If a low reputation user posts a question about the Holocaust, rather than being immediately posted, it should receive a notification telling the user that the question will remain visible (and editable) by that user, but will not be published until it’s approved by either the community or a diamond moderator.

The question should remain invisible even if the user edits it. That way, users won’t be able to circumvent the filter once it’s been triggered. In addition, users won’t receive any warning before the question is submitted. This will avoid people trying to misspell the word to circumvent the filter.

Update: I had a look at a sample of questions, based on which ones I flagged. It looks like a sizeable proportion lasted longer than 2 hours (if you define it as precisely 120 minutes). While the questions I flag may be a biased sample (for example, I can only flag questions which haven't already been deleted, and I'm more likely to flag questions which were posted when Australians are awake), I think that "virtually all ... within 2 hours" is an exaggeration.

12 questions that lasted less than 120 minutes: 6 minutes 16 minutes 28 minutes 54 minutes 55 minutes 1 hour, 6 minutes 1 hour, 6 minutes 1 hour, 10 minutes 1 hour, 10 minutes 1 hour, 13 minutes 1 hour, 33 minutes 1 hour, 40 minutes

14 questions that lasted more than 120 minutes: 2 hours, 10 minutes 2 hours, 42 minutes 2 hours, 43 minutes 2 hours, 52 minutes 2 hours, 57 minutes 3 hours, 31 minutes 3 hours, 32 minutes 3 hours, 50 minutes 5 hours, 32 minutes 5 hours, 57 minutes 6 hours, 8 minutes 6 hours, 28 minutes 6 hours, 33 minutes 8 hours, 20 minutes

While this is a limited sample, it already results in a total of 75 hours of Holocaust denial on skeptics.SE, with the average question lasting 172 minutes and a considerable number lasting for multiple hours.

Update 2 March 2019: In an answer to this question, fredsbend asks whether holocaust denial posts are a significant proportion of trolling posts. To examine this, I examined my flagging history, examining "helpful" rude/abusive flags, and looking only at questions. (This may not list some questions listed above, as I may have used custom flags on those questions). There are 18 such questions. Questions that explicitly mention the word "holocaust" make up half of such flags, and make up a narrow majority of questions that ended up being deleted. I have made a total of 294 flags overall, but that's including flags for comments, flags for answers, declined or retracted flags, flags about "[citation needed]" banners, and various other matters.

Mentions the word "holocaust", either in the title, the body, or with the tag (9 questions): https://skeptics.stackexchange.com/questions/43439 https://skeptics.stackexchange.com/questions/42754 https://skeptics.stackexchange.com/questions/39608 https://skeptics.stackexchange.com/questions/38322 https://skeptics.stackexchange.com/questions/38317 https://skeptics.stackexchange.com/questions/38264 https://skeptics.stackexchange.com/questions/36478 https://skeptics.stackexchange.com/questions/36336 https://skeptics.stackexchange.com/questions/36103

About Nazi Germany, doesn't mention the word "holocaust", one question involves an individual primarily known for his role in the holocaust (2 questions): https://skeptics.stackexchange.com/questions/42900 https://skeptics.stackexchange.com/questions/2260

About Jews, but not about Nazi Germany (2 questions): https://skeptics.stackexchange.com/questions/37543
Are American media controlled mostly by Jews? (2015 edition) (marked as duplicate of an open question)

Not about Nazi Germany or Jews (5 questions):
Did the former prime minister of UK say this about the Koran? (still open)
Did Adria Richards tweet "Black people CANNOT be racist..."? (still open)

