# Vote early, vote often

I would like to echo a post that Scott Morrison made on Tex.SE. I think this is definitely something we are experiencing here on Math.SE. A link to the post here, and the text of the post:

"I'm a moderator from MathOverflow, and this "question" is actually unsolicited advice, based on our experience from the initial launch of MathOverflow.

We should encourage everyone to vote positively as often as possible!

Every Stack Exchange site will eventually end up with a different "base level" of voting --- that is, the expected number of upvotes for a question of a given level of excellence. (This effect occurs because people see a good question, but already with a certain number of votes, and think "oh, I would have upvoted this, but it already has enough".)

It's easy for us to affect this "base level" by encouraging high levels of upvoting now. We're setting the standards, and this really will have an effect.

(On MathOverflow, we were very active about this early on, specifically encouraging all the initial round of users to vote early and often. You can compare statistics, and see that the average vote total for a MathOverflow question is much higher than on any of the other SE 1.0 sites.)

In case it's not obvious: the rationale for wanting this base level to be high is that it provides better positive feedback to good contributors."

Also @Scott thanks!

• There one on stats.SE too. meta.stats.stackexchange.com/questions/314/… – kennytm Aug 16 '10 at 15:58
• Can't agree more. – KalEl Aug 20 '10 at 14:34
• Am I the only one that thinks it is strange that easy questions get a lot of upvotes and harder questions do not? – Jonas Teuwen Nov 26 '10 at 23:28
• I dont think it is strange. I think it is annoying. I think it is sad. I think it is backwards. But not strange. Maybe I am wrong, but I would explain this simply as the fact that simpler questions have a wider audience which understand it, thus more people will think it is "good". What is significant are the questions that appear simple, but are not! I have said it before, and will say it again, one of my favorite questions of all time is the clothing topology question. This question was great for many reasons, but consider the difficulty. Easy to ask, very difficult to answer completely! – BBischof Nov 29 '10 at 5:48
• @JonasTeuwen From one possible perspective, one can understand an easy question much more quickly, which makes for a faster upvote. A harder question usually has more nuance, which requires more careful examination. I admit that I sometimes upvote questions that look like they're appropriate, without validating the mathematics of it all. There is, after all, something to be said for providing a small reward to a user who takes the time to properly write a question (even if the math involved might contain mistakes!) – Emily Dec 18 '12 at 22:15
• Will you upvote a question you don't understand, @JonasTeuwen? :=) – 1LiterTears Jul 14 '13 at 22:13
• Maybe this should be "vote early and often, but ignore trivial questions and in particular, ignore cheap answers to trivial questions". My impression is that the imbalance between the reward for cheap (but fast!) answers to trivial questions and the reward for sophisticated answers to demanding questions gets bigger and bigger. – azimut Nov 17 '13 at 20:10

For this I suggest that people who (can and do) read and understand more advanced questions be a bit more generous in their upvotes. If someone asks a commutative algebra question in a clear way and shows that they've made some kind of attempt to answer it, maybe that's worthy of an upvote even if the question itself is not very interesting to you. And it goes doubly for answers to upper level questions: I think that if you read someone's answer and can see that it is correct, then you are in possession of information that some other people would like to have and you should upvote to convey that information (namely, that the answer is correct).

• +1: By upvoting more advanced questions and answers we are encouraging people to ask those kinds of questions and give those kinds of answers, which hopefully would lead to more such questions and answers on the site. – Mike Spivey Feb 7 '11 at 21:47
• Just upvoting this answer is not enough!! This is very true! – Asaf Karagila Feb 8 '11 at 3:05
• @Pete Great Answer. – user5783 Feb 16 '12 at 6:50
• I was just going to leave this page without upvoting... xD – Jaideep Khare Feb 4 at 18:53

I would like to make a note that

This effect occurs because people see a good question, but already with a certain number of votes, and think "oh, I would have upvoted this, but it already has enough".

is really true. I've recently seen some questions get the first $10$ or so upvotes really fast, and then the rate of upvoting decreases drastically. I watched this behaviour more closely on one of my own questions: about $1/3$ vote per view in the first 30 views, $1/100$ votes per view in the next $200$ views.

