This site is probably in the most unique position to examine how belligerent people get when discussing topics on which there is much more room to disagree than there is room to agree. People's political leanings often (maybe even usually) become intransigent by the time they become eloquent enough to state their views in a clear manner.

There is any number of meta posts which complain about lack of civility here, about blatant violation of procedures aimed at elevating inquiry over patchiness and so on. The answer that often comes from the mods is that the community has decided this way by its voting.

What this clearly ignores is that votes themselves act as lash out and re-enforcement mechanism. If people voted to inform others of the quality of posts, these problems would not be so pronounced. Instead people often vote to boost their damaged egos or to bruise the egos of those they see as their ideological enemies.

In order to lessen the ego-boosting/ego-bruising impact of down-voting, I would like to ask that it be considered that the reputation loss for downvoting be 3 points instead of 1 point. This doesn't have to be permanent, of course. It can be an experiment for a limited time (start for 2 weeks for example). After that time passes, either the community or the mods can evaluate the impact that this change will have had and either roll it back, make other experimental adjustments or make the change permanent.

In order to stop the same effect on votes to close, I would also propose that a close vote (which is current free and effective way to act punitively towards those with whom one disagrees ideologically) carry some reputation penalty. At the moment, the close votes simply don't act the way they are supposed to. They are not exercised against posts which are poor quality. They are primarily used to censor. There are occasions when poor-quality posts appear here. But the number of posts which attract close votes are many multiples of the posts which are genuinely poor quality.

A close vote should be costly. Maybe their costs should even increase progressively for each individual until some sort of "reset" happens. For example, 5 points for the 1st close of the day and an arithmetically increasing cost for each consecutive close vote (8,11,14,etc.) resetting back to 5 every 2-3 days.

A more controversial measure would be to apply retroactively (and I don't even know if that's possible with the tools available to mods) rep decreases for downvotes, but, at the same time, allowing people to withdraw their votes at any time. Currently votes are locked after a fixed period of time, so it would not be fair to reduce people's reps without allowing them to rethink their votes to figure out if they really meant it.

There is already a system in place to reduce impulsive punitive behavior. The number of flags that any person can raise is limited. But this measure is not effective in a high-controversy low-population community such as this one.

  • Downvoters, please, leave a comment suggesting how this question can be improved.
    – grovkin
    Jul 28, 2018 at 0:16
  • 5
    On meta, people don't always vote on the quality of a question, but also downvote to show disapproval of the idea (this is different from main, where it's about quality). I think you could focus and reformat (headlines, etc) your question to improve it, and also show why you think this is necessary; the punitive behavior you see might not actually exist, or be clear to everybody. But ultimately, I downvoted because this is 1) not something politics.SE can do 2) Not something it should do. It would encourage the exact opposite behavior that we want to encourage (bad questions and bad answers).
    – tim
    Jul 28, 2018 at 8:04

4 Answers 4


First, What you're asking for changes to things that are the same throughout SE. Meta.StackExchange.com would be a better place to propose such changes. The mods here can't adjust how many points you lose when you downvote.

Second We don't really have a problem here with downvoting too much. If anything, the most common voting-related result of bias that we have is people not downvoting when they should be.

SE Uses gamification to encourage people to post higher quality content. We want the voting system to be actively encouraging it all the time. We don't want it to become a system where downvotes are thought of as so heavy handed that people don't use them.

And if people are downvoting your answers, it usually means that your answers failed in some way. Either your answer is not correct, in which case it deserves downvotes, or it is correct, but doesn't sufficiently justify itself, in which case it still deserves downvotes.

And for the record, you can represent the right wing here and still get upvotes.

  • 3
    First, you can represent the right wing in the answers, but not so much in the questions though. When you get 8 downvotes and no comments as to why, despite the usual "downvoters please comment" request, it's hard to attribute it to anything but bias.And that wasn't even technically right-wing. It was completely down-the-middle request for info... without any interpretations.And second, there is a main difference from the other sites on this one.This one is political by name.And everyone has political opinions that trigger them.The problem is high-controversy + low population.
    – grovkin
    Jul 27, 2018 at 23:38
  • 1
    It's informative, of course, to know that mod of individual communities cannot adjust how much effect voting has on the voters' reps. I didn't know that and I thought it may be different since comment upvotes used to give rep on other sites (not sure if they still do). So I just assumed that these are tweakable settings.
    – grovkin
    Jul 27, 2018 at 23:40
  • I am a little confused now. It looks like receiving downvotes on Meta.SE does cause rep loss. Receiving downvotes on Meta.Politics.SE does not cause any rep loss. Is it that the mods cannot control how much rep is lost/gained, but can control whether it is lost or gained? Or what's going on?
    – grovkin
    Jul 28, 2018 at 0:55
  • 5
    @grovkin: Meta.SE is a meta site in terms of how it gets used, but it is ultimately a distinct SE site, like Politics.SE, Writers.SE, and so forth. It therefore has its own reputation, distinct from any other site, which can be acquired and lost, etc. Jul 28, 2018 at 5:45
  • Re "Either your answer is not correct ... or ...": this leaves out the common possibility that a correct and sufficiently justified answer is downvoted by voters who prefer wrong answers. As with pending residents of a gulag, etc.
    – agc
    Jul 29, 2018 at 4:19
  • 1
    @grovekin Have you considered the possibility that your question (not sure which one you are referring to exactly) might have been ill-received not because of political reasons but simply because people thought that it was not a good question?
    – Philipp Mod
    Jul 31, 2018 at 8:35

I don't think that this would be a good idea.

