This question drew my attention to the method used for minority opinions in questions that don't have a clear cut answer. While I understand the democratic approach to answers, what seems to happen in practice is viewers read through the top answer (or maybe the first few) and upvote or downvote as they see fit. This makes answers below those receive less attention in general than others which is frankly undemocratic, similar to the suppression of opposition in totalitarian regimes (not that this is to that level). Anecdotally, the posts I've had that are in minority opinions received 6-8 votes total (up or down) compared to up to at least 20 for the top answers in some questions (This seems to be widespread).

Is there a way to fix this problem? Do we not see it as a problem?

Here are a couple solutions I've thought up (none of my arguments are based on my answers, but could come up in other situations)

  1. Display upvotes and downvotes separately sorting by upvotes only.
    • This system would allow minority opinions with substantial merit on one side of a potentially close debate to be seen despite the downvotes from the other side.
  2. Do not display downvotes from people who upvotes other answers (or answered the question themselves)
    • Since this tactic essentially votes twice and exasterabates the condition I described earlier. Of course there is no reason these votes shouldn't be factored but displaying them and considering them for answer order is problematic.
  3. Allow for a promoted minority opinion.
    • Enable an answerer with an opinion directly contrary to promote the answer (pending enough votes similar to reopening a question) to be placed on the second position on the order hiding the vote count. Then, when it s received enough attention (by days or up/downvotes), a vote could be started to determine if it should remain there or be returned to it's original state permanently.
  4. Any combination of these options

Here are some other examples of why this system would be an improvement by allowing opinions contrary to the majority to be more easily accessed:

  1. https://rpg.stackexchange.com/a/67301/41726

    • In this case a reasoned argument for the other option to affecting the magic item is given without much hope of ever being seem from the much more accepted modifying the item answers
  2. https://rpg.stackexchange.com/a/63618/41726

    • In this case, the answer provides an elegant solution not accounted for by the too voted answers that addresses the question in a unique way but is likely not to get the attention of the other answers since they appear so widely accepted
  3. https://rpg.stackexchange.com/a/107356/41726

    • This answer outright states it's intent to provide an alternative to the other answers which for the most part support finding ways to make the rolls work. It is a wholly different approach to the question which isn't disqualified as a non-answer but may be the right answer to another asker who finds this question with a slightly different mindset.
  • 2
    \$\begingroup\$ This seems related to the fastest gun in the west problem. \$\endgroup\$ Mar 10 '18 at 15:36
  • \$\begingroup\$ @dopplegreener very similar indeed with more emphasis on the desire (from me in any case) to create a database here for high quality rulings, which may be suppressed in some cases by the current system. \$\endgroup\$ Mar 10 '18 at 15:46
  • \$\begingroup\$ Ultimately, what would be the goal of any change? More upvotes for minority opinions? \$\endgroup\$ Mar 10 '18 at 16:46
  • \$\begingroup\$ @HeyICanChan More visibility of minority opinions. When multiple answers are relatively similar, minority opinions get pushed below and many viewers will see the few answers at the top which all seem to agree without seeing the disagreement in controversial questions \$\endgroup\$ Mar 10 '18 at 16:48
  • 4
    \$\begingroup\$ But what mechanism do we have that can sort similar answers from dissimilar ones? Or even contrary and agreeing? \$\endgroup\$
    – Rubiksmoose Mod
    Mar 10 '18 at 17:11
  • 2
    \$\begingroup\$ Option 2 is a complete no-go for me as it makes wide-ranging assumptions about the way everyone votes in general. There are countless very good reasons why you might up-vote one answer and downvote one or more others to the same question. Hiding this voting information would be extremely damaging to the way the stack works imo \$\endgroup\$
    – Wibbs
    Mar 10 '18 at 17:46
  • \$\begingroup\$ @Rubiksmoose I suggest that a person who believes there answer is a minority opinion (a la US Supreme Court) nominate it as such and then the vote goes through a similar system to how close votes work \$\endgroup\$ Mar 10 '18 at 18:06
  • 4
    \$\begingroup\$ I'm amazed that a two day old question with 4 answers is prompting this. I'd suggest taking a look at some of our highly-voted questions that have ten, fifteen, even twenty answers attached to them. I'm not saying that your stated issue is or is not a valid one--just that this example seems like a really weak example. \$\endgroup\$
    – nitsua60 Mod
    Mar 10 '18 at 18:39
  • \$\begingroup\$ @Nitsua this question drew my attention to the problem because it made me look deeper. Not as the primary cause for thought. \$\endgroup\$ Mar 10 '18 at 18:40
  • 4
    \$\begingroup\$ I guess I'm saying that if you want a proposed change to the system architecture to be taken seriously, I suggest you find some better examples to put in front of readers. They're out there. \$\endgroup\$
    – nitsua60 Mod
    Mar 10 '18 at 18:41
  • \$\begingroup\$ You'll also need to raise it on the overall stack meta, as what you're proposing changes a fundamental design decision about the way the whole network works. It's not something that can or should be changed on an individual site \$\endgroup\$
    – Wibbs
    Mar 10 '18 at 18:49
  • \$\begingroup\$ @Nitsua I added a few examples \$\endgroup\$ Mar 10 '18 at 19:00
  • 8
    \$\begingroup\$ Given the current voting trend on this suggestion I just wanted to point out that voting is different on meta: "On posts tagged feature-request, voting indicates agreement or disagreement with the proposed change rather than just the quality or usefulness of the post itself." So the votes are not saying that this discussion is unwelcome, just that there is disagreement about your proposed feature. \$\endgroup\$
    – Rubiksmoose Mod
    Mar 11 '18 at 3:36

