It seems MO awards +10 for each vote a question gets, as opposed to the +5 we get here. Is this intentional? (Have I made an error?)
Other than this, is the point system the same as on MO? Should it be exactly the same?
$\begingroup$
$\endgroup$
1
-
2$\begingroup$ Right now we have 484 users and it looks like roughly half of them have enough reputation to vote. Hence our best questions should already have > 200 upvotes. However, so far we have only 4 questions with > 20 upvotes, and I don't think that all of our questions are really so bad or marginal that they don't deserve more votes! Whether you get +10 or +5 is completely irrelevant if people don't vote. $\endgroup$– Jukka SuomelaAug 24, 2010 at 15:25
Add a comment
|
3 Answers
$\begingroup$
$\endgroup$
7
Making the question upvote rep be 5 (as opposed to 10, same as the answer upvote rep) was a deliberate decision made by the stackoverflow team.
You can read more about it here: http://blog.stackoverflow.com/2010/03/important-reputation-rule-changes/
-
2$\begingroup$ For a mathematically-oriented site, this is bogus reasoning: "We know that answers have more intrinsic value than questions, and the reputation balance should reflect that." Know? How? If the SO folks want to say "we want shorter questions so the ads don't get pushed onto the second page", they should just come out and say so. But if that is the concern, there are other possibilities, like a Wordpress-like "more" fold which people would be encouraged to insert after the first few sentences of their question. $\endgroup$ Aug 19, 2010 at 2:22
-
6$\begingroup$ I think the logic in the document makes sense. That having been said, I don't really see what's the big deal. Are reputation points (and slight rules regarding them) actually so important? $\endgroup$– Lev Reyzin ModAug 19, 2010 at 4:51
-
$\begingroup$ I'm inclined to agree with Lev on this. $\endgroup$ Aug 19, 2010 at 5:17
-
1$\begingroup$ Well, it's not terribly important, but it isn't unimportant. Tim Gowers once said on MO, "Mathoverflow's rewards and incentive structures are silly in a way, but I think they are a quite important form of silliness." I agree with this. $\endgroup$ Aug 19, 2010 at 5:40
-
$\begingroup$ I guess if it affects the way people behave, it's important whether or not these points really mean anything. In that case, it still seems better to me to place more value on answers than questions. $\endgroup$– Lev Reyzin ModAug 19, 2010 at 15:43
-
$\begingroup$ the discussion on enabling math support indicates that we need large community support to request changes in the scoring mechanism. Right now there appears to be a division on this issue, so we'll need to let things develop a bit. Maybe someone can post a question on meta with two opposing points of view that can be voted on. $\endgroup$ Aug 19, 2010 at 20:55
-
2$\begingroup$ @Lev Increased reputation (supposedly a measure of one's contribution) grants additional powers such as voting to close or delete items, retagging, creating new tags, editing items, and so on. On Stack Overflow, there are pathological cases of users amassing reputations in the thousands with little more than a deluge of puff questions. Reputation is important in that it helps keep signal-to-noise high and vandalism low. $\endgroup$ Aug 24, 2010 at 13:58
$\begingroup$
$\endgroup$
6
That's correct. MO is on the SE 1.0 system, and they have the power to hack their code any way they like. We're on the SE 2.0 system, and I don't believe we have a lot of flexibility in how we change such things. Certain of the thresholds are slightly different, but I think overall things are the same.
I don't quite understand the ramifications of these differences, so am content to let things lie for now. If we feel strongly about this at a later time, we can initiate discussions on meta.stackoverflow eventually.
-
$\begingroup$ This can be changed. Cf. meta.stackoverflow.com, which awards +10 for questions. $\endgroup$ Aug 18, 2010 at 7:04
-
-
$\begingroup$ Since most users are going to be common to MO, it might have been better to have the same scoring system. This way users can compare reputation points across sites. $\endgroup$ Aug 18, 2010 at 15:08
-
2$\begingroup$ I think to encourage more significantly fleshed out questions, it would be good to use +10 as MO does. We now have several nice but brief questions with in-depth answers. This is rational behaviour, but it would make more sense for the background and motivation to be up front in the question, instead of added as part of an answer. $\endgroup$ Aug 19, 2010 at 2:16
-
$\begingroup$ This answer is incorrect. They can not and did NOT hack their system to change the voting patterns. They are running an older version of the software which had different voting weight. See @Moron's answer: meta.cstheory.stackexchange.com/questions/37/… $\endgroup$ Aug 20, 2010 at 3:45
-
$\begingroup$ @Charles Stewart: See blog.stackoverflow.com/2010/03/… : “on Meta, the value of a question upvote will still be +10” I do not know why, but as a matter of fact it is designed that way. I do not think that it is configured on a site-by-site basis. $\endgroup$ Sep 6, 2010 at 2:21
$\begingroup$
$\endgroup$
4
If a great question generates several well-thought-out answers, it deserves a great deal of credit. Otherwise there is little incentive to ask careful, well-thought-out questions, but we will keep seeing all these one-liners that have started to appear as questions. I really dislike this. Many researchers have several hundred questions of that form in their notes, but these are not valuable without some fleshing out.
A prototypical good question should provide some background, motivation for why the question is important, perhaps sketch some consequences, and ideally provide a link to relevant literature. This requires at least as much effort as generating an answer.
If we encourage short questions with little background, we are going to see few novel research-level questions. Instead we will see fashionable questions which we have all seen before, or based on the pattern of "take a pair from the complexity zoo, generate question, rinse&repeat". These can be stated without significant preliminaries.
I would like to see more questions which are significant contributions by themselves, where someone poses a novel connection, or an interesting issue that came up in their work. To encourage this, it would be useful to award at least the same number of points as to answers.
-
$\begingroup$ you could always use downvotes to signal your displeasure. not ideal, but I've started doing that. I agree with your complaint about one-liner questions $\endgroup$ Aug 24, 2010 at 21:35
-
$\begingroup$ True, but I'd like to keep downvotes for truly bad questions and answers. Sometimes the one-liner is a perfectly OK question, and could become a great question with a bit of fleshing out. $\endgroup$ Sep 4, 2010 at 22:49
-
$\begingroup$ Awarding +10 points instead of +5 for an upvote for a question will certainly encourage some users to post more questions, but I am not sure if it works as an incentive to post well-thought-out questions (good) or as an incentive to post many one-liner questions (bad). Apparently, it seems that the SO people made the change from +10 to +5 to discourage asking many quick questions. blog.stackoverflow.com/2010/03/… $\endgroup$ Sep 6, 2010 at 2:29
-
$\begingroup$ Tsuyoshi Ito: that makes sense, thanks for the link. However, I remain sceptical: a timer for questions would seem more useful than using the sledgehammer of weights. Something like: maximum two questions with no upvotes can be asked at one time. $\endgroup$ Sep 6, 2010 at 20:54