Mine is simple. I up vote any answer I think will be of value to the OP and to others with a similar problem. If I think the answer will only be of use to the OP (this usually stems from the question being very localized), I will up vote the answer when the answer takes a broad view of the question and goes sufficiently beyond the local needs of the OP to have archival value. I always feel sad when such an answer is lost because the question gets closed.

To be fair to all who answer a question, my policy requires me to revisit a question whenever a new answer has been added to up vote again if another worthy answer has been added. I accept this responsibility.

I'm pretty sure some others have different policies. I don't claim mine is best. In fact, I suspect it might be too simplistic. I'd like to see other policies expressed. One might be so much better than mine that I will replace my current policy with it.

Do you have a different policy? If so please write it up as an answer.


I already discussed my criterion for rating a question as good in this answer to Mr.Wizard's question. That set my question up-voting policy because a good question is worth an up vote.

  • 7
    Damn! I upvote anything I like and downvote anything I consider harmful. Now I feel in the stone-vote-age. Jul 28, 2014 at 3:39
  • @belisarius Is someone who is about 5k of rep in front of me a thread which I should consider harmful in your kind of definition? ;-)
    – halirutan Mod
    Jul 28, 2014 at 16:47
  • @halirutan That's the only possible definition :) Jul 29, 2014 at 0:34
  • @belisarius I thought all frogs == harmful was also valid in your book? (coincidentally, I also meet your other definition)
    – rm -rf Mod
    Jul 29, 2014 at 13:19
  • @rm-rf Qualifying frogs as harmful is called an eulogy Jul 29, 2014 at 15:14
  • 1
    @belisarius ah, old testament voting. Count me in. I think there was talk of rain of frogs there as well.
    – Yves Klett
    Aug 1, 2014 at 5:42
  • 2
    N.B. closed questions aren't normally deleted, so the answers are not lost. Only if they are downvoted and have no upvoted answers, or if they're not voted for at all and have no answers whatsoever. Aug 5, 2014 at 22:26

8 Answers 8


My policy is simple too: Try to act, always. It is important that this doesn't automatically means an up-vote. It includes down-votes, close-votes, reopen-votes or comments. The idea behind this is simple, if I take the time click on a post to read a question, I have already spend some of my time. Usually, I do this because I find the question interesting or I see that I probably can answer it easily. Anyway, in my opinion, leaving this page without any action, will turn our community into an ignorant crowd I see so often on other SE sites.

My specific policy is the following:

  • When I answer a question, I up-vote the question in 90% of the time. There are corner cases, but in general a question which interests me enough that I write an answer, deserves an up-vote from me. The most annoying posts are, where people have written several answers (with up-votes) and the question has only one or two up-votes. Look at this post: 6 answers, 17 up-votes in total and no one voted for the question? Either the question is really bad, then it deserves to be down-voted or closed, or it is a valid question another beginner probably will stumble upon sooner or later, then there should be at least some up-votes.

  • If a question has many answers, I try to read through all of them and give an up-vote to each that has value for later visitors in it. Sometimes, I start reading from the bottom, just to see whether the current order of answers is really due to quality or due to other SE effects like fastest gun in the west. In my opinion the difference of votes between answers is often not fair. Take for instance this post, where I gave the accepted answer. Don't you think that the other answer deserves some more up-votes and the the difference of currently 9 votes is only because many people just up-vote the answer at the top? What if I tell you that @librik works for Wolfram and has an awesome knowledge about low-level I/O? Go ahead, read his answer (and maybe look over another one of his posts) and then come back and rate the quality of the information he has given again.

  • If a question is not well formatted or lacks of important information, then I usually vote to close and comment. If I think the question has no value whatsoever, then I additionally down-vote, because questions with a score of -4 or lower are no longer shown on the main-page. In the case of answers, I down-vote only if the answer contains wrong information or it really doesn't add any value to the issue. This case is really rare.

Finally, I don't think that everyone's specific policy is that important. It is fundamentally important that you vote. Otherwise, and here I'm pretty sure, our main site will soon look like the main site of this community and this is not what we have worked our asses off to get this site going.

  • 3
    I, too, up vote a large percentage of the questions I answer for the reason you gave. I didn't mention that because I am primarily interested in criteria for up-voting and wanted to keep my question focused. But I'm happy to have an answer more general than I ask for. Thank you very much for your well-reasoned response.
    – m_goldberg
    Jul 27, 2014 at 22:22

I suppose I should answer this because:

  • According to the records I have cast more votes in total than anyone else

  • I am also one of the "very high rep members" that m_goldberg refers to

  • As a Community Moderator my conduct should be held to a higher standard
    (i.e. if I'm doing it wrong people should criticize me)

As it happens my policy is somewhere between the existing answers and it has changed over time.

When the site was young I voted for every well written question and every useful answer that I had the time to read and the ability to comprehend. I felt at that time that it was important to keep activity and morale high and I hoped that my action supported that.

Quite soon however I noticed that vote totals often bore no resemblance to my subjective evaluation of posts: Why do excellent answers languish with few votes and what can be done to combat it? That made me consider the weight of an individual vote and the effect of vote dilution. I became hesitant to up-vote every Answer to a particular question as I realized this does nothing to rank or differentiate them.


  • On Questions of interest I skim the answers and vote for the one(s) I feel to be the best.

  • I vote for any answer that teaches me something significant and applicable. Contrary to m_goldberg's assertion this happens with great regularity. Admittedly I would never claim to be one of the most knowledgeable people here, nevertheless I do not imagine anyone knows everything there is to know about Mathematica, not even Wolfram himself.

