I believe if users behaved as they would from a game theory perspective (i.e. users compete for reputation), we would find nobody giving any upvotes to anyone else on the site. Let's discuss a fictitious, ruthless, cold-hearted player of the game - Alice.

If Alice answers a question and she reads Bob's answer to that same question, and Alice believes Bob's answer is really quite good, Alice will not vote Bob up - due to the fact that this would likely place Bob's answer above hers.

Alice, while just browsing the site, should have less of a problem upvoting other answers she agrees with, since these do not interfere with her own answers. But why should she do this? Remember - users are competing for reputation. Why should Alice let anyone gain reputation without any benefit to her? She will just leave them be.

The only situation that Alice (the cold-hearted, ruthless competitor) will vote for something is a question which she answered. This brings more attention to the question, which may bring more attention to her answer, which may bring her more reputation.

She even has no reason to upvote a correct answer to any of her own questions, although she may accept them to provide incentive for more answers to future questions.

It cannot be explained why anyone upvotes any answers at all (perhaps a feeling of generosity, and a desire for the site to work?). As it stands, people are rather uncharitable about their votes, which can be seen from many questions that have 50+ views and somewhere between 0 and 1 upvotes on any of its answers.

I'm just trying to point out that site mechanics do not support voting for answers - and if users lose their sense of generosity which is providing the votes currently, the incentive to answer questions may likewise diminish.

Should something be done about it?

  • 6
    I'm not upvoting this... – John Gietzen Jan 2 '10 at 0:13
  • I will, however, upvote all the answers thereto. – womble Jan 2 '10 at 1:28
  • 11
    This "game" is not zero sum, and there is no cost to me if You get a lot of reputation. – dmckee Jan 2 '10 at 2:57
  • 1
    @dmckee: That's not true! Do you know how many calories such a click wastes?! – Time Traveling Bobby Jun 15 '12 at 11:19

12 Answers 12


In theory, there is no difference between practice and theory. In practice, there is.

People are upvoting stuff. I think we're fine, and nothing needs to change.


Reputation is mostly worthless. It's a fun little game to play, but if that's really all you care about, then you're probably doing more harm than good anyway. This is one reason why all suggestions involving systems of monetizing reputation are treated so harshly: the last thing we need is anything that would encourage more users to answer questions they have no interest in apart from gaining rep!

Personally, I often up-vote other answers to questions that I've answered. Heck, if it's good enough, I'll even delete my own answer to draw more attention to the answer that I like... And I'll also down-vote answers that I don't care for, a practice certain other users find distasteful because it can be used to game the system...

if users lose their sense of generosity which is providing the votes currently

...then the whole site is doomed, regardless of what changes are made. Not saying that won't happen eventually, but the only thing that can be done about it is to continue encouraging users who do care about more than whatever number appears next to their name.


If Stack Overflow was a site where various children aged 7 - 9 who were extremely competitive collaborated, maybe this theory would show. I think the people here are smart enough to realize that these sites only work if people legitimately vote up positive things and that it's not a competition.


I agree that we need more question voting.

I just implemented a new gold badge to encourage question voting; this has been in the works for a while.

Voted on 600 questions and more than 25% of votes were for questions

  • 4
    How can we see how many questions we've voted (as opposed to total votes)? – user27414 Jan 2 '10 at 16:35
  • Great, this should help more voting. – Kevin Yap Jan 2 '10 at 23:14
  • 4
    Ok, so I should quit voting on answers until I get this badge? – mmyers Jan 3 '10 at 4:38

SO is not a competition.

  • 4
    I bet you only say that because you're losing. – mmyers Jan 2 '10 at 5:41
  • @mmyers - rep IS a competition on meta, where I'm ahead of you. Nyah nyah. – user27414 Jan 2 '10 at 16:34
  • I hit mostly for average, and I'm winning on that. :P – mmyers Jan 3 '10 at 4:40

You've made proper reasoning, but with wrong assumption.

User are NOT competing for reputation. They are competing for answers. For good answers solving their problems. And the way to get good answers is to encourage experts to provide that answers. And one of the ways to encourage them is to vote up good answers and questions.

By upvoting answer or question any time I find it usefull, I contribute to keeping StackOverflow as the network of experts willing to share their knowledge and help others. So do I answering the questions. And this in result increases my chances of getting good answer to my question, when I need help myself.


There are badges such as Supporter and Civic Duty which reward people who vote.

  • 1
    And "Altruist", which is specifically for upvoting competing answers. – Ernest Friedman-Hill Jun 15 '12 at 13:03

Maybe I'm missing something, but I don't understand why upvoting another answer will interfere with Alice's (using your example) reputation?

One reason for upvoting is being part of the game, having some control over the site.

  • Because then Alice's answer will receive less eyeballs, given that it's lower on the list. – womble Jan 2 '10 at 1:27
  • ups, didn't think of that... I prefer the answers ordered by date. – Carlos Heuberger Jan 3 '10 at 20:59

Should something be done about it?

No, site is fine. I sometimes browse questions, not to answer them but to look for good answers based on the content and comments. Sometimes I'll find answers that don't compile because the answerer was too fast to answer. The system will fix itself. Don't worry.


Game theory is an invalid theory, in that it is false that:

It cannot be explained why anyone upvotes any answers at all

People will upvote if they really like an item. Their values may not be yours, so do not expect any upvoting at all. For example, I can easily imagine a technically good answer getting no votes because it was written poorly. Such answers may confuse more than illuminate.

  • 1
    I wouldn't call game theory in general invalid, just Fragsworth's description of the game at hand. – womble Jan 2 '10 at 1:28
  • Well, I've studied it a minor amount [it is an enormous area so I won't lay claim to much], and from what I have seen, it continually and repeatedly makes the mistake that Fragsworth described. In its favor though, I believe it is used to good effect for certain statistical analyses. I just don't think it is a valid model of psychology. – AnneTheAgile Jan 2 '10 at 1:34
  • 4
    That would be because it isn't a model of psychology. – womble Jan 2 '10 at 2:03
  • Imho a model of psychology is necessary to answer "why" do or don't people do x. I'm glad we agree game theory is not one, but it is confusingly similar, which can lead to people asking "why doesn't game theory explain human action." – AnneTheAgile Jan 2 '10 at 16:18

altruism, n.

  1. Unselfish concern for the welfare of others; selflessness.
  2. Zoology. Instinctive behavior that is detrimental to the individual but favors the survival or spread of that individual's genes, as by benefiting its relatives.
  • Actually, it is rational selfishness that accounts for upvoting. It's very much in my long term interest. – AnneTheAgile Jan 2 '10 at 13:59

Answers are supposed to be upvoted by those, who find it helpful.

Example: I googled solution to a particular problem related to programming, and found a solution at SO. Bingo! Problem solved, I will upvote the answer.

The assumption that someone who answered that question would not like others answers, is irrelevant because those who are answering that question are not supposed to upvote others answers to this question.

Upvoting should be done by those who find that particular answer helpful.

  • 3
    This doesn't really address the question. Also, your quote: "those who are answering that question are not supposed to upvote others answers to this question." isn't true. If another answer is useful, you can still upvote it. Writing an answer doesn't mean you can't recognize other answers as good; sometimes you just have an alternate approach. – JMac Nov 16 '17 at 15:29
  • 1
    In addition to what @JMac wrote, we have the Sportsmanship badge for upvoting other answers to questions you have answered. When I earned the badge, I think I was doing what I was supposed to do. – Joonas Ilmavirta Nov 17 '17 at 7:29

You must log in to answer this question.

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