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.

I'm relatively new to stack overflow, but I've noticed one interesting element to the way the UI works.

Often, the earliest answer that is somewhat correct begins to be voted up by the community on frequently viewed questions. Later answers don't receive as much scrutiny because they fall below the fold and further down the users's scan path. It seems that the earliest questions gain momentum as reviewers can see the ranking and tend to push already moving answers up in the list.


This behavior may be corrected by the questioner selecting an answer, but I'm curious if there are any ideas for better ways to arrange the voting UI so that a couple of quick, partial answers don't push down a better answer, preventing it from ever climbing to the top of the list.

Thanks for your thoughts.

share|improve this question
    
See: meta.stackexchange.com/questions/9731/… –  Shog9 Aug 14 '09 at 16:36
    
No, it encourages quick poorly thought out questions. Old questions (that often have low quality or wrong answers) get buried under the torrent of new dumb questions. The recent cap on edits (5 max) is only going to make it worse. Apparently, somebody forgot to consider that quantity != quantity. Keep in mind that this comment comes from the guy who has 8 necromancer badges. I know a little something about providing a valuable answer long after a question's prime. –  Evan Plaice Oct 4 '12 at 1:25

5 Answers 5

up vote 6 down vote accepted

The UI is in favor of quick answers on purpose. SO is meant to be a site to get answers quickly, which is what separates it from the crowd. Sometimes this means that answers are way too simplistic to be a complete answer, but at the same time they usually provide something for the OP to work with, right away. A lot of the time, this quick response won't provide enough information to solve the problem from start to finish, but will give the OP enough of a boost to get them moving in the correct direction and solve the rest on their own.

Think of it as an endergonic reaction:

alt text

The initial energy barrier is the most important part. Once you can get past that, it's smooth sailing to the end.

There are cases where this initial energy won't be enough to overcome the transition state, which is when another user will come along and catalyze the process by providing a comprehensive answer, which will ensure that the problem is overcome. These answers don't get the initial rush of votes, but they get their own love by way of future readers and the OP -- who has the power of marking it as accepted (25 rep with up vote!).

It also seems that the most correct answer usually prevails in the end anyways, allowing SO to not only provide quick help to those who need it, but also serve as a reference for the complete solution of particularly tricky concepts.

share|improve this answer
    
M O O N - That spells transition state! +1, awesome answer. –  Tim Post Jun 21 '10 at 17:51

I don't know about stackoverflow as much as serverfault (which is smaller). But on serverfault this may happen a little, but only if they are roughly close for the most part, pretty old, or unpopular.

For an example of what Ian was saying, look at Evan Anderson's answers. He often answers slower then everyone, but gets voted up right up past them because the answer is just so much better.

share|improve this answer
    
I'm quick to the game sometimes, too, as evidenced by the "Grr-- Evan Anderson beat me to this one" comments that sometimes come up. Sometimes I have to answer verbosely because the topic warrants it. Sometimes I answer quickly, wait to get a couple votes (so my answer will stay high on the page) and drop on an edit that makes the answer much more verbose. I usually won't drop on an edit unless my answer really needs the edit to be correct or I've got enough of a vote margin that I won't drop below a bunch of other answers when I apply the edit. (I really overthink this game, eh?) –  Evan Anderson Aug 25 '09 at 4:31

The way things are set up may encourage fast answers, but the voting mechanisms encourages GOOD FAST answers, not BAD FAST answers. If you have a good fast answer, you just helped out the original author even more than an answer that was slow to make it. If you give a bad answer quickly, that just gives you even more time to get voted down to oblivion.

It should also be noted that you have a 5 minute edit window to stake your claim on a spot and then get a revised answer in there without effecting your position, but it should also be noted that that is also five minutes of time in which people can downvote a crap answer.

The door swings both ways and it can reward people who are both fast and correct as well as those who put lots of detail into their posts. More often than not the best answer will find it's way to the top.

share|improve this answer

I think at the beginning SO was more about helping others, now it seems there are "professional" bounty hunter out there that go for a quick gain and don't bother with more difficult questions.

share|improve this answer
    
...and I think they're going to down vote me to death! –  Khash Jun 21 '10 at 12:23
1  
Got any actual data for this? I have seen some people who started with the "bounty hunter" attitude but they all got corrected pretty quickly by snarky comments and downvotes, and either stopped contributing or changed their ways. –  Pëkka Jun 21 '10 at 12:24
    
can you provide links / data to support this assertion? –  Jeff Atwood Jun 21 '10 at 12:24
    
Gathering data to support this claim in an empirical way is going to take a good researcher a while. I am speaking based on my own anecdotes. I have seen comments like "with 30% accept rate [of the questioner] where is the incentive to answer?" which are logical but against the spirit of being helpful. The tokens are virtual after all. The point based system works because of human nature. But following that logic it shouldn't be surprising to see another animal nature kick in: Maximising gain by reducing investment. You can't just work with incentives partially. –  Khash Jun 21 '10 at 21:44

Fast answers are rewarded more than I think they should be, although fast answers that are bad seem to only get upvoted for a little. (I like leaving a comment why the answer is wrong, and then downvoting it.)

I don't think that answering the question within ten minutes is all that less valuable than answering within one. In some cases, where the answer is an obvious one that the questioner just blanked out on, it is. In most cases, the questioner will be better off waiting for a while and seeing what the community has voted up.

A five or ten minute delay between the time the question is posted and either any answers are posted or any votes will be accepted might improve the long-term usefulness of the question.

share|improve this answer

You must log in to answer this question.

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