12

In lots of proposals, you'll see a lot of questions with exactly 10 votes. You'll even see people saying that you're wasting votes if you vote for a question with more than 10 votes.

Most questions having the same number of votes means you can't tell whether question X is regarded as more suitable than question Y.

It also means that people aren't being fully honest with their voting. They won't go "Hey, that's a really good question, +1!", instead they'll go "Ok, what's the best question with fewer than 10 votes?".

I don't know how to fix this problem. Should the metric of "40 questions with 10 upvotes" be changed, or possibly the number of votes people can cast?

  • 11
    "It also means that people aren't being fully honest with their voting." This is partly the fault of the system. Even if I wanted to upvote 6 questions, I can only upvote 5, and so I was forced to be dishonest. – Joel Reyes Noche Feb 9 '14 at 0:46
  • 3
    leave this question at ten votes! – x-x Dec 12 '14 at 1:41
9

The first problem is, that people don't understand that moving to commitment phase too quickly won't help the proposal to be launched.

The second are miscalculated requirements. In theory you need only 60 followers for the proposal, but you need at least 40 questions with at least 10 votes on each. This gives you 400 upvotes (in optimistic scenario, without downvotes), so you need 80 followers for this. So many people are frustrated seeing that the restriction with followers has been long fulfilled and they are still far from 40 top questions.

The solution would be to balance the requirements, increasing the number of required followers, say to 120, or to give more votes to followers, say one vote for every 15 example questions.

  • 3
    Just a small point of correction - you actually don't need 80 followers, according to discuss.area51.stackexchange.com/questions/12248/…, because non-followers can vote on questions. – Daniel Schilling Feb 6 '14 at 15:56
  • 1
    @DanielSchilling I would say it's a very small point because if this site idea is on your field of interest and you really wanted it to take off you'd follow it. How much time do most busy people have to go around checking other people's proposals outside their own interests and responsibilities? There are certainly not enough hours in the day for me to do what I have to do let alone do that. It seems to me the vast majority of progress a site makes will be through its potential followers. – Geoff Pointer Feb 7 '14 at 23:03
  • 1
    In saying "people don't understand that moving to commitment phase too quickly won't help the proposal to be launched ...", I think you're making too many assumptions about what people understand and putting too much faith in established practice. Show me some valid research or at the very least a well reasoned opinion that shows that actively encouraging followers to stick to the current 10 vote minimum will (a) necessarily lead to a bad definition (b) that people don't actually understand your first assertion. – Geoff Pointer Feb 8 '14 at 0:19
6

Let’s have a look at a pathological example: Suppose a proposal has five questions that are outstanding to the extent, that they are preferred by everybody over all other questions. If everybody voted honestly (i.e., upvoted their five favorite questions), only these five questions would ever receive upvotes – no matter how viable the proposal actually is.

Though this case is constructed, it illustrates an inherent problem of the current system: It is only natural that there is a certain variation of popularity (in the honest sense) amongst example questions and the variation in the popularity of a given question amongst different users can only counteract this to a certain extent. As a consequence, if everybody voted honestly, a small number of questions would attract the majority of the votes. Thus a gigantic amount of acceptable example questions would be required in order for enough good but not very good questions to reach ten votes. Moreover, in this case (honest voting), very good questions would actually delay a proposal and it would accelerate a proposal if a question were deleted that turned out to be very popular.

So, while dishonest voting to accelerate a proposal is likely to be deconstructive eventually, purely honest voting is harmful as well and I would even guess that a lot of graduated proposals had never reached the commitment phase, if there had been only honest voting.

Now it is safe to assume, that no system will fully avoid tactical or dishonest voting or voting paradoxes (and for other cases, this has even been proven) and tactical voting is not neccessarily a bad thing, but I think that too much of it happens in the current system. Therefore it would be good to change the system (and not just its parameters) to restore constructiveness for the first phase.

Some ideas (numbers are just examples and other parameters of the system might need adjustment):

  • Change the requirements such that they account for the natural popularity distribution of questions: Instead of requiring 40 questions with 10 votes, require for example 5 questions with 50 votes, 10 questions with 30 votes and 15 questions with 10 votes.
  • For every vote beyond the 13th for a question, give an additional vote to the first voters for this question in order of voting. So, if the question received its 14th upvote, the first upvoter gets an additional vote; the next upvote grants the second upvoter an additional vote; and so on. This way votes on popular questions are not wasted. (This system should be more difficult to exploit than just not counting votes for a question that has already reached 10 votes.)
  • This is certainly a contentious issue. I'm getting negative votes for even daring to discuss it here. You make some good points. I think that what would be helpful would be if the SE powers that be posted some statistics on just how previous proposals have failed because of this spreading the votes issue. – Geoff Pointer 2 secs ago edit – Geoff Pointer Feb 14 '14 at 0:12
  • If there were only five questions on which all users agreed, then it would be almost certain that there were not enough questions to make the site viable. – rem Feb 14 '14 at 15:56
  • @rem: I never talked about a case in which there were only five questions about which all users agreed. I only talked about a case where all users favoured five questions above all others (about which they might still all agree). – Wrzlprmft Feb 14 '14 at 21:42
5

How many votes does a question have to have before it's considered appropriate? This process is about definition, not ordering the best questions. It seems to me, if you can get enough questions with ten votes each, that's already enough momentum to carry forward when the site becomes live.

Also, there are already many questions that seem like quite reasonable questions to me that aren't close to getting ten votes, that the questions that have got ten votes are being distinguished from. So there is already a significant level of discretion being expressed by followers.

I think it's a bit strong and unfair to question people's honesty in this process.

As I've mentioned in my comment to Lukaz L.'s answer, which, I realise, is anecdotal and not a scientifically proven assertion: People spend most of their time following their own interests and responsibilities which I think validates the latter part of his answer. On the topic of honesty, I've had a very brief scan of other proposals on Area51 but I haven't voted for anything outside what I'm following because I don't have the time or the expertise to properly and fairly assess matters outside my areas of interest.

So, in my humble opinion, numbers of followers are by far the main concern of any proposal and I see no harm in having a ten point threshold for deciding whether or not a question is appropriate and as I pointed out above there are plenty of good questions that won't make the ten point threshold.

The focus should be on how many such threshold questions should carry a proposal out of the definition phase. Some of the potentially super high voted questions will already be here and if we have enough followers with enough content more will surely come.

Is this really how the other established SE sites are measured? By their questions with really high numbers of votes?

I don't believe anyone's honesty is in question here. Oh, and by the way, there are still 4 votes going to waste on the Mathematics Learning, Studying, and Education proposal.

  • On the topic of honesty, I did just cast a vote on a question on another Area51 proposal that didn't require any specific expertise and only needed some ordinary life experience. – Geoff Pointer Feb 12 '14 at 1:28

You must log in to answer this question.

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