What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 127 Stack Exchange communities.

So I've been following this deletion discussion for some time now and I think I see a possible resolution.

A lot of the upset from "my stuff being deleted" appears to be the apparent unfairness of a moderator acting unilaterally. As a moderator elsewhere, I certainly understand that these decisions are tough calls to make.

Ideally, moderators are human exception handlers. Jeff said so. Moderators should step in when the community cannot, and otherwise defer to the community.

And so we arise at a bit of a predicament with old, highly upvoted/answered questions. To prevent content being removed, the system increments the number of delete votes required to remove it, one per every 20 upvotes (I think). This is all very well, but it creates the situation where you see this:

How many votes?

233 votes to delete a question basically means only a moderator can make that call. So, welcome to being between the rock and the hard place: if the moderator does not delete, they will face community pressure to delete from those who believe it should be removed; if they do, they will make the twitterverse very angry indeed.


Proposed solution: Cap the deletion votes required to remove any question at 20 (exact number to be worked out).

Justification:

  • If a question gets 20 or so deletion votes, chances are community opinion is strong that it should be deleted.
  • The deletion vote buffer is still wide enough that a question cannot trivially be removed if it received a lot of votes.
  • importantly: the community not moderators, can remove these questions if they deem it so necessary.
  • Even more importantly: moderators can justifiably say if you think it should be removed, vote to delete and know they're presenting the user with a way to have their voice heard.

In short, I believe that any SE site is run by us, and we need the ability to do our own cleaning (or not) and take the pressure off the moderators / put them back in the position of protecting us from ourselves (deletion wars).

share|improve this question
    
Related: meta.stackexchange.com/questions/8282/…. Note also the example I used there. It was not my proudest moment. –  mmyers Mar 3 '12 at 18:12
    
And how many votes to reinstate? –  Martin Smith Mar 3 '12 at 18:14
    
@MartinSmith The same number. –  Lorem Ipsum Mar 3 '12 at 18:20
1  
I'd also highlight this request of mine to prevent 5 users from usurping a deletion in progress. –  Lorem Ipsum Mar 3 '12 at 18:21
    
+1 Because the title is genius. –  Purag Mar 3 '12 at 18:26
6  
I'm not sure 20 is the right number, but I think a cap is a very good idea. –  Josh Caswell Mar 3 '12 at 18:47
add comment

2 Answers

up vote 4 down vote accepted

There's now a cap of 10 votes on the required number of delete or undelete votes:

from here on out, it will take at least three and at most 10 votes to delete even the most popular questions, and an equivalent number to undelete them.

Source: http://blog.stackoverflow.com/2012/03/reputation-and-historical-archives/

share|improve this answer
    
Have yourself a green tick! :) –  user142852 Mar 6 '12 at 21:14
    
@DCINinefingers Thanks! =) –  Anna Lear Mar 6 '12 at 21:15
    
Would it be possible to add it to the following page? stackoverflow.com/help/privileges/moderator-tools –  Final Contest Jan 18 at 4:11
    
@LaszloPapp Where, exactly? –  Anna Lear Jan 18 at 5:40
    
"It takes 3 votes, minimum, to delete a closed question. However, the number of delete votes required scales to the number of votes on the question and all its answers." -> Either after this sentence, or incorporating the information into that sentence. What do you think? –  Final Contest Jan 18 at 6:05
add comment

It might be worth mentioning the history and reasoning behind why delete votes were changed to make it harder to delete popular questions.

A relatively obscure (and somewhat opinionated) functional programming question was answered by Eric Lippert, who (in his usual eloquent fashion) made it a teaching moment. The answer was wildly popular; since the question was already controversial it got a lot of views, and Eric's answer was upvoted many times.

This question was subsequently closed and deleted by the community, which unceremoniously dumped Eric Lippert's answer. An argument erupted between the deletionists and inclusionists. The deletionists said that it didn't matter how good the content was in the answers if the question was bad. The inclusionists took the position that it didn't matter how bad the question was; you don't destroy good content.

The argument became so contentious that Jeff Atwood actually purged the entire post from the database, but he agreed with the inclusionists' argument, and subsequently instituted the 20:1 rule.*

And here we are today.


A lot of the butthurt stems from the fact that people are losing substantial rep from these questions when they get deleted. This is highlighted by the fact that it is very easy to see such deletions now, with the recent rep calc changes.

