Reply to a meta-stackoverflow answer

I have taken this into account in my comments, maybe. But the extra suggestion is marginal, and focusing only on that makes your answer weak (it does not "answer" my "original" proposal)

And in fact, I am not casting the OP in any radiant light, and I am saying exactly the opposite. They don't know what is good and what is not on the subject, so they have to trust votes, and to form their own opinion reading answers and comments (and reasons for downvotes)

Isn't SO aim to be a "manual" (as I've read in a comment)? A reliable reference with a Q/A structure? So the system should minimize the impact of "community members" that behave badly.

So let us suppose the "reason for downvotes" system is up and running. You ask a question. Then you come and read the answers.

Likely they are written by people knowing the subject, or just trying to help you at their best effort; but you are not childlike so you know that some answers maybe incorrect, at the same level of your own hypothetical answer to your own question. This is the average asker, exceptions may exist, and in fact exist.

Then you use however the votes and comments to have an idea on what is the best answer; likely you make your own test, if you can and if your question is "practical". At some point, you have one answer with, say, +15-4 and an answer with +11-4 (allowing to see up and down votes should be possible always, but this is another defect and another thing to talk about).

According to what you have understood until that point, they are both reasonable, but they are also different in several points. But since you expect readers/voters to be honest and to know about the subject, you guess that the +11 answer is indeed better than the +7 answer.

But then you read the reasons for the downvotes. Reasons for +15-4 are like this: "not fully std compliant, see [link]", or "this solution is O(n^2), while it could be easily O(nlogn)" or alike.

On your part, all reasonable. You compare the acquired informations with public comments, other answers, other sites, given references and so on, and you finally find that they are all "reasonable reasons": the final +11 score reflects it correctly.

Then you read the reasons for the +11-4 answer. They all are like this: "asdadsasa", "stupid", "he's wrong!" ... you've no other research to do, this +7 is indeed a +11. At this point, the extra suggestion would allow the asker to mitigate this odd downvoting.

Let us suppose now that the asker is a "moron". He, instead of "mitigating" the +11-4 answers, "mitigates" the +15-4 answers, making it a real +15. The system is still imperfect, but if you say that we have not to be "scared" by downvotes, and that they do not changes things so much, then with or without this extra power, everything would be the same, i.e. the defects are comparable to those of the current system, so the asker being malicious is not a disavantage, it is just like things are currently, too.

Therefore we can focus on the possible advantages.

First of all, let me add that "mitiganting" should not give back lost reputation. It should only adjust the score of the answer. Mitigated answers give no back lost reputation (I think it is better this way).

Now, let us focus on possible advantages. The aim is to help the asker, isn't it? Then, the asker must have as many tools as possible which help him to judge the answers. A reason to downvotes is another tool. Now, why not for upvotes? I explained already in comments that the "picture" can't be flipped, it is not logically symmetric. And moreover, people are scared by consequences of downvoting, while they are not, of course, by consequences of upvoting, so that you can read many more comments explicitly saying "+1 and why".

Now that I have explained it longly, let me reread your sentence above and let me interpret it under another light. Your are basically saying that the asker can't be entitled to vote the best answer, since doing so he autocamatically would say that all the other answer are not (so) good, and who is he to decide which answer is appropriate, and which is not? Moreover he's the asker, so we expect, in the vast majority of cases, he hasn't the necessary knowledge to decide it.

Community as whole should be the only judge.

At this point you can object that if the answer solves his problem, he's entitled to vote it as best answer. But, how does he decide among several answers that solve his problem too? Likely it will be his own taste, or cronological criteria, or even considerations on votes the answer received (and on reputation of the answerer). Again, having reasons for downvoting (and comments, which however already exist) can help him to take a decision. And of course the objection can exist only if the question is "practical"; but we don't care too much of "subjective" questions, since for them the whole voting system crashes miserably having no supports but opinions that, we know, they are what they are, expecially when the person trying to sell them failed embedding them in an axiomatic+logical frame.

To end this part, until this point, I don't see why the extra suggestion (which however I don't "push" too much) would meet such a strong resistence.

