Ok, now I admit I haven't been here very long, but I've been seeing a trend amongst the users here: they love to answer questions! This is a good thing, but I think this phenomenon is a little out of hand. For example, look at either of these:

I don't want to "call out" any of the authors for questionable motivation, but clearly the authors of these answers did not take the time to fully read and comprehend the respective question. Furthermore, I've seen a number of answers beginning with phrases such as, "Now, I'm not familiar with specific language whose feature is in question, but you might try code snippet (often syntactically incorrect)." Beyond lacking usefulness, these answers waste space and may work to confuse the askers.

So my questions are as follows:

  • Why does this occur? Is this a symptom similar to Slashdot's "first post" disorder?
  • Is there a good method to programatically detect such a response? (For example, Exhibit 2 talks of "listing files" in a gzipped file while the question mentions only "network streams" and "zlib" compression.*)
  • A down-vote moves the answer further from the question, and subtracts 2 reputation points. Is this sufficient? Should this reputation penalty scale to the users reputation level? (For example, Exhibit 1's author has 13k reputation; 2 reputation is nothing to him)

(*I recognize that gzip uses zlib compression. My understanding is that gzip is only mentioned when talking of files on disk, not when discussing network-or-other streams.)

  • I thought a down-vote takes 10 and costs 1? (I use the feature liberally ...)
    – cdonner
    Mar 8, 2009 at 0:20
  • 1
    It takes 2 and costs 1. Mar 8, 2009 at 0:25
  • The point being that it takes 5 times as many downvotes to cancel out an upvote - the theory is that this should discourage vendettas Mar 8, 2009 at 0:33
  • Don't think of the -2 as "punishment". Think of it as a partial offset of the +10 that an upvote would/did provide. The net effect is usually that it's worth posting an answer if anyone will find it useful.
    – Shog9
    Mar 8, 2009 at 0:34

6 Answers 6


Why does this occur?

If you read SO long enough, I think you'll find the "fastest gun in the West" problem discussed over and over. The first correct answer tends to get voted up a lot. Also, people here are eager to be helpful and sometimes they post too quickly without fully understanding the question. It happens. I don't think it's that big of a deal. You could always downvote such an answer to get the poster 1/3 of the way to a Peer Pressure badge. ;-)

Can you programmatically detect these answers?

I doubt it. But that's what the voting buttons are for anyway. The community voting system does a much better job than any automated program could. And even if we had a bot for that, it would only get one vote!

Is -2 rep per downvote enough?

I can't speak for everyone else, but to me it's the downvote itself that stings, not the -2 rep. Avoiding downvotes on principle is enough motivation for me to be careful when writing answers.

  • +1 for the downvote itself being the important bit. If I get a downvote with no comment, I always add a comment myself asking for the reason - even if it's not actually going to affect my rep.
    – Jon Skeet
    Mar 8, 2009 at 0:26

Exhibit 1's author has 13k reputation; 2 reputation is nothing to him

You might be surprised about that - a lot of the power users track their reputation carefully and are acutely aware of every downvote. So if you think an answer is wrong or unhelpful, don't take the poster's rep into account when deciding to downvote (unless it's 1, in which case it doesn't go any lower, so don't bother).

  • Even if the user's rep is at 1, I'd still down-vote. Its not about the rep, its about moving bad content to the bottom and letting good content move up.
    – AnonJr
    Mar 3, 2010 at 17:16

Sometimes I read or respond too quickly. I've learned to curb myself.

Hey, sometimes people simply make a mistake. Misunderstandings happen. Vote them down if you think it's justified.

Honestly, it's just a programming forum. Breathe deep - don't take it so seriously. If you want something to fret over, go read the economic news and start thinking about a solution for that. We'll all have much bigger problems if hyper-inflation sparks up than whether or not a code snippet was wrong.

Are they any more of a waste of space than this question? I think not. I'm sure that this opinion will earn me some down votes, but that's okay with me.

  • If you dislike the question so much, then why did you answer it? I find it hypocritical when someone answers a question AND votes to close it. Mar 8, 2009 at 0:31
  • Because I felt like responding, that's why.
    – duffymo
    Mar 8, 2009 at 0:41
  • it's not hyperinflation you should immediately fear, but deflation and permanent exchange rate fluctuations through quantitative easing. We are all being burdened.
    – Simon
    Mar 8, 2009 at 0:57
  • "immediately" - yes, not today. But when trillions of dollars are released into the wild, there's no telling where they'll go.
    – duffymo
    Mar 8, 2009 at 3:27

As for programatically detect such response, I doubt it is possible. Also, 13k reputation users are expected NOT to provide poor replies.

This occurs because people like to answer and help, or are seeking fast reputation, but lack the knowledge needed to answer properly.


I'm satisfied with the site's self-policing nature. Top-ranked users here are careful, just look at the last 10 answers of each of the top ten users. I think you'll find all their answers relevant and well articulated.

IMHO, most of the noise comes from non-programming related questions. Those get closed down quickly (too quickly by some accounts ;) Careless answers get also downvoted out of sight long before they endanger the signal-to-noise.


Don't forget that downvoting an answer does more than just take reputation away and move the answer further down the list, it also changes how other users view that answer. If bad answers are being posted they will be downvoted and ignored. I don't think we need much beyond that.

  • We could use comments by the downvoter saying what's wrong with the answer. Mar 7, 2010 at 0:26
  • Or you could just tell me what's wrong with it.
    – Shane
    Mar 8, 2010 at 14:37

You must log in to answer this question.