My impression is that most of the problem users on this website do not really care about their reputation score. They care about spreading their ideology. So increasing the penalty for moderating content would not be much of a deterrent for them.

It would, however, be a deterrent for the constructive contributors who close and downvote content because it is not constructive.

So discouraging closevotes and downvotes might in fact encourage the trolls to post more ideological content, because fewer downvotes and closevotes means they gain a higher visibility and are less likely to trigger auto-blocks.

Bottom line: If you don't want to suffer from ideologically motivated votes, do not post ideologically motivated content.

  • Well, I don't want to shoot down an opinion. But I would like to point out that I proposed an experiment rather than a conclusion that this experiment mus reach. Although I will concede that my hypothesis as to what conclusion it will likely reach is different from yours. However, since adjusting weights is quantifiable and behavior is somewhat quantifiable, the results of such (temporary) adjustment is more or less measurable.
    – grovkin
    Jul 28, 2018 at 16:28
  • 2
    BTW, the "bottom line" falls under the "if you don't agree, you must shut up; if you don't shut up, you must be destroyed."
    – grovkin
    Jul 28, 2018 at 16:39
  • 2
    Indeed the "bottom line" reads like ad usum delphini. Somewhat idealised and incongruent. 1. you cannot really avoid that; troll-motivated voters (ideologists?) will do it anyway. 2. At least the point "might be read as ideologically motivated" has to be used; genuine posts with good intentions are sometimes ill-phrased, from non-native speakers, from quite a different culture, newbies etc etc; some voters want to read everything thru their ill-fitted glasses. / Currently it's phrased as if the poster has the most control of voting behaviour. That cannot be true? Jul 28, 2018 at 18:01

Sorry, that is not a usable solution (leaving aside a separate questions of whether the problem you stated was diagnosed correctly in the first place, OR, separately, even if it was, whether the proposed solution would address the problem).

  1. It's trivially easy to earn rep on Politics. Combine political biases and bike-shedding, and posting a popular polarizing post almost guarantees a couple dozen rep points. Never mind lucking out into posting HNQ question or answer (just put the T*(&9 label on the question). As such, mere 3 points per downvote isn't going to deter many people, imho.

  2. The usual main problem comes not from abundance of downvotes on good posts (it does happen but not as often), but from overabundance of upvotes on crap. And those are impossible to remove due to flood of SO voters coming in via HNQ with their shiny 100 rep association bonus, but without a single care for post quality - the problem is not unique in any way to Politics.SE but people are even more likely to upvote crap here as it hits their political buttons.

  • Is this your gut feeling or do you have access to data which shows it? I am not trying to dismiss your gut feeling, but I am curious to know what data is available that I haven't seen.
    – grovkin
    Jul 30, 2018 at 18:38
  • @grovkin - observation based on reading nearly every single question on the site over its existence.
    – user4012
    Jul 30, 2018 at 20:04
  • so, an informed opinion, then.
    – grovkin
    Jul 30, 2018 at 22:01
  • @grovkin - it's not opinion. It's a conclusion based on facts. You may call it a rough Mark 1 eyeball estimate.
    – user4012
    Jul 30, 2018 at 23:32
  • Re "easy to earn rep": Not everyone wants to "earn" rep by gaming the system. Though perhaps that "trivially easy" method might eventually be made searchable, and could be made useful for identifying gamers fond of that strategem.
    – agc
    Aug 11, 2018 at 17:52

This seems like a pretty good idea. It's clear that Politics.SE is not quite like other SE communities, and vice versa -- there's no obvious reason a voting system designed for engineers necessarily ought to scale well to all other fields.

It's not obvious what the optimal downvote cost should be, but -3 points seems like a worthwhile experiment. One wonders what might happen if it were -10, -20, or -50 points...

OTOH it brings up the issue of discounting critical votes, which for wrong answers should seem to merit more reward than approving votes for wrong answers. What would happen if downvotes had a 0 penalty, or even a positive value?

Rather than speculate, what would help is some kind of simulator, perhaps based on some corpus of voting data here. So we could compare the outcomes of different reward formulas and constants given the same users, the same questions, answers, and votes.

  • Yes, a backtesting simulator could prove useful, too.
    – grovkin
    Jul 29, 2018 at 6:29

You must log in to answer this question.

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