Moral binding are inside the individuals, even though they can be manifested outside and coerced by society. If I read, as reason, "downvoted since I hate him", I feel morally bound to "mitigate" this downvote, and the downvoter will appear in my personal black list of unreliable users. You maybe won't feel that binding; your business. But I, as asker, am still interested in reading such a "reliable" reason.

This is already answered above in this very same reply. Let me be redundant, as someone told, just to be sure you've not missed a point.

If you say asker is not there to personally evaluate the opinion of others, you're stripping from him the right to vote the best answer (and to vote in general to answers to his own question, but I agree that it should be not allowed), since opinions and votes are about the answers, their quality, how and why they are right or wrong. So, perhaps it is not his responsibility, but indeed it is his own interest, if he has more information than non-askers, as we are supposing if the "reasons to asker for downvotes" system is at work.

If people are worried about "evil persons" noticing the "mitigation" of an answer, the mitigation could be done adding fake upvotes instead of reducing the downvotes. (Of course, fake upvotes does not increase reputation).

At this point I have again to stress that I am not so much in love with my extra suggestion. It was just an extra suggestion. But since you've taken it into account so widely, I have to explain how it is not so absurd (not more than the whole voting system!)

A stress on an extra alternative suggestion, with a maybe! You're simply great. Let me be redundant, and quote and analyse the whole parethesised extra idea.

First, notice that the form of the sentences is so that it should leave the impression that "it is all to be decided and discussed further".

(Extra idea: the OP can remove the downvote if the reason does not satisfy him/her, e.g. a reason containing just "a reason", just to fill it, is not a good reason... or maybe the OP can give a vote to the reason, and we could think about mechanisms where votes for the reasons influence someway the reputation...)

The part "a reason containing just 'a reason' just to fill it, is not a good reason" shouldn't look too hard, and anyway I've already discussed it before. With 'a reason' I meant a literal string, as "downvoted since I hate him" or "sadsasfasdad", i.e. explaining nothing.

We have two options. These exist since I thought: well, people are forced to give a reason, but people are mostly stupid or stupid sheeps, so we maybe would need something to "mitigate" the impact of "morons" (statistically, since as already treated above, asker can be "evil" too), and promote good people.

The same idea is behind of the thing called "reputation", and if you expect it works, more or less, you can't describe the extra alternative suggestion as totally absurd and incredibly unuseful and/or broken (while it is what it seems you are saying, more or less).

Since the reputation system and votes next to the answers exist already, it is logical to think that future "features" interact with these systems.

So, one option is to give to the asker the possibility to "remove" downvotes (removing them, means no giving back reputation; I have not explicitly said this, but nor the contrary, so if you deduced I meant that, you should have asked before deducing more).

Another option (or), maybe (i.e. maybe good, maybe not, who knows? it must be discussed, differently from me trying to make you be not totally biased) is to make the reason votable. The asker can vote it.

"we could think about mechanisms where votes for the reasons influence someway the reputation"

There's nothing fixed in this statement. Nothing decided. It is even open to couter-proposals, like another value similar to reputation exstimating for being good or evil downvoter; if you think reputation works, and if you find useful information like "accept rate", then there are no a priori reasons to object to this, unless you intend to say "SO is fantastic as it is, any change can only make it worse".

However, in this option of an extra alternative suggestion, I was just saying the asker could vote the reason, and this vote could influence someway the reputation. I believe there are good mechanisms we can think about. But you prefer not to try to imagine which mechanisms could make it better, or discuss why to you there exist no mechanisms that can make it better.

I am not telling my alternative extra idea is perfect, you likely "grasped" a weak part, but I have strong objections.

First: it is obvious that the "mechanisms" (which I did not specify), must be thought carefully and can be also very complex, to avoid it being exploited maliciously or create too much reputation-quake. In the worst case, it would suffer the same limit of the reputation and voting system itself.

Second: if people are more comfortable with comments, and do not add their answers, hardly their reputation can grow. It is the same for downvotes. If you are more comfortable with public comments explaining flaws, your reputation won't be influenced (positively or negatively) by downvotes-reasons. It is your choice. But you can do a counter-proposal, instead of a simple barren critic. For example, you can "flag" you comment as "flaws explanation", and the asker could vote on it exactly the same way of "secret" reasons for downvoting.

