed here, and declined.
    – user5582
    Commented May 27, 2013 at 19:12
  • @Sancho, didn't find it before, and after reading Jeff's reasoning I understand why they didn't implement it. But I'm still in favor of the differential cost, where the more downvotes you have the more each costs. Because it's ridiculous in my mind the the second most downvoted answer gives the writer +20 rep.
    – SIMEL
    Commented May 27, 2013 at 19:26
  • @Sancho, how about now?
    – SIMEL
    Commented May 27, 2013 at 19:40
  • Your question seems different enough, and I don't really have an opinion about rep issues. Just wanted to point out the previous discussion.
    – user5582
    Commented May 27, 2013 at 20:11
  • My issue is less with the reputation, it's more about low rated answers left on the site despite being heavily downvoted.
    – SIMEL
    Commented May 27, 2013 at 20:13
  • Ah, perhaps you should ask a question about the deletion policy. I believe answers with very low scores appear greyed out to non-signed-in visitors.
    – user5582
    Commented May 27, 2013 at 21:21
  • they appear grayed out to all
    – SIMEL
    Commented May 27, 2013 at 22:29
  • 1
    looking through MSO, there is a very old post about this issue labeled status declined, leading me to believe that this is by design that down votes are relatively minor.
    – Ryathal
    Commented May 28, 2013 at 13:28
  • @IlyaMelamed I've added a separate question addressing the expansion of the deletion policy: meta.skeptics.stackexchange.com/questions/2383/…
    – user5582
    Commented May 28, 2013 at 17:33

3 Answers 3


I am limiting my discussion to answers posted on the main site. Answers posted on Meta, and questions on the main site having different meanings for downvotes.

I don't have a strong opinion on this matter, but there are a few caveats that should be considered when evaluating the proposal above.

Punitive to OP versus Warning to Others

There is an assumption that the purpose of downvotes is to punish the poster.

This is touched on by the linked Meta Stackoverflow question:

When we started Stack Overflow, we wanted to make sure that downvotes were more of a visual and psychological motivator than a punitive action.

I guess I see the more important role of downvotes as an indicator of distrust of that answer, rather than distrust of that user.

(This raises another issue: +10/-5 answers are probably less trustworthy than +5/-0 answers, but they appear the same to visitors.)

Deletion of Unacceptable Answers

The answers you cite are distrusted by the community, but they follow community standards. Most of the answers that are downvoted heavily are because they don't follow community standards, and are deleted by mods (and to a lesser extent, by high rep users.)

They tend to be deleted before they get over a dozen downvotes. If punitive action was important, we should leave them around longer to be further punished. However, it is better they are removed as broken windows and improve the sites signal-to-noise ratio.

Biased Sample

Which brings me to my main point: the sample of the highly downvoted answers you have cited is unintentionally biased. These aren't the worst answers; they are the "worst" answers that do manage to follow the community rules. The worst answers have already been deleted.

  • Parts of what you say is completely true, but you doesn't fully address the problem. The worst answers are deleted, true, but if an answer is really bad, but follows all the rules, it's allowed to stay, and no one takes care of it.
    – Wertilq
    Commented May 28, 2013 at 14:26
  • @Wertilq That is an issue with the deletion policy, not reputation penalties.
    – user5582
    Commented May 28, 2013 at 16:27

Anyone with at least 15 rep can "nominate" a post for deletion by flagging it. Mods won't necessarily oblige, but for truly abysmal answers it's a fair bet they'll at least take a hard look. Flags become part of an author's permanent record on the site - too many, and moderators may take action beyond just deleting bad posts.

Anyone with at least 50 rep can leave (constructively) critical comments. Even if that doesn't result in improvement, it does provide a public record of the problems with the post.

Anyone at all can leave a better answer, demonstrating the problems with a bad answer by way of comparison. A bad answer can be valuable to the readers by demonstrating faulty reasoning if that is made clearly evident by way of down-votes, comments, and other answers.

Finally, the problem of earning net reputation from controversial answers is an old one, and a big part of why certain types of questions were discouraged and eventually disallowed on other sites. Given the subject matter, it is not possible to disallow questions that spark controversy here - hence the emphasis on very strict moderation (of both questions and answers). There's no sure-fire way around this; even weighting the penalty associated with downvotes might result in a net gain if the ordering of votes is favorable to the author, and doing so would dramatically increase the complexity of the (already very complex) reputation system. Over the years, I've come to believe that it's simply not worth fixating on; it is enough if the posts themselves are scored appropriately, worrying too much about edge-cases involving manipulative authors becomes a distraction with very little payoff.


No, we shouldn't change the reputation penalty. I agree with Jeff Atwood's answer here here.

  • Reputation doesn't matter, but some people think it does.
  • What matters to visitors is the quality of the answers, which is demonstrated by their net score.
  • In order to have this net score be meaningful, downvotes should cast as freely as possible. There should a very minimal penalty for both the caster and the recipient — enough to be noticed, but not enough to create a psychological barrier to downvoting.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .