-14

Recently, I've noticed that a lot of C++ questions get downvoted within minutes of them being asked. It must be the season of exams as we do see a lot of basic questions being passed by.

To me, that doesn't give nice optics to those first time users. I can understand that those questions are merged when asked several times or closed and downvoted when they ask us to do their homework.

However, there are also well asked questions according to my definition of it: understandable problem and reproducible code snippets.

Two examples I was involved in recently:

To me, both questions are asked well and clearly show a beginner in need for help. I agree, I might have taken more time to search for a duplicate, however, answering it takes about the same time and gives an answer tailored to their problem.

Why would both the questions and the answers get downvotes because people are at their first steps of learning the language?

closed as primarily opinion-based by πάντα ῥεῖ, Robert Longson, gnat, curiousdannii, Ward Jan 14 at 6:03

Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise. If this question can be reworded to fit the rules in the help center, please edit the question.

  • 2
  • Why would it be off-topic? – JVApen Jan 12 at 8:17
  • 4
    I can't speak for why people choose to vote the way they do, but I imagine a lot of times people in those situations are following the part of the downvote tooltip that says "This question does not show any research effort...". So that does seem like a perfectly valid reason to downvote (not that downvotes necessarily need to be justified). It's just a matter of opinion. Some feel those type questions deserve downvotes and some don't and they each have a right to those opinions. – n8te Jan 12 at 8:19
  • 3
    The first example you give was closed as duplicate by me, and I found those dupes within seconds with a google search. Your second example was close voted by me as off-topic for having a typo (very simple problem) and no debugging efforts were shown. I didn't DV any of your answers there though. – πάντα ῥεῖ Jan 12 at 8:19
  • 2
    @πάνταῥεῖ yes, why are those C++ questions off-topic? EDIT: The older question asks "Should one downvote answers to off-topic questions?" – Mari-Lou A Jan 12 at 8:19
  • 8
    Voting in any way (up or down) is okay. It may not be liked, but everybody can vote however they wish. If you start dictating how people should or shouldn't vote, the voting system becomes problematic. – Jason Bassford Jan 12 at 8:24
  • @jason I don't mind the down-voting, though I do want to understand what the reasoning is behind it. Looking at the list of recent questions and seeing 75% with a down vote feels uncomfortable to me – JVApen Jan 12 at 8:27
  • 1
    ... So, the comments here indicate that your assertion that those questions are well asked is false. – Raedwald Jan 12 at 10:24
  • @Raedwald Even if the question is poor, the answer might not be. As far as I understood, the main purpose of this portal is to help. And it's ok to send people back if their questions are not understandable. But if they are - which seems to be the case, otherwise providing an answer would have been impossible - I think providing that help is nothing bad. Please correct me if I'm wrong here. – Psi Jan 12 at 10:55
  • 2
    The title of your questions refers to "a well asked beginner's question". People may be downvoting based on not agreeing with that, or something else that makes them think it does not show any research effort; it is unclear or not useful. – PolyGeo Jan 12 at 11:38
  • cross-site duplicate: When is it justifiable to downvote a question? – gnat Jan 12 at 12:57
  • @gnat It's talking about votes to the answers, not to the questions – Psi Jan 12 at 14:51
  • 2
    @JVApen: "Looking at the list of recent questions and seeing 75% with a down vote feels uncomfortable to me" Alternatively, I feel uncomfortable with the fact that 75% of questions in the C++ tag are sufficiently bad to get a negative score. Maybe that's the problem that should be fixed. – Nicol Bolas Jan 12 at 15:34
  • @NicolBolas Agreed, I'm looking forward to the new 'create question' assistant – JVApen Jan 12 at 15:36
  • @JVApen: "downvoting questions is cheap" (the downvoter doesn't have to pay via any sort of rep reduction for downvoting) ... Therefor, just ignore anonymous downvotes of questions, since there is nothing to explain how the question could possibly be improved, so why bother. For short: niks van aantrekken ... – Pierre.Vriens Jan 12 at 15:58
19

I agree, I might have taken more time to search for a duplicate, however, answering it takes about the same time and gives an answer tailored to their problem.

Exactly, that's where it goes wrong. When you hover over the downvote button, a tooltip will appear that says:

This question does not show any research effort; it is unclear or not useful.

Their question is not unclear, but any research effort would've led them to find the duplicate via Google. And it's certainly not useful if it's the umpteenth duplicate of an 'undefined behaviour' question. Yes, it might be useful to them, but please keep in mind what the Stack Overflow community is trying to do: building a high quality 'library' about programming questions. Having multiple instances of basically the same question around is counterproductive, even if it has the benefit for the OP you mention: a tailored solution to their problem.

You state those people are beginners: the best (only?) way to learn programming is viewing examples and then tailor them to their own needs. Otherwise they'll just learn copy-pasting and not learning how to program at all.

As for downvotes on the answers (I failed to notice you were not the questioner): I can imagine that people want to discourage answers on such questions. Some users (group 3 in this graph; note that I'm not claiming you're one of them) thrive on the easy reputation these questions bring - even a -1 accepted answer gives you +13 reputation. People (from group 1) have resorted to downvotes on the answers after seeing that leaving comments doesn't help stopping this behaviour.