However to prevent to mix reputation due to answers votes, bounty or whatever, with "downvoter grade", another attribute can be added to the user, saying the average vote for downvotes' reasons.

I would call it "reasons-grade" (see above too). We expect that users with higher grade are better constructive critic (with the same "precision" in meaning of the accept rates, reputations, and votes)

And what is a question for on this Meta? Isn't it to expect answers that show dark sides, weak points, motivations, counter-ideas...? Or is it more like asking "Do you like it?" to receive "Yes/No" answers? Then, a quiz format would have been more suitable.

At the very beginning I received comments about the question being a duplicate, and focusing not on the "core" of my question. I had to annoy you all a lot to receive a bit of senseful feedback, and anyway inadequate, since it was mostly biased towards a unjustified convincement: it can't work; or unmotivated personal statement: I am not interested in such a feature. Still nobody has "analysed" the proposal (the main proposal I mean, no the extra idea) in a really analytical, critical and propositive perspective.

My light looks so bright since you are mostly in the darkness.

It is why I am with who proposed public comments for downvotes. But people are more concerned with their "reputation" and the fear for retaliations, rather than with doing the best thing to help. So the idea of comments readable only by the asker is born.

A variation could fix the problem: downvotes comments are public, but the user can choose to be anonymous (the system knows who made them of course, so that they can be flagged anyway; and then they should be numbered to allow replies to specific comments)

Other aspect of this objection are already treated above and here I add: asker interested in their problem got solved, will dig more "opinions" and "reasons", s/he will compare them, trust more the one well documented and with references, and will do its own tests (when applicable); and of course, using comments, can ask for more feedback to evaluate downvote-reasons to specific answers.

The down arrow says "unuseful". Now, unuseful for who? And why? This objection of yours is against the SO system, not in particular to my suggestion (extra or not).

My judgement of usefulness is built on trying to have the PoV of the asker and future readers of the question. If I have doubts, I do not push any arrow at all, and limit myself in commenting if I have something to say.

For some question, the usefulness is a fact. Given a question, if the solution is right, automatically it can't be unuseful. I am thinking about question like "help guys, I need to know how to find the transposed of a matrix in C", for example. If the given solution explain how it can be done, or give working code to do it, it can't be unuseful. And in this case it is easy for many many people frequenting the site, to evaluate if the answer is useful or not to the asker. And the asker him/herself, can check the solution in a practical way. Fine points (like the better way to do it) are harder, but a weak but working algo is not unuseful; who can see the weakness, will think it just doesn't deserve an upvote; but it does not deserve a downvote too, since after all it solves the problem that is the reason why the asker did the question. Commenting the weakness is more useful in this case, and it will be done, publicly.

On the other hand, who think that the algo, even though correct and working, is unuseful because of its weakness, should explain it; if s/he is scared by retaliation or counter-comments, s/he won't comment publicly and let the judgement to the asker. Which can use the provided information as already I described above, in another "section". The asker knows well (it's easy) that there can't be counter-comments to the "secret" reasons. But hey, how easy it is if s/he has doubts: s/he can ignore the reason, or add a comment to the downvoted answer, asking for explanation.

And notice, s/he can do it in a way that makes other commenter not aware of the fact that his/her comment was "triggered" by a reason to a downvote! Only the downvoter himself can notice it!

All others "pathological" patterns are just "noise" in SO, and they already happens, and adding more features, like the one I am proposing, do not change the situation from this point of view.

Where subjetive answers are expected, would be all another thing to be discussed. For what can concern my idea, a question tagged as "subjective" (or similar) could "block" the proposed downvote-reason system, and let people downvote, where a downvote in this case must be considered as "I disagree".

Being it complicated (tag can be changed, for example), the dowvote-reason system can remain, and asker knows that reasons like "I disagree" are acceptable and mostly what s/he has to expect. (We must consider that in general, the asker perception of usefulness of an answer to a subjective question is different from the one about "concrete" problems)

Already covered in above "sections".