  • When it seems appropriate I vote to reward effort. One example: sometimes its clear that a user has put a lot of work into an answer with little expectation of accolade; I try to support this beneficence by up-voting.

  • Like Simon Woods I am a bit capricious in my voting. Subconsciously I follow these guidelines, and a few others as well, but certainly I am not consistent in the way I vote all the time. I don't go though a lengthy internal justification process before I click the arrow.

  • In all cases I am volitionally biased toward the first correct answer using a particular method. I feel that this is important to not diminish original contribution. I may vote for a derivative answer if it is sufficiently different or superior but I vote for the older one as well.

A note: I have been quite sparing in my use of the down-vote as I know from my early experience on Stack Overflow that it can be quite discouraging but I am starting to use this tool more often as I increasingly recognize its utility.


I don't really have a firm "policy" - I have a tendency towards indecisiveness and I'd spend far too much time pondering whether or not an answer met my criteria (and whether or not my criteria needed revising!).

So I vote fast and often, without overthinking it. If I read an answer and think "yeah, good answer", I upvote. It might be good because I learnt something, or because it contains a great explanation of a tricky concept, or a lot of time and effort went into it, or just because it includes some nice elegant bit of code.


I wasn't around when this was asked, but I feel I should also be putting in my take.

My upvoting criterion is capriciously arbitrary: if I "enjoyed" reading the answer, or learned something genuinely new, or it is an answer that I would have written myself (also counting the case where it was written in a far better style than I would've done), only then do I upvote the answer. It is the exact same voting criterion I use on other SE sites, and I like to think it has served me well.

My downvoting criterion for answers is way more strict: I will only downvote if an answer is actively wrong/harmful, and the answerer has proven uncooperative in nudges (through comments) to fix his/her answer.


My approach is subjective. I upvote answers from which I learn something new and unexpected. I sometimes downvote an answer when I have the impression that it too heavily relies on comments or answers already given. Sometimes I downvote questions if the OP doesn't participate ("Do my homework but don't bother me with questions or comments"). Sometimes I regret that one cannot upvote comments.

  • 5
    Comments can be upvoted -- no rep accrues, though. See m_goldberg's comment to halirutan's answer. (2 upvotes right now.)
    – Michael E2
    Jul 30, 2014 at 17:07
  • @MichaelE2 Thanks, nice feature :)
    – eldo
    Jul 30, 2014 at 17:24
  • 1
    It seems you are saying that you do not up vote a question that has tutorial value for others but not you personally. I think such a policy is short-sighted because I strongly believe, when we are reading an answer, we should consider how it might have tutorial value for any Mathematica users, no matter what their skill level is. Up voting such answers will encourage more of them and help to build the library of good answers that Mathematica users can consult in the future. Building up that library is an important goal we should always have in mind.
    – m_goldberg
    Jul 31, 2014 at 1:07
  • 9
    Also, consider this: if the very high rep members of the community adhered to your policy, they would almost never up-vote an answer. I have good reason to believe, however, that they up vote many answers for which they personally learn nothing new.
    – m_goldberg
    Jul 31, 2014 at 1:08
  • @m_goldberg I'm relatively new to Mathematica and SE. Hopefully, with growing experience, also objectivity will go up. Your second comment ("high rep members...") makes a very good point.
    – eldo
    Jul 31, 2014 at 8:20

This is a Q & A page.

When I need answers to specific questions, I search with keywords and usually find fair questions but even better answers. If a Q / A element solves my question or brings me a good bit closer to the solution I rate also old solutions positive.

I do upvote "old" responses.

One time "Questioner" I ignore.

A good question is well written, well researched and reduced to the minimum, a "minimal working example" in the proper sense. Such questions are interesting because they show that the questioner is interested on his level of an answer. Such questions I upvote.


I think of my voting policy as very conservative. Every vote must be justified (at least ideally). Nonetheless, there's a lot of good content on the site, so it's not hard for even a conservative voter to have cast a lot of votes in aggregate. My voting is also very subjective, although it has become less so over time (cf. the second bullet points in each category).

Let's presuppose that anything I upvote is coherent and correctly formed, at least intellectually. (I normally correct minor formatting or writing problems, if there are any, before voting. This is particularly easy to excuse for newcomers to the site who bring with them fresh ideas.)

A question gets my upvote if

  • it asks about a interesting problem that I would either have to think moderately hard about in order to provide a good answer (which is also the best way to attract my answer), or (for more challenging problems, if I don't have time to answer, and where I have insufficient understanding of the subject to provide an answer of my own) where I expect I could learn something from the approaches taken by other people. Or,

  • where a definitive answer immediately comes to mind, but the question has extraordinary value for the community (e.g. it raises a commonly encountered issue in a particularly clear and incisive way).

An answer gets my upvote when

  • I learn something from it. I very rarely upvote an answer that doesn't teach me anything. Although this doesn't have to be especially profound--in performance tuning questions, when people collect others' answers together and benchmark the codes against each other, that also gets my upvote, because it provides the answer to a natural question and leads to a better understanding of the existing answers. Or,

  • it is a technically impeccable answer from a new user, even if not particularly enlightening to me personally. I do this because everyone needs some encouragement and a certain amount of points in order to be able to use the site fully and feel as if they're a part of the community.

I try never to vote for answers simply on the grounds of correctness. A correct answer must also fulfil one of the other criteria in order to get my upvote.


I only down-vote questions where the OP is argumentative. I up-vote interesting questions, even if poorly phrased. Otherwise, no vote.

However, if a question triggers good answers, I will up-vote for that. And up-vote the answers, of course.

You must log in to answer this question.

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