37

Cake Is Not A Lie hat: edit a post which then gets 5 or more upvotes

This hat encourages good behaviour: improving posts. This is exactly what a good badge should do.

Why not make it a permanent bronze badge? There is an existing badge to encourage people to edit at all, but this one would encourage users to put in enough effort that the post is actually improved.

It should only be awardable once. Making it awardable multiple times would lead to some annoying gaming (such as sniping edits on posts you recognize as good) and would need higher requirements.

There is a year-old badge request for editing a negatively-scored post which later gets a positive score, but I think the hat's criteria is a better choice. The negatively-scored posts that need the most help are probably below -2, and even with a good edit you'll need quite a bit of luck to bring it back over 0.

There are also a lot of very meh-quality posts with a score around 1–3 (particularly given the new review queue) which a good edit could improve enough to gain four more votes, but this behaviour would not be incentivised by the existing proposal. Requiring a relative change in score makes the tag somewhat less luck-based and encourages the positive behaviour in more situations.

(The badge also the unique attribute that it is already in use in a different form. We can look at how often the hat was awarded for an idea of how often the badge would be, and to some extent (despite the hat being "secret") we see if people try to game it too badly.)

Thoughts? Name suggestions?

|improve this question
  • I second this, sorta... – Austin Henley Dec 19 '12 at 19:06
  • 3
    My first thought was that this encourages people to edit and then upvote crappy posts, or posts that wouldn't normally deserve an upvote. I kind of wish the requirements of hats were secret - it would be more fun and less likely to affect user behavior in any specific way. – user159834 Dec 19 '12 at 19:08
  • 2
    I don't agree with the duplicate. They're very similar (and there are other similar existing requests), but this one is a bit different and has the unique attribute that it is already in use in a different form. We can look at how often the hat was awarded for an idea of how often the badge would be, and (to some degree, despite the hat being "secret") see if people try to game it too badly. This criteria may also be better than the one in the other question: if a post is below -1 or -2 it's very hard to get it positive again; a relative improvement is more consistently achievable. – Jeremy Banks Dec 19 '12 at 19:55
  • Too bad the spoiler markdown doesn't seem to support images – Some Helpful Commenter Dec 27 '12 at 21:43
  • 1
    Note that the cake hat, as currently implemented, is trivial to acquire, especially if you frequently edit your own posts. To make it into a meaningful badge, there would have to be some indication that the post wasn't going to earn those five upvotes anyway. Maybe make it require that the post had a negative score when it was edited? – Ilmari Karonen Dec 27 '12 at 22:28
17

I suggest Piece of Cake for the badge name. ;P

But in reality, it'd probably end up something like Improver to match the Editor badge.

|improve this answer
17

I like the idea, but we need to make sure badge addicts (not to confuse with badge hunters) don't destroy, annoy or make bad stuff to the site in their path to obtaining the badge.

There is always someone that put badges above rules, these people will do whatever it takes to get them (e.g upvote/approve anything when reviewing).

So we need to ask this question, in what way can someone try to game the system to obtain this badge?:

  • If someone finds a really good question or answer that will easily get +5 upvotes, he will likely be tempted to make an unnecessary edit in order to complete his goal. (bad)

  • People will start bumping old (but good) questions, because these are likely to get the +5 upvotes when they appear in the 'active' questions. (probably not so good)

  • People will start bounties more often after their edits, to get more visits and with this, increase the odds of getting upvotes. (good)

We need to work on these points, but I definitely like the idea.

|improve this answer
  • 11
    Those are basically my concerns. Go through Jon Skeet's answers, find one a year old, make a trivial edit, profit. I'm more for 0 or 1 score 'meh' answers .. why didn't they get more votes? Make them more awesome, and get the badge. The problems we're currently facing with bad reviewers just trying to get a badge more than illustrates that this would happen in reality. – Tim Post♦ Dec 20 '12 at 1:36
  • 1
    Selling badges for $500. First come first serve. – Andy Dwyer Dec 20 '12 at 11:35
  • 1
    Yes. We would need to slap a requirement that the post has a negative score at the time of the edit. – Mysticial Dec 26 '12 at 23:55
7

The behaviour that this badge should encourage (in my opinion) is to edit correct-but-badly-formatted (or typo-ridden) answers which would lead to them getting upvotes.

Unfortunately I think that this is not easiest way to get the hat (and the possible badge). The easiest way is almost certainly to do a minor edit to an already-good answer before it gets tons of votes.

