Which SO (not MSO) behaviors are we most "losing the battle" against?

Question

It occurred to me today that there are quite a few things that we, as human beings, are naturally inclined to do on SO but really "ought" not to for the sake of the site. They're the kind of things that we rely on the documentation, the culture, personal discipline, the incentive systems, user feedback and in some cases moderator action to keep our "basic tendencies" in check.

For most of these things, the system "works", but for some, human nature is collectively getting the better of us. By identifying those things in one place, perhaps we'll be better positioned to address them.

(Note: I realize there are some SO behaviors that are "controversial" in the sense that there is no consensus that they are "bad". I think listing them here is ok, since it's still helpful to be aware of the human tendency.)

Update #1: Not sure if this will sufficiently address the concerns of the "primarily-opinion based" close voters, but answers should point to metrics where possible (e.g. slope of close vote review queue size).

Update #2: For those of you who consider MSO as an extension of SO, this question is about SO exclusive of MSO.

Answer 1

1. Unhelpful commentary
2. Lazy, poorly-asked questions
3. Omphaloskepsis

Answer 2

Firstly, the non constructive commentary. If your comment isn't going to improve the post, don't post it. Stack Overflow is not Reddit; "OP is a fag"-type comments won't roll here. Yes, the comments we get here aren't that bad, but they're still bad.

Secondly, the ignoring of rules. I see a lot of high reps posting link-only answers. A more troubling behavior is answering questions that really should be closed. This behavior reinforces the "dump your code on SO and your problem will be solved" mentality that's out there.

If you have been here for long enough, you should know better than to answer obviously closeworthy posts.

Besides this, I have a suspicion that many >3ks don't even know which posts are closeworthy. It's fine if one doesn't want to take active part in the moderation, but this is worse.

Answer 3

How about random, petty down-votes to "get even" with someone who pointed out a flaw in your answer or begged you for enough information in your question to make it answerable?

I've received a couple today, including one on this nearly 5-year-old answer. I must have really put off some immature person who thinks 2 rep points is going to make me feel sorry for trying to make the site better.

Since I feel that most of my peers have long since realized how futile it is to down-vote people for revenge, I suspect these are more angry childish n00bs than high-rep users, but hopefully they find their way here too.

Over time, I have reduced or eliminated some behaviors that seemed to contribute to more frequent revenge-style occurrences of this (and I say that because (1) the frequency really went down immediately and (2) I don't think it's that my answers have gotten better). One of the main ones: explaining down-votes. Now I typically don't bother, and that's a shame.

I don't care about the rep, but the time I've spent trying to analyze whether there is a legitimate problem with my answer or it's just a random drive-by sour grapes down-vote that I can ignore has gone down as well. But it still drives me crazy trying to figure it out when it does happen - and all that will lead me to is contributing less in other ways, period. Again, that's a shame.

I think the main problem is that we put a very high level of emphasis on reputation when really, at the end of the day, the general thresholds are important to enable privileges, but - in the real world - 2 or 200 or 2000 rep points don't mean jack. None of it means jack. Particularly trying to chip 2 or 4 rep points away from a 100K user. Seriously? If a potential employer is looking for your SO rep on your resume (and cares if it's 5K or 10K or 50K), I don't know what to tell you. Knowing you're a contributor, period, should tell them enough about you. I know some pretty high-rep users who I wouldn't hire if they were the last person on earth, and I know others with low rep that are valued contributors but give away all their rep in bounties, or contribute in non-accelerated-rep-accruing activities, etc.

Answer 4

1. Asking and leaving as is poorly formed questions
2. Answering duplicate questions vs. searching for and V(ing)TC as duplicate
3. Downvoting poorly formed questions vs. commenting and/or V(ing)TC
4. Answering questions in comments (only)

All of these are "understandable" and arguably based of the traits of a great programmer (i.e. laziness, impatience, hubris). I think our challenge is making the "right thing" more attractive and/or the "wrong thing" more repellent.

In some cases this can be done with (dis)incentives, but in other cases it may involve substantively changing how the system works.

In all cases, I think it requires us to focus on what users are experiencing on the site and stop thinking that we can just "encourage" people towards different behavior (e.g. spending more time casting/reviewing close votes).

I didn't list "non-constructive comments" because while I've seen/experienced that behavior in spades on MSO, it seems like it's relatively "in check" on SO, perhaps because I'm not "hanging out" in the tags where this is more of a problem.

Answer 5

Overly broad list style questions in which everyone is encouraged to provide their own opinions about some problem without ever having a conclusive answer.

Answer 6

We're not doing nearly enough of what is in this link.

Answer 7

Straightforward answers without any explanations like code only answers.

Answer 8

While people have mentioned multiple times the presence of lots of poor comments, another related problem is not enough good comments. I try to comment every time I downvote a post or vote to close or just see a poorly formulated answer or question. The questions from a particular person won't improve unless you explicitly tell what's wrong (at least try to). That especially concerns downvoting 1-rep users to -infinity and leaving witty comments instead of linking to rules and shortly explaining why this happens.

Answer 9

Posting full solutions to bad questions, ones that are essentially "I've given up on learning this / too lazy to study documentation. Give me the code please."

These should be downvoted/flagged but are often just answered instead, which encourages people to make more.

I have to assume the answers are hungry for reputation points as it's really not beneficial to the asker, they will not learn from being fed code to paste into their programs.

You see this a lot with linq and regex questions.

Answer 10

phpBB-like questions where the OP expects community to find and fix some trivial mistake in his code, or asking for advice what he should learn/read/use.

Duplicate questions occuring because OP couldn't understand the answer, or was too lazy to understand it, or even worse, too lazy to google before asking. Generally, that chatty (usenetty?) approach, that anything that isn't on main page doesn't exist.

Both above are symptoms of general ignorance towards FAQ from the many new users. They don't care to read it, and they don't feel the need to abide to it.

Answer 11

Theft of content. StackOverflow encourages users to copy/paste into the answers themselves instead of linking offsite. This means code snippets I spent a very long time on are being copyied off my blog (some with and some without attribution). Google really likes StackOverflow and then always returns that answer above my own site where the content originated. This is frustrating to me because I'd like to help share these answers with the community which is why I posted them publicly, but I'd also like to build my reputation as an expert in the field and this is harder to do when my posts answering difficult questions are not viewed as the authoritative subject on the matter despite being the original author.

The answer to this problem without making the StackOverflow experience worse is difficult. Maybe StackOverflow could at least do an automated web search at the time of posting for duplicate content and then link back to the original source when the author has not or suggest to the author that they include attribution in that case?