# Rationale behind downvote button

First: as a new non-scientist member of Physics SE I want to say thank you for the wealth of knowledge, interest and patience with my ignorance I have encountered here.

Second: I was very surprised to see that there is a "downvote" function for posts and answers. After trying to think of scenarios where such a function would enhance the quality of scientific discourse I am still at a loss.

The strength of this f̶o̶r̶u̶m̶ question-and-answer website is that it is highly specialised, and it's assumed that the vast majority of questions are not understood or of interest for the vast majority of people, be they physicists or not. I for one would be very busy if I would take the downvote button seriously and start voting on all posts that are "unclear" or "not useful" to me, the same way I upvote those that I find interesting. (Edit after comment: by this I mean that I don't understand 99% of the questions, but that says nothing about their quality, sincerity or usefulness for others, only my lack of knowledge.)

I do understand the 'upvote': it gives an indication on how popular a questions is, which means that there's a higher possibility that I will find it interesting. And it gives me an opportunity to send a small appreciation to those who taken the time to answer my question, or asked questions that enlightened me. But how can the distinction between a post with 0 votes (not popular) and one with -10 (very unpopular) help me to find the information I want?

So my honest question is:

What exactly is the rationale for the ability to downvote?

Can you point me to a specific case here where downvotes has achieved the intended function, that could not be achieved with other tools available? Is there research and/or data that shows that such a button enhances the quality of questions and answers?

Note after comments: I fully understand that the caretakers of this site are rightly proud and protective of it. I see myself as a grateful guest on PSE, and by no means am I criticizing or proposing changes. As a naive empiricist I am honestly curious on how exactly downvoting can have the function stated over and over again on the site and in comments below, as for me it seems counterintuitive. I apologize if this was a provocative issue. Please feel free to express your critique by downvoting this question.

• ...how would you propose to express consensus that a post is wrong or not useful if not by the community overwhelmingly downvoting it?
– ACuriousMind Mod
Feb 14 at 17:10
• 1. I would propose to not encourage expressed consensus. 2. If you consider a post being wrong I propose you point out the mistake. 3. If you don't find it useful I propose you ignore it. Feb 14 at 17:17
• Downvotes are central to the StackExchange system. Questions about them should be on Meta.stackexchange. I will not that a proposal to get rid of them or change them or require a comment come up regularly and are not well received. Feb 14 at 17:35
• And an upvote does not mean 'popular' - as the tool tip says, 'This question shows research effort; it is useful and clear' Feb 14 at 17:36
• @JonCuster Thanks for your comment. I have no intention to propose anything, just understand its function in helping people learn about physics. Feb 14 at 17:39
• @JonCuster Yes, and many upvotes means that many people find it useful and clear = popular. Feb 14 at 17:40
• Hot Network Question voting clearly shows cases where 'popular' is neither useful nor clear (or well researched). I disagree with your equivalency. Feb 14 at 17:44
• I for one would be very busy if I would take the downvote button seriously and start voting on all posts that are "unclear" or "not useful" to me, the same way I upvote those that I find interesting Welcome to the club. Feb 15 at 2:34
• @peterh Thanks for your comment. "However, it does not say anything about the skill of the voters (...)" Yes, and for me it's assumed that every question is viewed by at least 10 people who doesn't understand it, before someone with an answer comes along. I just can't put my finger on how having these 10 people decide the usefulness of the question is helping the site. Feb 15 at 5:42
• If you don't find it useful I propose you ignore it. The problem with this approach is that ends up clogging the system. The site is primarily a resource: its strength is partly in the ability to quickly search for useful information. Thus, ranking the information is essential. Of course the ranking is not absolute and positive votes do not correlate with usefulness since trivial or fashionable questions get more votes than less trivial or less fashionable ones. The latter are much more crucial to the site than the former. OTOH downvotes do correlate better with poor answers. Feb 15 at 15:13
• @KyleKanos Ah, I see. To me it just seems like a very blunt and unhelpful tool, in a scientific context, to "suggest an answer is incorrect" by common voting, rather than pointing out the mistakes by comments or providing a better answer. Feb 15 at 15:41
• You can't make a statement that downvotes are harmful (or at least "not good") without evidence to back it up. The default sorting mechanism of answers (see the tour page) is descending by score (time is 2nd condition) and after -3, answers are "greyed" out to somewhat hide them. In some cases, such posts are even deleted. Feb 15 at 16:17
• If I may: I don't think the objective here is to improve the scientific discourse: the objective is to provide information to the readers or visitors to the site. Downvotes do this in part. It's not a perfect system, but except for snowflake posters it does work fairly well on average. (IMO this site does not have a downvote problem, but it does have a mild upvote problem, i.e. some people upvote too often compared to their downvote frequency.) Feb 15 at 16:25
• "What exactly is the rationale for the ability to downvote?" I think the actual intent here is "I don't like the reason for down votes. Let's have a debate about it." There isn't anything wrong with not liking down votes, but it seems like you already understand the points, and anyone who tries to offer further explanation is met with resistance. Feb 16 at 10:42
• @BioPhysicist Thanks for the comment. To be more precise: I understand the intended function of the downvote, but I do not understand how they can achieve this function. And therefore it would be interesting to see some data (and no, I don't know what that data would be, that's why I'm asking) on how they perform, so we don't need to have this debate. Feb 16 at 11:32