If they help, why are they also unnecessary if the whole aim of such a site is helping?

Haven't you just said to forget not the fact that OP is asking the question in the first place and "may not know this"?

So, who makes the evaluation and what does it make it good? The answer could be: "the community". So, indeed the asker has to evaluate the evaluation of the community... and to do that, public comments and "secret" reasons, votes, reputations are all what s/he can look at. One more tool, is just another clue helping his/her evaluation.

When you say "you", you should specify the assigned role. Am I the asker? An answerer? A simple reader/voter?

Let us suppose I am the asker. I need to understand why one user thought an answer I find useful is unuseful. I have already given reasons above. I (the asker) want my problem to be solved, and since I am not able to solve it, I expect others do and in the same time I expect to have enough clues to choose among several answers that all solve my problem. I.e. I would like to have also criteria to say which, among the given useful (to me) answers, is better.

I need to do it also since the system invites me to choose the better answer (o, and I would force this too; with a mechanism like this: if you have N or more questions where you've not selected the best answer, you can't ask more question; or, likely better, you can't ask more questions without loosing reputation; like if your question starts with a, say, -2).

The flaw in your point is this: the user X should not evaluate how useful it is for him/her the answer. S/he should try to evaluate how useful the answer is in the context of the question and from the PoV of the asker and of the average user/reader of SO.

If you (reader/voter) do not do so, evil scenarios appear.

For example:

• I am not interested in 80% of the arguments treated in the questions on SO; thus, I find the answers unuseful to me.
• I have already my own working solution to 10% of problems treated in SO questions, so their answers will be unuseful to me.
• I find 5% of the questions maybe interesting, but it happens that in my everyday life I won't find that problem at all; so all the answers are unuseful to me.

At this point, I should have casted downvotes on all the answers of the 95% of the read questions, so far. And nobody could say me I am a silly stupid downvoter, since I have just explained why they are unuseful to me, and, if taking into account the picture you and others gave, those explanations are reasonable and unobjectable (and unnecessary)

What does it happen indeed? It happens I ignore answers to 95% of the questions read so far. And if I upvote one of those, it is because of the context (the question and the asker, not me and the question), or because they are in the 5%.

This is why I think people saying "it is unuseful" (downvoting) should explain someway the meaning of their own "unusefulness" judgement.

I have already explained it. The logic is not symmetric, can't be reverted.

When you point out that something, to you, is right, nobody can ask you to tell the list of wrong alternatives that demonstrate that yours is the right one. This is since the set of all possible answer is infinite (let us call it A); the subset of possible correct answer is finite (let us call it R).

So people downvoting at most have to list the finite subset of correct answers; they can say what "moves" the answer outside R.

People upvoting, at their best, can do the same but they would be repetitive. What could they explain? It is correct, since it is in R, and it is in R since it is correct? Or do you think they have to list everything outside R to demonstrate that in that list there isn't the answer?

Notice that the very same argument is still valid if R of an user does not match perfectly (or at all!) R' of another user!

If the answer is in the intersection, the users agree and then can upvote, and can't say much more about this upvote. On the other hand, if the answer is outside the intersection, the user with its own R' will be able to point to what "moves" the answer outside R' and so into the infinite set A'\R'.

Differently saying it, you can describe correctness or badness only in term of the finite subset that contain the right answers (to you). This is the reason why you can have very much to say if (to you) it is wrong, but not too much if (to you) it is right, except describing the subset R, which would be tautological and not effective as explanation for people disagreeing.

Still another way to say that, with another PoV: dialectic comes from contrast/opposition. If I agree (with an answer), the only things I can say are those that confirm the answer and my agreement; i.e. something not at all useful (namely, it's taken for granted)

No, in fact. And how does this demonstrate the unusefulness of the downvote-reason system?

Already covered in previous "sections".

People finding them useful, can still vote them up.

(Have I said that they can't?)

Upvotes are not more sacred, as explained they are just a different matter.

