These are a few of the reasons why I have difficulty in participating in the StackOverflow community. I was once a very active user, but due to these reasons, and a few unstated, I am unable to participate in the community anymore. 

  1. The Eternal September Issue. Many new users of StackOverflow [SO] rarely ever follow the guidelines of the community. I’m not sure how to solve this, but it is annoying to see questions posted as a plea for help. Stackoverflow moderates its self as a very terse question and answer site. It’s not a discussion forum. [This is a crutch and a gift] Another issue with this is that duplicates show up despite the crotchety moderators complaining about it. One example of this is questions asking about where to find free stock quote data.

  2. Questions that deal with software development that are not inherently technical are frequently down voted/closed. One example of this is that a question asking about a specific data set [for training/development purposes] was closed for the reason that it “didn’t fit with the community.” Until there is a dataset related StackExchange site that the question can be moved to, that was a perfectly acceptable development related question. By a “StackExchange site that the question can be moved to” I mean a fully functioning site, one not currently being “developed”/on Area51. Closing the question as “not relevant” does not help the author, nor does it help people that are looking for a similar dataset.

  3. Down voting as a means of closing a question. To vote-to-close a question a user must have 250 reputation points. This does not take very long to get, if you participate in the community. Down voting should be a way of saying, this is either wrong information, misleading, or not helpful. If the user believes that the question should be closed, but does not have the reputation needed to close, then make a comment and give evidence on why you believe that this should be the case.

  4. The down voting of correct, but not exact-answers. This gets to be an issue when there are questions that can have multiple answers. For an example: A question about optimization of a Java application, the common answer will be to use a more efficient algorithm. That answer will probably be the most voted answer. However, another valid answer is that the process could be rewritten in a more low-level language and connected to via a pipe [Socket, inproc, JNI, etc] to the main application. The latter is better suited for rather unique situations, but it is still a valid and workable answer. From my experience, the second answer will be downvoted, despite that it gave correct information. I have had a discussion with a moderator about this [Shog9] and, according to him, that strategy is a perfectly acceptable strategy to down-vote a creditable answer.

  5. Timing/Duplicate answers. Since I have stopped participating in the community, I have not seen this as often. However, when there is a question posted there is frequently a rush of answers to be posted there. After some time, someone will post a duplicate answer, and get it to be voted on more than the original answer. This is harder to find, but it does happen and it’s very frustrating.

  6. Misattributing credit on answers: If there are other answers that assist you in answering the question, please cite those other authors. It is polite and it gives their question credibility. [Also those that you cite, you should vote for their answer as well].

  7. This is another one of the odd cases on StackOverflow. A few of the “Exact Duplicate” questions are not duplicates due to minor, but important, differences. I cannot come up with an example now, but commenters, frequently, are quick to claim that it is an exact duplicate without verifying the claim. Sometimes the accusation isn’t backed up by evidence. Links to the other questions are sufficient evidence.

  8. The value of reputation: After the global recalculation, the site’s creators made a bold statement that participation is not valued on the site. The recalculation devalued questions, and the new policy was applied retroactively. This caused a loss of reputation. The creators of the site made a claim that “reputation was useless,” which falls contrary to their claim that reputation is a “value of how much the community trusts you.” Operating on the prior claim would make a statement that a user with 500 reputation points is as valued as Jon Skeet. [A well known user in the community, and an author of many technical books]

EDIT: Correction: Vote to close requires 3k in reputation, the ability to downvote requires a reputation of 125 or better. I still stand by my statement that downvoting