The problem with upvotes on these soft questions is that it is "unearned" rep, and everyone knows it. Reputation is supposed to be a measure of the effort you put into helping people with their programming problems, but these soft questions generate rep out of all proportion to the amount of effort, skill or knowledge required to answer them. These questions were never meant to accrue the kind of rep they do, which is the reason the community insisted on Community Wiki for them.


So... Your deletion cap idea. I think it's a good one, and here's why: questions can get Redditted. Those kinds of votes don't speak at all to the quality of the content; all they do is measure how many views the question gets (and it's a lot: the legendary Regex question currently has 442,000 views. Most of the votes there don't really indicate how good the content is.)

I also think 20 is the right number for the delete votes cap.

*It's been awhile, so I apologize if my facts aren't completely straight.

share|improve this answer
2  
This sounds like an epic biblical story or something. –  Purag Mar 3 '12 at 18:40
    
@Purmou: It kinda was. –  Robert Harvey Mar 3 '12 at 18:45
2  
Those who cannot remember the past are doomed to think that it was peopled with morons. This is definitely useful context. –  Josh Caswell Mar 3 '12 at 18:45
    
I haven't seen the post in question, but it sounds like a case where the question should be edited to match the answer, not closed. But how is this relevant? If it was a bad question with one stellar answer, wouldn't it still have a fairly tame total score? –  Gilles Mar 3 '12 at 18:48
    
@Gilles: There were a lot of upvotes on the question as well. 100 upvotes on only the answer would have kicked up the deletion requirement by five additional votes. –  Robert Harvey Mar 3 '12 at 18:51
3  
I'd like to bring your attention to meta.stackexchange.com/a/51018/893 - "For God's sake, if a question has valuable content, just leave it!" (your words). –  Greg Hewgill Mar 3 '12 at 19:03
    
@GregHewgill: Yes, that was part of the epic battle. And the question to which I refer: Why do people think functional programming will catch on?. As you can see, it's not even in the database anymore. –  Robert Harvey Mar 3 '12 at 19:04
    
@GregHewgill: Note that not a single person during that discussion argued that Eric Lippert's answer was off-topic, or inappropriate in any way. –  Robert Harvey Mar 3 '12 at 19:44
1  
Has that question ever been archived somewhere? I wasn't here during those days and I'm quite curious to see this question and Eric Lippert's answer. A quick google search didn't show anything. –  Mysticial Mar 3 '12 at 19:49
    
@Mystical: web.archive.org/web/20100208233119/http://stackoverflow.com/…. It's a mess, though; none of the original formatting was preserved. –  Robert Harvey Mar 3 '12 at 19:55
    
Ah, thanks! Still better than nothing. Though it doesn't have Eric Lippert's answer on it (not the first page at least). Maybe it was archived before he made the answer? –  Mysticial Mar 3 '12 at 20:04
    
Great writeup, as always. One comment I would make is on the statement "A lot of the butthurt stems from the fact that people are losing substantial rep from these questions when they get deleted." In my opinion, the most contentious arguments have come about due to deletion of content that people thought had value, like here, and reputation lost didn't really enter into that. I think people don't care as much about that old reputation as they do anything that took a little work to write. –  Brad Larson Mar 4 '12 at 2:01
    
For example, I was glad to see this question get taken back behind the shed, because I didn't deserve the reputation gained from what was just me giving an opinion in the middle of a language flamewar. I'd even removed that one from the top answers list on my Careers profile, because it wasn't quality content. I think reputation-based arguments can be put to rest pretty quickly around these question deletions, but the content issues are tougher to resolve. –  Brad Larson Mar 4 '12 at 2:04
    
@BradLarson: Point taken. But the contention has become a bit more focused by the fact that people can now see the affect these deletions have on their rep instantaneously. –  Robert Harvey Mar 4 '12 at 2:05
2  
@RobertHarvey - Yeah, that's how I caught the latter of those questions, and it might start some arguments in the near term. I think Shog9 states it nicely, and I agree with the proposal here to make it more practical for the community to deal with these questions. It's a lot easier for people to tee off on individuals, but if 10 or 20 people all agreed that something needed to be deleted, they're going to have a harder time arguing against that. It'll also work against the (misguided) notion of a cabal of moderators ruling with an iron fist around here. –  Brad Larson Mar 4 '12 at 2:22
show 10 more comments

You must log in to answer this question.