Explaining an upvote can't add too much interesting information, even if the upvote is done not only for friendliness or whatever. It would just stress the fact that two users have a non-null intersection of their "R" (see above), and this intersection is already "delimited" by the answer itself.

There's a "shaded area" you could think about, but there are considerations in my favor even in those cases.

Explaining a downvote can help always (unless it is of the kind "downvoted since I hate him", but in this case we are in the domain of the "noise" I've already alluded before), even if the reason can't have a direct counter-comment.

I don't see any well argumented reason. When there are those reasons you're talking about, they do not focus mainly and exactly on the suggested idea; or they are widely unsupported by any kind of reasonable motivation.

How it would not be helpful and how actually would not "strengthen" the system can be summarized this way: it is not helpful and does not strengthen the system, in fact it does not solve the very same problems we already have with users being "morons" ("noise" to be more polite).

The question is: can the asker benefit from it? This should have triggered a discussion far before the closing of the question as duplicate, it should have triggered "counter-proposal", or more extensively motivated objections. How much had I to annoy you before to obtain a bit of feedback that can be considered as such? My answer is (and yours too, I suspect): too much.

But this is my only suggestion. Why do you treat it a part? Above, was you talking about someone else suggestion?

You all (if not all, many) have said that explaining the flaws of an answer is useful. How could it become not useful if tied to a downvote? It becomes apparently half useful if secret, ok. But I explained why there's this requirement, and how the problem is already fixed by itself: the asker can comment and request for feedback on the reasons. Downvoters fearing for their reputation are saved, and in general the system helps forcing people to be "honest"/serious when downvoting.

This is a reason to eliminate votes and comments tout court. Or it can't be used as a reason to explain why downvote-reasons are not desiderable (being secret or not, with or without the extra idea).

It's a supposition I can contrast simply stating my convincement that it would help indeed. The only way to prove it would be to do it. If a Q/A site comes into light without a voting system, people can say that adding it, does not ensure benefits. Then it is added, people gets accostumed to, and everything looks fine, and another conservative loop can begin.

Adding features have a cost, and there can be "technological" limits, I can understand it. But then let us shift the conversation to those topics.

S/he is still free to comment, replicating the reason publically; the system could be made so that s/he does not need copy-pasting, so that it would be enough to check a checkbox to see the reason as comment instead of being secretly given to the asker.

I agree, but it seems this kind of proposal meets two kinds of strong opposition: I want not to be the target of retaliations, and I want not to be forced to give explanation.

My suggestion tries to fix the first problem. As for the second, forcing people to give explanations can't make SO worse; in the worst case, it makes it the same. In the best case, it makes SO a more useful place.

"Honest" downvoters have always a reason they can share and can't be too much annoyed by being forced to comment. Superficial downvoters could start to become less superficial, or reduce the number of easyly given downvotes.

So forcing giving a public opinion would benefit the OP far more than just letting them see one user's opinion!

Ok! But my proposal is not to remove what currently is working, so how would it possible that the OP sees just one opinion? Comments are still possible and downvoters are free to comment their downvote publicly.

As I thought it initially, it seems they have to copy-paste the reason as comment, if they want it public. Ok, then as already said, let us allow them to make their reason a public comment.

Who has already this habit, won't notice any change: he currently click on the down arrow, then write a comment, or first write the comment and then click the down arrow.

With the "new" system, he clicks the downarrow and he's prompted to enter the comment-reason. If he wants it public, have to click a checkbox (or viceversa: the default could be the public comment).

If you want to comment and not downvote, you still can.

If you are the kind of downvoter always in hurry, then you have to slow down and think about your reason for downvoting. This can't make the place worse.

From the asker PoV, a secret comment that can't have a direct counter-comment is better than a silent unmotivated downvote, or in the worst case, it is the same. If the asker gets doubtful for a downvote, or for a secret comment, in both cases, as said, he can ask for feedback in comments.

I don't see how my proposal would reduce the amount of information the asker can use to evaluate the community evaluation and the answers themselves.

And this is fine and not incompatible with my proposal.

(Even though about this, I would allow everyone from the very beginning to be able to see upvotes and downvotes separately).