I don't mind that the expected number of upvotes for a question of a given level of excellence is rather lower here compared to other SE sites, but I think it is unreasonable to not vote when you like a question, just because someone out there liked it first.

I certainly know that will never stop me.

• @BillDubuque But this way you block other good answers which in turn may never get seen. You have >50k reputation, is it so important to get even more? I am sure, that the OP will read your answer whenever it was voted high or no. Imagine situation where everyone would do like that, this is madness! I agree that this is a design flaw, isn't there a way to show the answers with same vote count in random order? – dtldarek Apr 18 '12 at 15:18
• @dtldarek Huh? My point is that, due to the design of the platform, quickest answers receive the highest exposure. So if, as I, one desires to seriously attempt to teach here, then one is forced to rush answers, else they'll receive little exposure. Perhaps you view the the site more as "answer the OP" vs. "teach the community". I strive to do both. I'm not here to gain "rep" but to teach. Indeed, I've done the same for many years on similar forums (e.g. sci.math newsgroup) which have no rep measures. The one-dimensional SE votes don't provide good measures of anything (except answer speed). – Bill Dubuque Apr 18 '12 at 15:50
• @BillDubuque Alright, I am sorry, I misunderstood your comment. But then, shouldn't we actively work on proposing a better solution for this problem? This would allow the teachers to work on the answers instead of struggling with time and maybe also to decrease the probability that a really good answer would go unappreciated. I admit, I really do view the site more as "answer the OP", because I think those that want to be taught will just ask for it, I do not wish to impose myself on anybody. However, I admire your quest and wish you the best of luck! +1 for your attitude ;-) – dtldarek Apr 18 '12 at 17:34
• Relatedly, it bothers me when questions with large negative scores get "sympathy upvotes". Because the benefit of upvotes is much greater than the punishment for downvotes, that can often mean that people get net rep gain from terrible questions. (At the same time, there is a tendency when a post has negative score to view it more critically, so posts can suddenly nosedive in score as well...) – Ben Millwood Jul 15 '12 at 11:31

This is very important, I completely agree that people should be voting more.

Interestingly, Stack Exchange has recently changed the user page to include not only reputation, but also "voters." This shows a list of the top voters for the week, month and year. Hopefully this should encourage more people to vote more often.

Also note that this answers Carl Brannen's question.

Edit: I believe this issue needs to be discussed more as the amount of voting occurring on the site has dropped over the last $6$ months.

• Actually, I think this SE has one of the least problems with voting, moreso than most Stacks, including Stack Overflow. Most helpful answers I see upvoted at least once, and even "easy" questions manage to get 2-3 upvotes. – Uticensis Apr 28 '11 at 20:22
• I agree with the edit, Eric, but I think this merits a whole new thread. – Asaf Karagila Dec 25 '12 at 8:20
• @AsafKaragila: I am planning on creating a new meta thread soon (likely for the start new year), to bring this to the attention of the community once again. – Eric Naslund Dec 25 '12 at 8:21
• Very good. I'll be waiting. – Asaf Karagila Dec 25 '12 at 8:22
• Has there actually been less voting? Or has there just been less voting per question (or per answer), since the number of questions has gone way up? – Gerry Myerson Dec 26 '12 at 7:12
• @GerryMyerson: I have been running some queries. The number of questions has almost doubled, and the number of votes has gone up a lot as well, however votes per question is down. The reason for creating a new thread is so that new community members see it, and adopt the community standard for voting. An unfortunate side effect is that old community members see the same post yet again. – Eric Naslund Dec 26 '12 at 8:10

It might be possible to get some statistics on this by examining the ratio of "total votes" to "days active" or "length of membership".