What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 131 Stack Exchange communities.

Possible Duplicate:
Max downvotes per question/answer

I recently posted a wrong answer. This answer was accepted by the OP (thus making it indelible), and has, at this time, -7 votes. I got tired of losing rep, so I CW'd the answer.

But it got me thinking: what is the point of an answer which has less than -1 votes?

An answer with 0 votes could mean "highly controversial" (many up/down votes canceling), "neutral" (not a bad answer, but not great), or "not seen yet" (recently posted, not evaluated yet.)

But an answer of -1 means: "Wow. This was so bad and unhelpful that it needs to be marked as such. I trust the answerer a little bit less too, making them less reputable on the site." A negative vote-count is pretty bad and embarrassing.

Which leads to the question: what is the point of answers which are so heavily downvoted? It eventually gets really really rude and snippy.

Is it really so bad when someone asks a poorly-conceived question because they misunderstand the website, assignment, or technology they want to use? (Case in point, there is a huge problem that clients complain that their website is "taking too long" to develop - so the vague question "How long should a website take to build?" is actually a very reasonable and relevant problem for a programmer to solve.)

Bad answers are far more dangerous: they propagate bad knowledge. While this answer is vastly unhelpful, -18 votes is only relevant to show that it is worse than a post with -17 votes, but really, once a post is negative, who cares? While I have no idea what the "JS" is in this person's post, half of the answer is pretty correct and reasonable. And I don't even know what the (downvoters'?) coments mean.

The only explanation I can think of is some sort of bandwagon effect. "Aargh, we'll show them! How dare they waste my time with such an inane response?" Its obnoxious and makes us look like twits.

So I can think of two solutions (if anyone else feels this strongly about this! Which is likely nobody, but hey.)

  1. Cap the vote score to -1 or -2 or something. Just so everybody gets the point.

  2. Have increasing cost to the voter for negative downvotes. How about this: if i is the vote count after performing a downvote, and i < 0, then the points lost could be expressed as:

n = abs(i)

points lost = 2^(n-1)

So the first negative vote would give -1 points to the voter. Then -2, -4, -8. So that, by the time the post has a vote count of -4... it stops becoming worthwhile to join that bandwagon. (one could also use fib(n), rather than 2^n).

Alternately, one could increase the rep requirement to downvote for very negative values of i. You need 1000 rep to vote something to -2, but then 1500 to vote it to -3.

Do others feel like this is a problem?

share|improve this question

marked as duplicate by Shog9, random, Lance Roberts, Ladybug Killer, Jeff Atwood Nov 29 '09 at 5:15

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

so how does your proposal deal with the situation where a post gets downvoted (score: -1), then upvoted (score: 0), then downvoted again? the score is now only -1, but it's the 2nd downvote, and if the upvote gets rescinded, the score will now be -2. –  quack quixote Nov 28 '09 at 20:05
The idea would be to calculate the rep at the time of voting. So if the vote "would make the resulting score -1", then no matter how many times the post has flipped between 0 and 1, the intent of the user is to vote the post to -1. So the idea of rescinded votes wouldn't matter - since this targets the intended score (at the "point of clicking") rather than what the eventual score is. –  user3788 Nov 28 '09 at 20:19
Largely a duplicate of meta.stackexchange.com/questions/24826/… as well as meta.stackexchange.com/questions/28737/… This suggestion includes a few alternate ideas for discouraging down-voting, however I think the answers and comments on the two earlier suggestions do a pretty good job of making it clear that this isn't really a problem at all. –  Shog9 Nov 28 '09 at 21:29
JS = Javascript –  Greg Nov 28 '09 at 22:21
(In other words, the author of that answer was making wild guesses) –  Shog9 Nov 28 '09 at 23:15
@Shog9: Wow - I had no idea those other questions existed! Had I known, I would have certainly posted my thoughts there instead! –  user3788 Nov 29 '09 at 19:11

4 Answers 4

If you don't like the negative votes, instead of saying "oh well, it's wrong, sorry", edit the answer into something right.

share|improve this answer
I'm visiting this (outdated?) post because its related to one i posted recently. And whats interesting to me is how much stack users insist and demand that either down-voting is NEVER wrong, or something like "O just change it and the benevolent society of stack will fix the downvoting". I've found the stack community to be very generous with help and scrooge-stingy about credit. I believe the average stack power-user doesn't want "Noobs" to get credit. –  Dave A Oct 11 '12 at 4:23
@DaveA, don't look for conspiracy where there is only laziness. By the time the 'noob' gets around to fixing their mistakes, the down voters will have lost interest and not notice the change. You don't get 64k rep by watching the same old questions over and over. –  Paul Tomblin Oct 11 '12 at 11:17
not so much conspiracy as the loss of interest you're talking about. Seems people are more interested in hazing the noob than educating them. That's what drive-by down-voting tells me. "I want to pee on you, but not to explain why". The lack of interest in aftermath is part of it. –  Dave A Oct 11 '12 at 19:37

(Note: my C/C++ isn't strong enough to talk about the merit of your specific answer. I will assume that the downvoters are correct and that your answer is incorrect and misleading. Obviously downvotes for a correct answer are always bad, but occasionally happen.)

