For anyone not familar with a "protected" question, here's a pretty useful case for the tool… and a counter-example where it make, umm... less sense.

A protected question prevents answers by very new users. Questions should be protected when they are garnering lots of views and newbies are adding "me too!", "thanks!" and possibly even spam non-answers — incidentally, which seems to happen most often when a question is blasted out to large numbers of people who do not understand Stack Exchange.

But English SE has "protected" almost 1,000 questions. That's over 5%, almost twice the #2 site, and 20 times more than most of the network! I'm not sure why most of those questions needed protecting at all.

Preemptively Blocking Users?

The protect-question tool is extremely useful when "we've seen enough poor answers." But when you start to proactively protect questions just in case, you are blocking the most basic functionality of the site for a large number of users. That's not how Stack Exchange is supposed to work, and that's not the purpose for which the tool was designed.

Having said that, I don't run this site. You do. So I'm asking about the policies behind the use of this tool, and if perhaps a little introspection and adjustment is needed.

  • 2
    Looks like the collective response here is we don't think there's a problem of too many questions being protected. Me, I wouldn't mind having the current number even if you needed, say, 50-100 points to answer. I'd be more interested in content quality than in site democracy/equal access. Commented Oct 2, 2012 at 20:28
  • 1
    @FumbleFingers No, there doesn't seem to be a problem; or at least it's safe to say the issues have been well considered and somewhat specific to this community (which is what I was hoping for). So thanks for the discussion and the well-thought-out replies, everyone! Commented Oct 2, 2012 at 20:37
  • 2
    Quite likely Reg has protected a few questions unjustifiably by some people's lights, but personally I think he nearly always makes a good call. Of course, the rest of us can always kick up a fuss and reverse a peremptory question closure - that just needs five reopen votes for the rare mistake. But I suspect if we made "protecting" a voteable option for high-rep users we'd have a much stronger "slash and burn" policy than we do already. SO.ELU is unusual - people sometimes think if they know their own language, they're qualified to answer more than they should here. Commented Oct 2, 2012 at 23:33

4 Answers 4


In defense of your "less sense" example, I would suggest that when a question gets sent to the multi-collider because it has something involving sex or programming or both in the title, said question garners way too much attention and votes. Think of it as an attempt to keep the quality of the site higher. It's not a successful attempt, because the "protect" method works about as well as the rhythm method does for preventing unwanted pregnancies, but at least it is an effort.

Look at the highest-rated questions and answers on this site. Some of them are good, but many are not in keeping with any rational set of standards regarding English and its usage.

What is the origin of ZOMG?
What is the factual basis for pirate speech?
Is there a word for a person with only one head?.

Really, can you honestly say such questions are worthy of the attention they get? We have many questions that are worthy but damned, by comparison, with faint praise. Consider this answer given by Kosmonaut, which was upticked a mere four times. Compare it with the fifth-highest answer to the "word for a person with only one head" question. That answer, unresponsive as it may be, garnered twice the number of votes.

Look, you people set up a series of sites that crowd-source answers, and yet you then turn around and tell us (as Jeff Atwood did) that we're allowing too many crap questions and answers. I think he was referring to something like the slang words for breasts question, but it doesn't matter. The point is, ELU is unique in one important respect: Everybody who comes to these exchange sites speaks some kind of English, so everyone feels entitled to weigh in on the bikeshedding issues at the expense of question quality.

I don't know how to remedy the situation. But blaming the victim isn't an option that seems very useful at this point.

  • 1
    +1 for the comparison. I had never heard of the rhythm method (it goes under a different name in Italy), but your meaning is very clear and made me laugh.
    – Paola
    Commented Oct 3, 2012 at 15:36

Everybody calm down and look at this:

About 1190 protected questions

About 791 of the protected questions are closed

That's right. Some 791 out of the 1190 protected questions are also closed. So for the sake of this discussion, these are false positives. 66% false positives, mind you.

I just went ahead and unprotected some of them, because I'm nice like that and tomorrow's a day off anyway, but frankly, that's a waste of time, is it not.

And yeah, what Kit says. And what Rob says. And what Tom says. Whenever we get a question about the f-word, or the n-word, or the c-word; the p-word, the t-word, the b-word; the j-word, the g-word, the l-word; or any of the other words Jeff used to get all up in arms about, we absolutely do protect them preemptively. Because that's what we all agreed upon two years ago in bright daylight.

