Is posting an incomplete answer then immediately improving it helpful or aggravating?

Question

Recently I have been posting an answer that is not yet complete and then continuing to make additions over the next few minutes. Undeniably I enjoy getting first post, but there is another reason that I have been doing this: to get some ink on the page so that others know what is coming.

I personally dislike putting time into an answer only to see a longer and better one of the same method pop up before I post. I would rather that the person somehow inform me that an answer of type XYZ in imminent so that I would not start or continue working on mine.

I think that a "stub" answer that actually contains meaningful content, and not just a comment or a "me first" placeholder, is reasonable and potentially helpful for the reason above. On the other hand if this practice is too much gamesmanship or simply annoys people I shall cease doing it.

What does the community think of this?

Answer 1

Please do not take my answer as a critique of any regulars here

The Mathematica site has a problem that is a good problem to have: we have some really expert, really dedicated users who can answer a wide range of questions very well. And these experts live right around the globe in many timezones, so they can answer just about any time. This is very good for people with questions.

But it has a side effect that some might not see as so positive: it is hard for less experienced users (or those of us who can't -- or shouldn't -- devote much time to the site) to find a question they can answer, that hasn't already been answered really well. This will be a particular issue for those who learn by trying to find out the answer for themselves: speaking for myself, I am a much better Mathematica programmer since I started participating on this site and its StackOverflow tag predecessor.

I have personally posted answers and then revised them quickly. Usually it's because I thought of something else to add rather than deliberately putting a stake in the ground. But the knowledge that there are fast-posting experts has definitely often added to my haste in clicking the Post Your Answer button.

This is an inevitable issue as the site grows and attracts new people who would like to be regulars. I think the fact that mid-range-reputation users have posted answers to this question has shown how keen some relatively new users are. The regulars in the SO mathematica tag need to be mindful that we are growing a community here, and new members want to feel that they are contributing. This does not mean that high-rep users should hold off from posting, but breathing for five minutes every so often won't hurt.

Some final, short observations designed to encourage people to post answers anyway.

• The existence of long and highly upvoted answers has not stopped some newer users from posting a subsequent answer to the same question (recent example). Indeed, sometimes it is worth doing so even if there is an accepted answer already.
• For the simple questions, there is often more than one way to do things, so more answers are fine.
• StackOverflow is actually designed around getting multiple answers. (Area 51 recommends an average of at least 2.5 per question.)
• Even if your answer isn't accepted, it will probably still attract some upvotes.
• Posting a duplicate inadvertently isn't a crime, and the later one might actually be explained better. You can always delete afterwards if necessary.
• You can gain reputation by posting some really awesome questions, too.

Conclusion: while there is no harm in posting a partial answer and then editing, there is no great benefit to everyone else in doing so. The rest of us should remember that we shouldn't hold off from posting answers even if someone else has already done so, unless they really are duplicates.

Addendum: I just wanted to note an important point exemplified by this recent post. b.gatessucks's answer was first, correct and accepted. But by providing a more extensive answer that explained in more detail why things work that way, my later answer currently has a couple more votes and the OP found it very useful. Personally, this is my favorite kind of answer to write. The question was quite simple and many of us could quickly come up with a drive-by solution, but I find building understanding to be more rewarding, regardless of the rep points.

Answer 2

I was reluctant to post an answer, so as not to seem whiny. But, personally, I find it aggravating. Ultimately, two main reason people answer questions on any SE site is for the perceived value in reputation or for the good feeling of helping others. But, many are frustrated by the "fastest gun in the west" problem.

See here and here and here and here. And Google a bit, too.

Personally, I generally prefer giving a solution I've tested, rather than a "this should work" solution. My approach tends to take longer, and by the time I've worked out a possible solution, someone else usually beats me to the punch with a quick-and-dirty answer that when the comments come in "Hey, this doesn't quite work" then gets fixed up and is more likely to be chosen as the answer, and thus to be seen by future visitors.

This means then when I see a question posted where I don't instantly know a 2-second answer to, I tend not to bother working on an answer. This inherently isn't a problem from the questioner's perspective, as the rushed answers tend to evolve into excellent solutions, but people need to recognize that the rush to first post has a tendency to put off some potential answerers.

Consider this answer ultimately a neutral position, recognizing the potential off-putting consequences

Answer 3

I downvoted this: I rarely add answers because some are really fast in doing so, leaving not much room for me. I don't mind it, there are much smarter people around here, and I'm not here for collecting points but mostly for work and secondarily for learning new stuff about Mathematica. But I don't think that monopolizing a question (even for just a stub-status) is helpful for the community (see next para). Also I don't think that cross-checking others' intentions is viable.

I do agree with Eli in that seeing that some people are extremely fast in giving answers discourages others, mostly newcomers. Of course, they tend to give weaker answers, but the fact that there are not much deletion/rollback going on on this site is (at least partly, I think) the consequence of weaker posters not posting too much. And it also discourages veteran Mathematicans to add anything more if there is already a fast answer that is also an overkill. Again, I don't mind it at all, as I do try to work inbetween SE sessions. But since the answers-to-question ratio seems rather important for a beta, this issue has a particular importance here that is definitely more than anyone's personal preference/grievance.

Concerning quick and dirty answers: It is a bit like an oversimplified Darwinian struggle. Are you not fast enough? Well, be faster than the one who beat you, or perish and remain in silence. As I'm a biologist, I have nothing against harsh selection, and it definitely improves the site's content. But as inbreeding/elitism increases, diversity decreases.

(Should I remove the comment below the question?)

Answer 4

yoda posted these four links in Chat regarding this question:

[1] [2] [3] [4]

The consensus appears to be that this is acceptable, even desirable. Specifically this answer touches on precisely what I described, and it has +75/-1 votes. Of course consensus there and here may not be the same, and consensus itself is not everything.

Consider this answer the pro-stub-post position and vote accordingly.