That's would be the only drawback I see to making this a badge: people will game it.

|improve this answer
  • 2
    People game all the badges (or at least they try to). You could make the criteria a "meaningful" edit where that's determined by the % of the post you edit. An already good answer you tweak to correct a typo would have a low % whereas one you reformat and perform wholesale spelling and grammar corrections would have a high % – ChrisF♦ Dec 20 '12 at 11:25
  • @ChrisF but percentage requirement would encourage more sophisticated and potentially more harmful gaming. "Lorem ipsum dolor sit amet, consectetuer adipiscing elit..." - stuff like that :) – gnat Dec 20 '12 at 11:47
  • @gnat: I don't think that would work, at lorem ipsum would not cause the post to be upvoted (unless it was hidden inside tags or something similarly invisible). Maybe raising the bar high enough is enough here: as long as there are more-easily-gamed badges, it should be safe. – Joachim Sauer Dec 20 '12 at 11:49
  • 1
    @gnat - You're right (as usual). The more I think about this the more I think I may have to retract my implicit support for this idea. – ChrisF♦ Dec 20 '12 at 11:49
  • @JoachimSauer it'll be invisible if enclosed in <!-- Lorem... --> - which would make it less harmful of course. But the real problem would be when gamers will add truly innocent looking garbage that's harder to catch as abuse than easily recognizable "Lorem ipsum" – gnat Dec 20 '12 at 11:56
  • ...overquoting API javadocs is one example of gaming I am thinking about. It will look 200% innocent... but in the same time it could be actively harmful when essential, brief quotes are replaced with wall of text serving only the purpose of pumping edit percentage, right? – gnat Dec 20 '12 at 12:16
5

Couple of considerations to ensure that badge like that is awarded fairly.

It should not be awarded to posts that are proven to be going well without editing.

For example, as an editor I would find it embarrassing if I was awarded badge for an edit done here.

  • This basically means posts with sufficiently good positive score are not eligible.
    It would be likely safe to consider posts with non-positive score, or maybe with score less than 3. The downside is this sort of closes the opportunity to reward edits that made +100 of a +10 post, but I for one could live with that.

It should not be awarded for posts having good chance to go well without editing.

Posts at the top of all have good chances to get their portion of cheap populist upvotes without editing; as an editor I would be embarrassed to get a badge for an edit that easily.

  • This means only posts with score proven to be sufficiently stabilized are eligible.
    Given requirements, I'd say these are posts having no score changes for 2 days at least.

It should not be awarded for "co-located" edits.

If someone else does a badge-worthy edit that s the post, I would be embarrassed to get a badge for subsequent spotting and fixing of a minor typo there.

  • Safe way would probably be to allow only one badge per post.
    This would definitely prevent awarding badges for co-located edits. On a downside, this would introduce a chance for somewhat unfair awards in cases when really great edit follows the bump caused by some minor change but I for one could live with that.

Regarding eligibility of edits done to own posts, consistency considerations should be taken into account when deciding on this, namely the fact that current ignore such edits.

I would also consider establishing a line of badges instead of a single one. Graded by the amount of eligible edits done by a user, it would incentivize persistence in making edits valued by community.

A single rocket jump edit may happen by a pure chance, eg as a result of a luckily co-located change as mentioned above. As such, it hardly deserves more than a bronze badge.

A bunch of 5-10 eligible edits is a good indication of an editor being consistent hit maker, probably worthy of a higher (silver) recognition.

Finally, an outstanding stream of, say, 50-100 successful edits suggests somewhat of a pulitzer scale achievement, if one takes into account the sheer amount of upvotes it presumably triggered (50*5=250 at the very least, more than amount required for a golden great question / answer badge).

|improve this answer
4

I seem to have got this hat for editing (updating) one of my own answers. If this is made a badge, I think it should exclude editing your own answer.

|improve this answer
  • Exactly what I thought (happend to me too). – Felix Dec 27 '12 at 17:02
  • I don't agree. If you make an improvement on a post that then is substantial enough to get it more upvotes, why should you not get the badge? Are you not improving content? What difference does it make what user created the post? As always, the emphasis should be on content and not users. – Lix Dec 27 '12 at 18:35
  • 2
    I made the post, there were some comments, I updated it to reflect those, and upvotes continued to come in. I don't think that's at all the same as fixing a crummy post into something worthy of upvotes – Kate Gregory Dec 27 '12 at 18:37
2

I don't like that.

I already see too many minor edits in the edit review queue, obviously made by users hunting for two reputation points. The main reasons there isn't more of them might be that old reputation hunters can't get those two points.

I wouldn't want to also see many minor edits of new questions seeming promising.

Too many minor edits aren't painful just for reviewer, they're also painful for people making good edits which are prevented due to other ones.

We have enough editing in my opinion, and enough reasons to edit, we don't need that.

|improve this answer
  • Damn... I made a gravatar image just for the cake I have on SO and it really doesn't fit on meta... I need to edit a promising post fast... :\ – Denys Séguret Dec 20 '12 at 10:57
  • Darn ... I felt tempted to fix the "szz" in a minor edit, but you were faster. – Joachim Sauer Dec 20 '12 at 10:57
  • I'm sure you'll see something else to edit, as I know my English is very poor (yes, the 'z' is near the 'e' on my keyboard, which means it's not qwerty)... Don't forget to also upvote my answer 5 times :) – Denys Séguret Dec 20 '12 at 10:58

You must log in to answer this question.

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