Low quality questions on the front page send a signal to first-time visitors that this is not a site worth exploring. Downvoting low-quality questions helps to keep them off the front page.

• if only for this reason downvoting is useful. Feb 17 at 17:39

After trying to think of scenarios where such a function would enhance the quality of scientific discourse I am still at a loss.

I think the first problem is that it's no scientific discourse we're aiming for here. This site is a question and answer site to which we are (voluntarily) providing (potentially) useful information to readers. Whether the content is useful or not is intended for the community to decide, literally why voting is important.

I suspect that it might be useful to consider scores on posts as being akin to grades on homework or tests: it provides a useful metric to others as to the knowledge content in the assigned work. While we may be used to the percentile grading system, it wouldn't be too difficult to adjust the scoring to see how far above or below the median (or even passing) would be. For instance, instead of a 90%, you got +5 to a mid-B or a 60% being -25 to the mid-B. In this sense, it is easier to see how bad you did (due to negative score) versus how good you did (positive score).

Downvotes are very much the same here: they convey a measure of knowledge content in the post: those with useful content are positively scored while posts with little useful content are negatively scored. So really it's just a measure of the relative worth of a post, given our own understanding. And this measure of relative worth is a nice by number attached to the top of your post.
Comments, on the other hand, are tiny and buried under the post. So not only do you have to read the entire post, but then what could be dozens of comments to parse the usefulness or potential issues of the post. It should be obvious that a big fat $$-X$$ at the top is a lot more clear indicator of problems than a tiny "this is wrong!" at the bottom.

In regards to alternatives you've proposed,

1. No expressed consensus
• Then everything is correct and there is no reason to vote up or even comment on anything. In fact, why bother asking or answering?
2. Comment the mistake/error
• This is a good first step (I've argued so previously), but it does not, by itself, work to replace votes because (a) comments are generally temporary & such a comment mentioning the error might be irretrievably erased and (b) there is no guarantee that future viewers will actually read the comments to find the error.
3. Ignoring the post
• Then not only is everything correct, but you are actively damaging the site because users who don't know better might think the bad post is actually good when you really know it is wrong (#2 below addresses the case when you cannot judge the content).
4. Flagging it for moderators to review
• Flags, as discussed in the Help Center, are for specific instances of posts (e.g., someone using an answer to reply instead of a comment, spam), rather than a catch-all for bad content. Using moderator intervention in this manner is actually not in line with their wants and, even worse, then relies on these 6 people being, effectively, the ultimate arbiters of truth.

Hence, none of these provide any meaningful way to convey an incorrect/problematic post; voting is the only reliable mechanism to indicate a usefulness measure (positively or negatively) while an answer still exists (for deletion, see this and this Help Center article).

