29

I noticed today that a request for protection came up due to the the large number of me-too answers on a trending question I answered recently. This got me thinking, since trending questions get far more attention than the average question, should questions that trend be automatically protected?

On the upside, this does seem to be a common problem and the auto-protection could be put in place only while the question is trending. On the downside, it would make it more difficult for new people to participate in answering a trending (highly visible) question so it could be a barrier to entry.

Thoughts?

28
  • 2
    This seems like a good idea to me.
    – Linuxios
    Mar 6, 2013 at 14:36
  • 2
    "looking for long answers..." notice: 1) wouldn't make a barrier to entry and 2) would give users and mods sufficient justification to delete metoo crap as a blatant violation of the "rules set for this game"
    – gnat
    Mar 6, 2013 at 14:44
  • 3
    @gnat Seeing how many bounties you're putting up and the sheer indifference from the higher-ups, I think it's safe to say that we may need to take things into our own hands. 1) Get a gang of 3k users to close-vote everything stupid that tops the multicollider. 2) Offset the multicollider by manually redditing everything else that is good. (That is: Give even more attention to everything else that actually deserves it.) I don't like this latter method since it will lead to spamming. But it has proven effective in getting attention to those hidden jewels.
    – Mysticial
    Mar 19, 2013 at 5:40
  • 1
    That said, I also don't like the idea of abusing close-votes and downvotes for the purpose of denying visibility... But there's no other alternative. In the recent months, I've been seeing people doing more and more of this. (That is: improperly close-voting a valid question to deny attention and thus leading to ridiculous close-reopen wars...)
    – Mysticial
    Mar 19, 2013 at 6:08
  • 1
    @Mysticial guerilla! :) You're reading my mind; I was thinking about it, too - even prepared legal justification for er "corrective voting". BTW regarding voting, per my observations even 3-4 coordinated DVs could make a substantial difference in hotness score; score sensitivity to manipulations like that is explained here - "Yet another indication of issues with current formula..." Interesting to note that neither "my" nor "your" variant of formula wouldn't be as sensitive to such tricks but oh well
    – gnat
    Mar 19, 2013 at 6:41
  • @gnat Haha... There really aren't too many options. So we're bound to come to the same ideas. :) Although I hesitate to recommend this, extending the coordinated downvotes to all the answers as well also helps to lower the hotness score. But anything beyond that is clear into the territory of "suspendable abuse". (Such as coordinated rollback wars (to get a post locked) and red-flagging.) (Not saying that improper and coordinated close-votes and downvotes aren't already "abusive".) But the fact that these nasty ideas even come to mind shows that there's a clear problem in the current system.
    – Mysticial
    Mar 19, 2013 at 6:59
  • 1
    @gnat Agreed. And it's worse than that. They don't help as much and they cost -1 to cast. Furthermore, IIRC, only scores from the top 3 answers are counted towards sum(Ascores). And the top 3 answers are usually the ones that are good. So now you're downvoting potentially good answers. On the flip-side, downvoting the tail of 0-score answers will let them be delete voted. But that needs to be done fast before the masses upvote them out of negative again.
    – Mysticial
    Mar 19, 2013 at 7:07
  • anyway, my plan for next 3-5 weeks is rather peaceful: I am going to put 3-4 more related bounties (until my rep falls dangerously close to 2K) - after that, I'll likely be open to consider er more radical measures :)
    – gnat
    Mar 19, 2013 at 7:08
  • 2
    Nah, don't waste all your rep on this. If we hear nothing anything after this one - assume it won't happen. (This comment trail should be a hint that things might get ugly if nothing is done.) I've already been testing the waters a bit in our C++ chatroom with coordinated deletion of late answers to popular questions. There's probably enough of us who are interested in targeting dumb things that top the multicollider. (since a number of us do regularly speak out against them... but only in chat)
    – Mysticial
    Mar 19, 2013 at 7:24
  • hey! I recently saw a wonderful example of a terribly mediocre question with 20+ FGITW answers that had no chance at collider because of several DVs that strategically "balanced" about the same amount of upvotes. If memory serves, it was in C++ tag :)
    – gnat
    Mar 19, 2013 at 7:29
  • 1
    @gnat I just tested the "abuse CVs to deny visibility" idea on this question. (I linked it in the C++ room.) Well... it got ugly...
    – Mysticial
    Mar 26, 2013 at 23:02
  • 1
    @gnat It actually has no effect. The time decay is updated every top of the hour. You just happened to catch it at the top of the hour.
    – Mysticial
    Mar 27, 2013 at 12:00
  • 1
    @gnat I was the first to upvote it. :)
    – Mysticial
    Aug 6, 2013 at 7:28
  • 1
    @gnat I can't say I'm completely satisfied with it though. Genuinely good questions that deserve to be stuck at the top won't be able to stay at the top. IOW, it's "solving" the problem by equally nerfing all questions (good and bad) at the same time.
    – Mysticial
    Feb 4, 2014 at 19:25
  • 1
    @gnat - I concur that this is addressed by that. May 12, 2014 at 13:28

2 Answers 2

20
+150

As soon as three posts from low-rep users are deleted by mods or the community, the question is auto-protected. This threshold is pretty good in my opinion, it avoids shutting down external users from the start, while still protecting the true crap magnets.

Protecting is a far more dangerous tool than it appears at first glance, as it can prevent good content from ever being written in the first place. I think we should generally be careful when to use this tool.

1
  • 4
    That makes sense, perhaps it could be an optional feature for each site to decide the setting on depending on the normal level of activity on a particular site. Some of the smaller sites can tend to be hit really hard really fast and have fewer moderators to deal with it. Mar 6, 2013 at 14:42
8

As Anna Lear points out in another question, proactively protecting questions is generally a bad idea.

I generally keep an eye on Hacker News and other programmer sites for when our questions hit the front page. We as moderators try to keep an eye on these questions and actively moderate the bad stuff out quickly.

There is the possibility of 'going too far' and over moderating a post just because it's highly visible. We want to keep that from happening, as that can color the outside world's opinion of Stack Overflow.

6

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