  • Very good answer, but are these C++ questions massively downvoted or do they have 1 or 2 downvotes? – Mari-Lou A Jan 12 at 8:27
  • Thanks. I don't know how many downvotes they get (maybe πάντα ῥεῖ knows, he keeps a close eye on the C++ tag) – I only see the special cases featured on Meta or sometimes via SmokeDetector. – Glorfindel Jan 12 at 8:29
  • 1
    @Mari-LouA Usually those questions aren't really massively downvoted, but I (and probably others) working at that tag use DVs to spot off-topic questions, and yes I also DV most off-topic questions I am voting to close along. An exception are duplicates I hammer close, but that depends a bit on how obviously the duplicate was to find. – πάντα ῥεῖ Jan 12 at 8:35
  • Especially for a beginner trying to figure out what's wrong with their code, the possible duplicate might not even be obvious. Even when you mark the question as duplicate, a beginner often struggles to transfer the errors in one particular case to another one. Then, he states that the duplicate is not a real duplicate, because it's a different case (because of the lack of understanding). In my eyes, answering questions that feature specific code with an answer taylored to that specific code is not wrong. However, I agree that you cannot dictate on how to vote. – Psi Jan 12 at 10:48
  • 4
    @Psi Who said that site is for beginners to learn programming? It's not. – πάντα ῥεῖ Jan 12 at 12:29
  • @πάνταῥεῖ Where did I state that I assume the site is for beginners? I said, I understood that the site is for finding and providing help. It says nowhere that it is only for beginners, but neither it is stated anywhere that it is only for advanced programmers. I would appreciate if you respond to what I actually said. – Psi Jan 12 at 12:39
  • 5
    @Psi SE is not a personal help desk. Most users (including me) believe that we're trying to build a library of high-quality Q&A for future readers, with a side effect that it also helps the OP. Now, since I can only see the 2nd question, what are the chances in the future that any users will stumble upon that question (with a non-descriptive title, no compiler error message) and learn something? – Meta Andrew T. Jan 12 at 14:42
  • @andmyself Maybe you get me wrong here. I do not believe that poorly asked questions are suitable. All I said was, that in some cases a beginner might struggle identifying a duplicate due to the lack of understanding the issues. This topic is about downvoting good answers to those questions whithout even doing anything more than that. Whenever I downvote an answer, I explain why I did that. But, if you have a look at the second question, it got downvoted, but no one explained why, nor did the question get marked as duplicate. – Psi Jan 12 at 14:49
  • @Psi I see, apparently I missed the real question because the title of this meta post doesn't mention about downvoting answers. That said, regarding answers getting downvoted, it's a controversial topic that has been discussed many times, like Why do I get downvotes for giving a good answer for a bad question? – Meta Andrew T. Jan 12 at 14:58
  • @andmyself True, the question got changed. I just wanted to point out that I do not want to question that poorly written questions should be revised. I just wanted to state that beginners might have troubles identifying a different question to be related to their problems. – Psi Jan 12 at 15:01
  • 1
    @Psi and then those beginner become our interns and then we spend hours or days fixing our codebase after they get their hands on it because they only learned with copy and paste code. It is important that OP's Struggle with the duplicate and figure it out - it really is – LinkBerest Jan 13 at 2:18
  • @JGreenwell just read what I actually wrote please... What I understand from what you are saying is "I'd love to get rid of all these nasty beginners". I did not say let them do what they want. I said: when you dowvoted an answer to a beginner's question, you should indicate where you would see improvement. Otherwise you won't stop the unwanted behavior or you discourage anyone to do anything anymore here. And then people wonder about the group of apathetic people in the diagram... – Psi Jan 13 at 3:05
  • @Psi no, what I'm saying is "I'd love to teach these beginners something which requires them to struggle with a concept" - and is a direct statement based on your comment because the attitude of "just tell them" is harmful. I personally don't care if people comment on downvotes because I've done it enough to understand why people don't like being attacked when they do take that extra step to try and help - it stops being worth it very quickly. – LinkBerest Jan 13 at 3:13
  • @JGreenwell exactly. And therefore it makes sense that you not only dowvote without providing any reason. Nobody will learn anything that way – Psi Jan 13 at 3:15
  • 1
    @Psi I neither stated whether I commented on downvotes or not (if you really want to go through my decent collection of comments on SO you could find out which side of that philosophy I fall on). I said - "I don't care if others do" because "I get why people don't". Not my place when I at least see their point. However, any statement that a duplicate is hard to understand so we should just give the answer or not correct the person who is giving the answer - is, well, harmful. Hence, let them struggle and learn (same philosophy I have with graduate students reading their first research papers). – LinkBerest Jan 13 at 3:20
11

As the tooltip states, for both answers and questions, downvotes can mean that the post isn't useful. There is nothing about it being correct or being wrong.

One thing I'm sure about (I'm speaking about the second Q&A you linked): neither answer nor question were useful to me. But I would also go so far as to say, that it wasn't useful for anybody.

It was not useful for the OP of the question: you gave him fish instead of teaching him how to fish. If I were the OP, some years down the road, I would probably be thankful for the snarky comment about the need to learn how to debug.

It was not useful for you: You didn't learn anything answering this question and this answer isn't going to help anybody in the long run.

Even more, it was harmful for the community in the long run.

Do you still remember the time, when one scrolled the Google search page until one saw a StackOverflow link and clicked on it, because one knew that is where one would find the answer?

Do you still remember the time, when one didn't have to scroll at all, and the StackOverflow link was the first link?

And do you remember when you started to go back to the Google search page from a StackOverflow link and try another one? You might not remember, but north and Google do.

So what happened? Such questions happened. They are not very high profile but they are many and the probability to hit one of them is very high - and they aren't useful.

Where does that leave us? We again have to scroll for StackOverflow links and then still go back and try the next one.

So why didn't I downvote your answer?

  1. Because it wouldn't help at all: somebody will see the downvote, think "but the answer is correct" and upvote your answer. It would have netted you at least 8 reputation points, which probably will not lead to change of strategy from your side.
  2. Who am I, to throw the first stone?

What I do instead: Trying to resist my urge to answer such questions, which is sadly not always successful.

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