For other remarks on downvoting in general,

1. Don't have the time to downvote
• Well this doesn't preclude you from actually downvoting, it just says that you're either not interested in curating or plain lazy (it really doesn't take much time or effort to hit that button as you see posts).
2. One may not understand the content
• Great, it happens to all of us (because we're not all experts in all fields). Fortunately, this doesn't preclude you from voting on subjects you do know, just the ones you don't know.
3. There's more people who don't know enough about a given topic than who do know about that topic
• This is probably an idea you had based on your experience with the Hot Network Question effect (see this Meta post and links therein), but in most cases, this is not true. Most posts only get a few upvotes. More advanced or specialized topics get even less.
• Thanks for taking the time to provide an answer. It does seem that my question was very unclear, but I will try to comment on some of the misunderstandings caused. Again, I am not here to tell anyone how to run a website, the proposals you listed are either given by me on specific questions on existing alternatives for downvoting, or given as alternatives in PSE guidelines. But all this can be found in comments to the question. Feb 16 at 5:51
• (1) Consensus can be useful, and it would still be expressed by upvoting. I am hesitant to say that encouraging a consent in a community is good for finding the correct answer, but this might have been a misunderstanding from my side of a comment above. Feb 16 at 6:04
• (3) By ignoring I mean on an individual basis: if I personally don't find to specific question clear and useful, I will not judge it but leave it to someone else who do might find it clear and useful to interact with. If the question is inherently unclear and useless, it will be ignored and have 0 upvotes. (4) I agree that flagging seems a bit drastic for most cases. Feb 16 at 6:07
• (1) My 'don't have the time' remark in the question was not to be taken as a serious argument. I was was just trying to illustrate the absurdity of me downvoting the questions I find unclear and not useful, as I, due only to my own ignorence, find most questions unclear and not useful for me. I now understand that this remark leaves too much room for misunderstanding. Feb 16 at 6:15
• (2) Again, my worry is that people who don't understand a certain question will judge it 'unclear and not useful', therefore downvoting and burying it before someone who actually understands the question can interact with it. I admit that I have no evidence either for or against that this happens, but in my mind it seems likely. Feb 16 at 6:19
• "Consensus can be useful, and it would still be expressed by upvoting." - no, popularity and audience size would be expressed. An answer posted five years ago can be compared to one posted five days ago based on their scores and the ratio of downvotes to upvotes. Without downvotes the score is only measuring how many people saw the answer mixed with how many liked it (and frequently, how many like the user that wrote it). We already have view counts and timestamps, so any "upvotes only" mechanism is misleading as well as redundant.
– Nij
Feb 16 at 6:30
• (3) 'There's more people who don't know enough about a given topic than who do know about that topic' This might be another misunderstanding due to poor phrasing from my side. Cont. below: Feb 16 at 6:30
• As a scenario: a very deep, thoughtful and complex question on and obscure topic is asked. Due to its nature the first 100 people who view the question does not understand it. 10 of these gives the question a downvote because it seems 'unclear' and 'not useful' to them. Despite the -10 rating one of the few people who understands the topic finds it, leaves an enlightening answer, and upvotes. The question is now left with -9 rating. How is the judgement of the 10 people who downvoted the question of any relevance? Feb 16 at 6:34
• '...it might be useful to consider scores on posts as being akin to grades on homework or tests..' Yes, and by the current PSE system the rating of the tests would be done by voting among the other students. Which might be effective, but it might also be unprecise. Feb 16 at 7:43
• No, it really isn't. Downvotes simply don't happen to the same extent as upvotes. If they were really as influenced by interpersonal opinion instead of content rating, that wouldn't result. Making up examples that don't reflect the reality of voting only indicates the naiveté in the discussion - there's no solid basis where it has a foundation worthy building on. Fix that first, and then suggest changes, if you still think they're needed.
– Nij
Feb 16 at 8:59
• @erikm you have posted way too much content in the comments here for me to possibly respond to (which really goes to prove my point in (2b)). Unless you didn't actually mean to ask, "Why do downvotes exist? How do they help?", then your question was perfectly clear. You just don't like the answers. Feb 16 at 14:53
• @erikm It's fine to not like the answers you get on Meta. But if you think your question wasn't answered, then rather than making 6 lengthy comments debating the points, respond with why you think the post doesn't answer your question. Feb 16 at 15:08
• Searching for 'downvote' on Meta yields 14,250 results. Seems unlikely that the OP has actually gone through them all (and I certainly won't). The bottom line remains - downvotes are core to SE sites and complaining about them here does nothing useful at all. Feb 17 at 14:10
• @erikm With due respect: downvoting is an earned privilege so the suggestion is that people who downvote questions do so even if they do not understand a question or an answer is rather condescending. Feb 18 at 2:46
• @erikm feel free to do so. Your worry that “people who don't understand a certain question will judge it 'unclear and not useful', therefore downvoting and burying it before someone who actually understands the question can interact with it” is IMO unfounded and overly general, and rather different than suggesting that, with the exception of some sensitive users, the (down)voting system works reasonably well. Feb 18 at 21:02

Think there is a mechanism like, you cannot change the vote of a downvoted and commented answer, before the answer is not re-worked. Accordingly, I would love to see the common attitude, that a downvote comes with a qualifying comment, so the author has a chance to re-act and common quality of both, Q & A, is raised.

Else downvoting remains on the level of selectionism.