Lastly, as chance would have it, I elaborated on my personal policy slash general impression in a comment just a couple weeks ago.

[P]opular questions that are one step short of bikeshedding are obvious candidates [for getting protected]. However, less popular questions that do have a definite answer probably account for just as many protects. If an elaborate answer has been sitting accepted for two years, and someone new comes along and posts a half-assed one-liner that adds nothing except typos and lols, and the question gets bumped, and someone else sees it and posts yet another pointless one-liner — we don't really need that. And then of course there are questions about offensive words we protect preemptively as soon as we see them. – RegDwighт ♦ Sep 20 at 19:24

Edit: some more Google stats.

  • "Protected by KitFox" — 5 hits
  • "Protected by simchona" — 5 hits
  • "Protected by tchrist" — about 19 hits
  • "Protected by Jasper Loy" — about 144 hits; "Will Hunting" — 41; 185 in total.
  • "Protected by RegDwighт" — about 414 hits; "ЯegDwight" — 348; "RegDwight АΑA" — 55; 817 in total.

So yeah, as Kit hinted at, any and all flak for this should go my way.

  • 3
    The 'l' word? The 'p' word? I gotta get out more often.
    – Mitch
    Commented Oct 2, 2012 at 19:35
  • 2
    obviously the b-word is Balrog. Commented Oct 2, 2012 at 19:39
  • @Mitch You don't know The L Word?
    – tchrist Mod
    Commented Oct 2, 2012 at 19:40
  • 1
    @Mitch: penis, poo, pedobear, lesbian, liberal (tell me to stop)...
    – RegDwigнt
    Commented Oct 2, 2012 at 19:44
  • @RegDwighт: Thank you.
    – Mitch
    Commented Oct 2, 2012 at 19:52
  • 3
    Reg: the stats are tilted toward you because, as moderator, you can protect questions immediately; the rank-and-file must wait 24 hours. Sure, your high numbers could mean you're "overprotective," but they could also mean you're simply efficient at nipping things in the bud. (That's all the flak I plan to send.)
    – J.R.
    Commented Oct 2, 2012 at 20:15

I think Reg can probably answer your question best, but I will put my two cents in.

It is often the case that we get new answers on a very old question and that sparks a slew of new answers on the same question by other new users. Protecting a question when that happens helps mitigate the negative impact of an answer bump.

We proactively protect questions that we think may garner a lot of attention because they are controversial or include words or language that attract spammers or other types of non-constructive, opinion-based, soapbox types of answers. These are probably more likely on our site because of the types of questions that may be asked here.

Lastly, you need fewer reps to answer a protected question than you do to upvote, so I think we are primarily discouraging drive-by and bandwagonning answers from users who are not much invested in the community.

I'm interested to hear what others think about it.

  • 2
    Since you asked what others think, I'll chime in. I'm at a loss to figure out what all the hullabaloo is about. How many points do you need to answer a protected question – 10? A lone upvote will get you that. If protecting questions meant you needed a rep of 1,000 to leave an answer, I'd think maybe the O.P. has a point. With such a low threshold, though, I hardly think protection leads to any disenfranchising. I can't accept the O.P.'s premise that this blocks "a large number of users." Besides, a new user can leave a decent answer somewhere else, then later answer a protected question.
    – J.R.
    Commented Oct 2, 2012 at 20:04

Questions related to sex and/or profanity tend to attract bad answers:

Similarly, questions with very high vote totals and/or view counts also tend to be lightning rods for bad answers:

By “bad answers”, I mean ones that will end up being deleted, usually from being non-answers, offensive answers, or spam answers. If they're doing to be deleted anyway, it seems like less work for them never to be created in the first place. Yes, sometimes these are just very low quality answers, without much research backing them; those take longer to get out of the system, if in fact that ever happens.

First-time posters are often unfamiliar with how the site works, and what it even takes to be considered a good answer. A lot (but not all) of these bad answers are from drive-by first-time posters, which is the only thing that protecting a question staves off. It really doesn't seem like much of an impediment to ask a poster to first get one single upvote on one of their previous answers, or two on one of their own questions, before posting an answer to one of these lightning-rod questions. I mean, does it?

Protection is easily reversible, and doesn't seem like much of a barrier anyway. If I am wrong, please let me know.

You must log in to answer this question.

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