In the tag wiki, there are defaults for scoring if not specified (in bytes). Recently I have seen some questions what to do in case of a tie. My proposal is to make this a default in the tag wiki. The two things I have really seen specified as tie breakers in challenges are:

  1. Earlier post wins (e.g. Draw a Random ASCII Art Quilt, Hodorize a String)
  2. More Upvotes wins (e.g Programming Tetris Blocks (Literally))

I don't think we need tag-specific defaults for this, but I would like to propose the following global default:

If no tie-breakers are specified, or all specified tie-breakers are exhausted, the earliest answer wins. When determining which answer was earlier, the timestamp of the first valid edit which gave the current score counts.

  • \$\begingroup\$ I will accept this answer if no one objects in a few hours. \$\endgroup\$ – user34736 Apr 17 '15 at 15:33
  • 2
    \$\begingroup\$ @Reticality Give it the weekend. Someone might propose a different default. \$\endgroup\$ – Martin Ender Apr 17 '15 at 15:33
  • 1
    \$\begingroup\$ It might be worth mentioning the two benefits of having a default. First, you save characters in the challenge post. Second, it serves as a safeguard against authors with poor foresight (e.g., "I didn't know there would be a tie, and now I am going to add this arbitrary rule to break it."). Perhaps there are other benefits. There is also one downside: defaults have low visibility, almost like a hidden rule that you can only find by digging into the code-golf tag. I think describing how these pros outweight the one con would support the answer more than "it makes sense". \$\endgroup\$ – Rainbolt Apr 17 '15 at 15:41
  • 3
    \$\begingroup\$ @Rainbolt I have removed the assertion that a global default is a good thing in general. That is a much bigger issue (because we've got several other kinds of defaults, including default answer formats, default counting for code golf, as well as standard loopholes) - if anyone wants to contest that defaults are a good idea in general, I think that deserves its own meta post. \$\endgroup\$ – Martin Ender Apr 17 '15 at 15:58

It just occurred to me that my suggested default tie breaker doesn't necessarily make sense for popularity contests. Therefore, I'd like to suggest an amendment to the accepted answer, and if this gets a reasonable amount of support I'll edit the accepted answer.

In the case of popularity contests, if there is a tie, the latest answer wins.

This may seem counter-intuitive but my reasoning is this:

  • If two answers have the same number of votes, the later answer must have accumulated them more quickly.
  • In addition, the later answer also managed to catch up with the earlier one despite the FGITW effect.

Both of these indicate that the latter answer is actually the more the popular of the two.

  • 1
    \$\begingroup\$ Sometimes, on somewhat older questions, there is an opposite of FGITW effect where if a new answer is posted, it shows up in the feed and gets more attention. Sometimes, even more than the original answers. \$\endgroup\$ – Optimizer May 21 '15 at 12:46
  • \$\begingroup\$ Could just wait until one of them has more votes... :D \$\endgroup\$ – sirpercival May 21 '15 at 12:53
  • \$\begingroup\$ @Optimizer With the bigger popularity contests that's unlikely, because the early answers will get all the attention from being on the HNQ. \$\endgroup\$ – Martin Ender May 21 '15 at 13:18
  • 1
    \$\begingroup\$ @sirpercival Sure, but some popularity contests have strict deadlines. \$\endgroup\$ – Martin Ender May 21 '15 at 13:19

You must log in to answer this question.