The system has various ways this is supported.

  1. You can sort answers not by votes, yourself. Look at a question and look at the top of the list of answers - see that (active, oldest, votes) option? You can do "active" and get answers getting newer activity.

  2. You can bounty an answer, one of the reasons is "deserves more attention."

You seem to think downvoting is more of a problem than it usually is. The normal pattern is that some answers get upvotes and others don't. Downvotes are not usually given to "all other answers" - they're given to particularly bad ones, or ones someone disagrees with very strongly.

Take your example What can I do when I accidentally gave out an overpowered item?. At the time of this writing, the votes on the answers are (+73/-3), (+172/0), (+103/0), (+70/0), (+15/0), (+7/-1), (+7/-2), (+4/-1), (+3/-1), (+2/-1), (+4/-5), (+2/-3), and so on down to (+1/-6). You will note:

  1. A minority answer is the one that got accepted, raising it above other higher rated answers
  2. Most good answers have no downvotes. The really poor answers at the end do, but all the answers that have decent upvotes aren't downvoted by all the others who agreed with another answer.

There's not a lot to do about "people don't read many pages of answers" or "people don't go back to read new answers to old questions" (though the latter bumps the question and gets put in the "Late answers" review queue, which gets eyeballs esp if combined with viewing by active).

TL;DR it's not a problem; to the degree that it is there's compensating functionality in place already.


I'm not going to tackle the larger issue right now, but I do want to make an important procedural observation:

This feature request could only be implemented for the whole Stack Exchange network, not just RPG.se. Consequently, any debate we have here has at best an infinitesimal chance of causing such a change, since our little site is unlikely to merit overhauling the bedrock systems for over 200 sibling sites, many of them vastly larger than ours.*

There is merit in discussing the minority opinion issue, but discussing it only through the lens of nigh-impossible technical solutions may detract from the utility of the discussion.

* In practice, the most likely means by which a discussion here could have technical impact is if the technical discussion here was used as an example, along with similar discussions on other Stacks' local Meta sites, to demonstrate widespread support to change the whole network's systems. Then a [feature-request] at the network's Meta site, Meta Stack Exchange, would have the beginning of evidential support for a discussion of technical change. Then, if it got widespread support and the network owners agreed, it would go into their design & development queue, to be implemented at some point.

  • \$\begingroup\$ I'd be glad to out in the research to support the change on the full network but I have a question, how often does this kind of change happen (is it worth my time). I'm very new to the network \$\endgroup\$ Mar 10 '18 at 19:04
  • 2
    \$\begingroup\$ @DavidCoffron Not very often — the devs get lots of ideas thrown at them, and only the best that align with their overall goals for the network tend to stick. Then there's dev and testing time, sometimes on the order of years even for changes that are popular and well-received. Looking through MSE [feature-request] and MSE [feature-request][status-deferred] can give an idea of the process. \$\endgroup\$ Mar 10 '18 at 19:17

I'm going to take a crack at this even though my points are very basic. I want to say upfront that this my answer (like your question) is not at all a commentary on any of my own answers (even the one you linked where we clearly disagreed).