What's the point of downvoting something heavily? To show that it is categorically wrong. That's particularly important when the answer is accepted - an answer which has just -1 and is accepted may have been downvoted by a couple of people who misunderstood it, but it was actually fine.

If something has been voted down to -7, that's a clear indicator that the answer is wrong and should not be trusted in any way, shape or form, unless it's got an edit to say something along the lines of "Yes, I got it wrong - here's the right answer." (I usually get rid of the incorrect text at that point to avoid misleading people.)

I have absolutely no problem with something being downvoted a lot, any more than I have a problem with something being upvoted a lot.

As for being worried about losing reputation through it - the answer was accepted. Even if you only got 7 downvotes, you still gained a point of reputation in total despite posting something so wrong that it was downvoted that heavily. How can you feel hard done by?

Personally I would suggest editing the post to wipe out all the previous text, just to say "This answer was incorrect. I will delete it when it's been unaccepted." Then add a comment to the question requesting that the original posted unaccept it. When that's happened, you can delete it. (I've been through a similar situation, although my incorrect answer was upvoted instead of downvoted. When I realised my mistake I made it correct, but still requested that it be unaccepted in favour of another correct answer, so I could delete it.)

share|improve this answer
"I have absolutely no problem with something being downvoted a lot, any more than I have a problem with something being upvoted a lot." - that. –  Shog9 Nov 28 '09 at 21:18
Seems to me as thought people are missing or avoiding the point. I'm coming here 3 years later. But I'm a recent (heavy) user of stack and I've found the stack community to be very helpful but deeply disparaging, even hazing. I've had answers accepted and down-voted by some anonymous visitor. The fact that the question poster and another person trying to help "fixed" the problem by up-voting doesn't change the impression I get frequently that while I'm welcome to visit and get answers, I'm not soooo welcome to help others. –  Dave A Oct 11 '12 at 4:28
@DaveA: Yes, comment-free downvotes happen sometimes... but really not that often, in my experience. –  Jon Skeet Oct 11 '12 at 6:12
@JonSkeet, you've been involved in stack far longer so I lean toward deferring to your experience. On the other hand, most of the down-votes I've noticed have been anonymous. Anonymous isn't necessarily bad, but often seems "irresponsible". I compare them to drive-by shootings. –  Dave A Oct 11 '12 at 19:40
A downvote on Stack Overflow is comparable to (attempted) murder? Words fail me. –  womble Oct 13 '12 at 6:41

You did get a minor technicality wrong. Most C++ compilers define NULL like:

#define NULL 0

So the compiler doesn't know it's a pointer.

The downvote explanations don't strike me as particularly enlightened:

Notice that your linked site says "NULL is defined as a null pointer constant" - not as a null pointer!

So a "null pointer constant" is not a "null pointer" ?

The next comment (with 5 "great comment" votes) reads:

In C++, NULL will not be (for example) "(void*)0" -- that's clearly not allowed.

But Microsoft's Visual C++ compiler thinks it's OK for NULL to be a (void*):

NULL == (void*)0   // True
void* p = NULL;    // Compiles

So your answer is not exactly right, but still pretty informative and helpful. Unlike the technically correct answers, you explained the issue in a way that the original poster understood. So you got 15 rep for the accepted answer. Looks like you actually lost rep by marking this as community wiki :)

share|improve this answer

All the big negative downvote means is that N more people (for a -N score) thought "This answer is not useful" than thought "This answer is useful". If you would like to avoid getting downvoted into oblivion, I would imagine that not posting answers that aren't useful would be the best course of action.

share|improve this answer
what about the useful answers that get down-voted? –  Dave A Oct 11 '12 at 4:30
What about them? –  womble Oct 13 '12 at 6:40
so those people get discouraged and stop posting. And some worthwhile answers will get passed over as a result. Neither is good for SO. But as much as I value SO, I see its not my concern. –  Dave A Oct 20 '12 at 1:32
If an answer got downvoted, it was by definition not useful to at least as many people as there are downvotes (assuming rational voters). –  womble Oct 26 '12 at 0:04
cause every down-votes is for good reason? I see this logic from so many users with high reps. I've concluded that many ENJOY down-voting. Proving my point. –  Dave A Oct 26 '12 at 17:46
Your conclusion is not proof. A downvote is someone expressing their opinion that a particular question or answer was not helpful. That is the "good reason" for the downvote. It's also the definition of a downvote. You're free to posit irrational voters if you like, but the problem you need to solve then is irrationality, which will extend far beyond downvotes. –  womble Oct 29 '12 at 23:05

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