0

My understanding is that tag badges require a minimum number of answers for a tag to address the tag badge exploit. However, I haven't read anything mentioning any requirements for those answers (other than the total score meeting the badge threshold). If someone satisfied only a badge's score requirement, what would stop them from posting a bunch of low-quality answers to satisfy its required number of answers? (Presumably they'd post short but correct answers or just regurgitate someone else's answer to avoid having their answers deleted or downvoted.) Should tag badges instead require a minimum number of answers with a positive score?

Or probably this isn't worth implementing (and explaining) since people who meet the score requirements are probably good citizens who wouldn't try to game the system?

3
  • I have to get some sleep or I'd do it, but somebody should really run an sede query against gold badge holders to determine just how big the impact of this is, if any, before proceeding further with speculative conversation.
    – Jason C
    Commented May 28, 2015 at 3:45
  • (So why the downvotes? Is it because this is such a terrible idea? Did I apparently not ask a clear and thought-out question?)
    – jamesdlin
    Commented May 28, 2015 at 7:26
  • meta.stackexchange.com/questions/47634/…
    – Jason C
    Commented May 28, 2015 at 10:15

2 Answers 2

2

I think you are right this can be exploited, but I wonder how big the impact actually is. Therefore it might not be necessary to fix it.

One concern though is that since that post, having a gold badge in a tag does give an extra privilege: dupe-closing a question. That might make the impact of the exploit bigger than it was before (who cares about the 'free' badge?).

I think this could be addressed with an extra requirement as in your post: at least 50% of your posts in that tag should have a positive score.

  • Bronze: 10 questions out of 20 should have a positive score;
  • Silver: 40 questions out of 80 should have a positive score;
  • Gold: 100 questions out of 200 should have a positive score.
3
  • 25%, really? I mean, when does a good answer actually have a non-positive score? Maybe make it 75%? Commented May 19, 2015 at 9:02
  • This is just to prevent messing with the score. If you need a decent amount of positive scoring posts, you will pass. This will also take into account helping users that don't have the up-vote privilege yet. Personally I have no problem raising it to 50% though. It is just an arbitrary number. Commented May 19, 2015 at 9:04
  • 1
    @ChristianRau: updated. Commented May 19, 2015 at 9:05
0

You could also limit the positive impact of one answer by like +20 or +40 for example. That way a really great answer can't give you the gold badge right away. You still need a couple great answers to get your gold badge.

The problem when putting a minimal percentage of >0 answers is that people would hesitate answering on below average questions by new users because they know they aren't much likely to get an upvote. I would agree with a percentage of non negative answers though.

2
  • I never suggested a percentage of positive scoring answers. I suggested an absolute number of positive scoring answers, precisely because I agree that people should not be discouraged from posting answers that might not be upvoted.
    – jamesdlin
    Commented May 28, 2015 at 7:22
  • @jamesdlin this was more as a comment to the other answer from Patrick Hofman which does suggest a percentage of good answers. I do agree with putting a requirement of an absolute number of >0 answers.
    – meneldal
    Commented May 28, 2015 at 8:07

You must log in to answer this question.

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