You say that the problem is the lower voted answers receive less attention, but I think that only #3 would potentially accomplish that (and definitely not without creating serious side issues).

What constitutes a "minority opinion"?

The SE system is designed to find the one answer that best answers the question.

The whole system is based off the idea that a community of experts can through up and downvoting give rise to a singular best and correct answer to the question. In this system there are not "minority opinions" so much as answers that are judged to be less good (to varying degrees) answers to the question than those with more votes.

In this system, both the up and downvotes are key and meaningful indicators of the perceived quality of an answer.

#1 Regarding sorting by upvotes only

The first problem here is that this does not seem to solve the problem you perceive. You imply that downvoted or low-upvoted answers do not get read or voted on as much as the top answers and seek to try to equal that out. However, if an answer is getting a lot of downvotes it is being read and people simply see an issue or issues with the answer (there are many reasons to downvote an answer).

The second problem is that it seems to be contrary to the way the SE voting system has been set up to work. Changing the sorting of answers to ignore downvotes would mean that an answer with +1/-10 (an answer that clearly has exceptional problems) would be displayed at the same level of community approval as one with +1/-0 which seems to defeat the entire purpose of downvoting and makes the entire idea a non-starter for me.

You also say that the downvotes are "from the other side" but I disagree with that assessment. This implies that there are only two sides for one thing (which is definitely not always the case). For example, there is no reason that I couldn't or shouldn't downvote an answer that does not meet whatever standards even if I do not agree with the other answers.

#2 Hiding downvotes from people that have upvoted/answered

What you perceive as voting twice is not cheating the system. It is the system working as intended. If a question has a terrific answer and an awful one, upvote the former and downvote the latter. The fact that you have voted on other answer does not mean you have a conflict of interest or that my opinion of other answers is any less valid. Anybody in the community with high enough rep is welcome to do the same. What makes you think that if this was enacted, that people wouldn't just choose to downvote the "bad" answer instead of upvoting others?

#3 Promoted minority opinion

I am disregarding any potential technical complications of such a request because I have no idea if they would be problematic or how bad they would be.

I am confused, isn't the current best way to promote an underappreciated answer (possibly a minority opinion) just to upvote it? It seems highly unlikely that people would want to upvote the answer through another system when they declined to vote on it the normal way the first time.

In the end, the person that asks the question is the one who decides the answer that is most helpful by selecting it

All these systems deal with vote sorting, but, in the end, it is the person who asked the question that gets to choose whatever opinion they want and select it as the answer. Votes are simply used as a means of sorting and indication, but it the querent find the minority opinion more helpful they are welcome to choose it.

  • \$\begingroup\$ I see the system as a way of multiple experts to realize the right answer within a group. When one cannot be agreed upon, onlookers should be able to see the contrary but possibly valid answer (again not based on any of my answers in particular). In this case, I would certainly support a minority opinion that has the chance of being a better ruling even if i would personally support another. There is the case where this is not one right answer and next best answers sometimes deserve the right to be seen so onlookers can make informed decisions \$\endgroup\$ Mar 10 '18 at 18:35
  • 1
    \$\begingroup\$ @DavidCoffron but the 'right answer' is the one selected as such by the querent, no more, no less \$\endgroup\$
    – Wibbs
    Mar 10 '18 at 18:37
  • \$\begingroup\$ @Wibbs except the questions are stored and then seen by others who may support a different "right answer" and they can't ask again because of how the system treats dupes. This makes the right answer the first thing seen by all readers while I believe support for valid but minority answers (especially in types of questions without hard rulings) would benefit the site \$\endgroup\$ Mar 10 '18 at 18:39
  • 2
    \$\begingroup\$ Regarding #3: bounties are another traditional and system-supported way of "promoting" an answer one feels is underappreciated. \$\endgroup\$
    – nitsua60 Mod
    Mar 10 '18 at 18:40

You must log in to answer this question.

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