10

In the Hebrew language proposal, a few questions with many up-votes have been closed as a duplicate of older questions (with fewer up-votes).

The convention has always been to defer to the question which was asked first. But when the object is to collect the best possible example questions during Definition, sometimes this does not make sense.

After nearly one year of development, we are very close to reaching the necessary number of questions with many up-votes. Losing some of our best questions seems to be a bit problematic to me. I wanted to start a discussion about which questions should be closed as duplicate in situations where the first question isn't necessarily the best for the proposal.

Which duplicate should remain; the first or the best?

5
  • I clarified your question to apply to the broader problem I've been seeing. Feel free to edit if my changes are not on point. Commented Aug 9, 2017 at 22:03
  • If Robert's edits are acceptable, I don't see the point in the question. The duplicates should be targeted towards the best candidate, regardless of its age or vote record.
    – Nij
    Commented Aug 10, 2017 at 5:24
  • @Nij: I agree with you that the duplicates should be targeted towards the best candidate. The reason for the question is that currently age is the only criterion for closing questions. I only mentioned the situation of the Hebrew proposal since currently in my opinion closing only based on age might cause problems for this proposal.
    – Christian
    Commented Aug 10, 2017 at 6:27
  • Age is correlated with quality - an older question is more likely to be edited into good shape and have answers that were good to start with, or also fixed. But there is no age criterion, we have closed several duplicates to the newer question because it was far better, despite repeating some content, and deliberately in some cases as a.canonical question.
    – Nij
    Commented Aug 10, 2017 at 7:31
  • @Nij Answers/arguments one way or the other should be posted and discussed below, otherwise all was have is random thoughts, no consensus, and the status quo will continue. Commented Aug 10, 2017 at 14:31

2 Answers 2

5

On other SE sites we occasionally close an older, weaker question as a duplicate of a newer one that is asked better. If we can do that on sites where we don't have a goal of developing a specific, limited corpus, then surely it should be ok here where our votes and questions are limited to a few per user.

If we continue the current practice we create perverse incentives. Some people will move their votes from the good questions to the weak ones, which rewards and penalizes the wrong users. Others won't move their votes, and the proposal is penalized by reducing the number of qualifying questions.

When there are duplicates, the better question should be the one that counts for advancement, even if it's newer.

A point has been raised that older questions attract more votes just by being older, even if they aren't the better questions. That can be true on other sites where those early votes are locked in, but on Area 51 people can and do move votes. I don't have data, only anecdotes, but I know that I move my limited votes around when better questions come in and I know from my reputation history that others do too. How common it is I can't say, but Area 51 is built to support that. So I think users should indicate the better question through voting and those closing A51 example questions as dupes should take votes into account when deciding on the direction.

2
  • I love your reasoning for questions with similar subjects/clarity/quality. But is there a quality consideration when one is simply better than the other? The bulk of voting often goes to the early bikeshed examples, but sometimes a better example is posted later. In the Final Evaluation, when I'm trying to build the strongest case for getting a site confirmed, picking the best example is sometimes at odds with pushing a proposal through as quickly as possible. I'm just struggling a bit with ignoring a quality call in favor of "the one that counts for [most] advancement". Commented Aug 11, 2017 at 0:23
  • On other sites older votes are locked in, but in Area 51 we can and do move votes - especially because we only get 5 upvotes so need to move and not just add. We can cancel or move votes at any time. So I argue that here age isn't as big a factor. (Feel free to edit that in, or I'll do it when not on mobile.) Commented Aug 11, 2017 at 3:28
0

Votes are representative of age as much as anything else, in almost every Stack Exchange site that exists. Neither aspect is a reasonable gauge for deciding the duplicate target.

Age is correlated with quality - an older question is more likely to be edited into good shape and have answers that were good to start with, or also fixed. But we have closed several duplicates to the newer question because it was far better, despite repeating some content, and deliberately in some cases as a canonical question. Thus age is a weak proxy at best.

The fact that Area 51 requires a number of questions voted to a certain level should not affect this. It defeats the purpose to have the best questions closed simply because they came after the highest-voted questions.

If votes are not being moved when this occurs, it speaks implicit volumes about the long-term engagement of users in the proposal.

You must log in to answer this question.

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