  • 3
    +1 great idea. But I'm not sure if the platform supports something like it. This might need to be raised at meta.SE.
    – tim Mod
    Feb 10, 2019 at 20:40
  • 1
    I am not sure how this could work. What is "programmatically speaking" a question about the Holocaust?
    – Sklivvz
    Feb 10, 2019 at 21:09
  • @sklivvz just regex for Holocaust and maybe “holohaux”. Feb 10, 2019 at 21:25
  • 3
    there is no existing mechanism that we could use for this as far as I know, so I think it is very unlikely that we would get something like this. I think we can handle this the usual way, by flags, close votes and mod actions. And I also think you're severely underestimating the creativity and determination of some kind of trolls, dumb word filters rarely stop them for long.
    – Mad Scientist Mod
    Feb 10, 2019 at 21:34
  • @Mad Scientist - that’s why I advocated not being given any warning before the question is submitted. Feb 11, 2019 at 1:02
  • 1
    @tim Since it's your edit: "it already results in a total of 75 hours of Holocaust denial on skeptics.SE, with the average question lasting 172 minutes and a considerable number lasting for multiple hours." Is this a problem? The questions bolstering not denial are still up, so that's thousands of hours and growing. It's a strange way of looking at it. Besides, I didn't realize our user base was made up of such fragile minds, the kind that can't handle challenging thoughts with nonsense.
    – fredsbend
    Feb 26, 2019 at 7:33
  • @fredsbend Yes, it's a problem because it normalizes Holocaust denial. We really don't need to give equal weight to "both sides" on this issue. Most people will be able to be "challenged" by this nonsense and just see it for what it is. But a small percentage will start to doubt obvious facts, and those already believing it will feel emboldened by others openly sharing it. This sort of content hurts the image of skeptics.SE, and it strengthens an already growing hate movement.
    – tim Mod
    Feb 26, 2019 at 7:53
  • 1
    @tim I wouldn't say each side is getting equal treatment. 75 hours of exposure vs thousands, as you've said above.
    – fredsbend
    Feb 26, 2019 at 8:00
  • @fredsbend And ideally, the correct proportions would be 0 hours vs thousands. The question is how this can be achieved, and if it's worth having to make potentially complex changes to the platform.
    – tim Mod
    Feb 26, 2019 at 8:04
  • 1
    @tim The ideal is 0 for all spam and trolling. Arguing for this topic to get different treatment is special pleading. If 75 hours is a lot of the total spam and trolling, then you might have an argument. You'll have to show that.
    – fredsbend
    Feb 26, 2019 at 15:23
  • @fredsbend Spam is annoying, but it is not relativizing the mass murder of 6 million Jews and contributing to an ideology that seeks to exterminate all existing Jews. And once a mechanism is implemented, it could be expanded to cover other topics, though I wouldn't know of any that attract questions in violation of the code of conduct at this high a frequency (if a mechanism is to be implemented, I'm sure the stackexchange team can analyze their data and potentially include other topics).
    – tim Mod
    Feb 26, 2019 at 15:45
  • @fredsbend And for spam, my experience has been that the number is very close to 0 here (my guess is that there is already a filter in place for this). Even for issues such as blatant racism or sexism, I'm having trouble coming up with even a handful of examples. For Holocaust denialism, it was easy to find dozens of examples.
    – tim Mod
    Feb 26, 2019 at 15:55
  • @tim Your search for holocaust questions came from your own flags, right. Could be a selection bias. I realize there's no comprehensive search for non mods, but Oddthinking did search and thinks it's not an issue, which is good enough for me. Either way, you're still doing a lot of special pleading. Like I said under Lang's answer, "a dogmatically aggressive approach [to any history] is more tasteless than the denialism itself, and is certainly not the attitude of a respectable skeptic."
    – fredsbend
    Feb 26, 2019 at 16:19
  • 1
    @fredsbend You commented a lot in this thread. That looks to me as if you want and should write answer. Seems to me even if you'd just agree with 1of2 answers, your position & content isn't reflected in either existing answer? Feb 28, 2019 at 23:07
  • @LangLangC Ok, added an answer.
    – fredsbend
    Feb 28, 2019 at 23:50

3 Answers 3


This is a feature-request. It is worth understanding the status quo.

First, we should understand that genuine, good faith, questions about the Holocaust are not taboo here. We have many examples, including at least one that goes directly to the claims of holocaust deniers:

That said, Skeptics.SE is not alone in the SE network of regularly (once or twice per month) receiving bad faith questions from Holocaust denialist trolls trying to advance their political positions.

I personally find Holocaust denialism rather offensive. I appreciate that many people find it deeply offensive and upsetting. I am keen to strike the right balance: we allow light to shine on the question, while not giving oxygen to the anti-semitic trolls.

Users with sufficient rep can see deleted questions about the Holocaust.

Some were deleted by mods. Some were automatically deleted by users flagging the questions. Some were deleted by automated systems, based on the users behaviour on this and other sites.

I looked at a sample. Virtually all of them were deleted within two hours of being posted. [NB: See update below.] Some within a few minutes.

This one was an exception. It took 48 hours to be deleted by a background process, even though it had an upvoted answer; I don't understand why it was deleted, but I imagine it was based on the OP's behaviour on other sites.

So, the critical question is: does a question once or twice a month surviving for less than a couple of hours give sufficient oxygen to the trolls to warrant adding a new feature to let mods decide whether a controversial question should be seen at all?

I'm leaning towards No, but this is a result of weighing up of several factors, and I completely understand some people would conclude the answer was Yes.

Update: There has been some criticism of the numbers in this answer - particularly the claim above "Virtually all of them were deleted within two hours."

My initial statement was from a quick and dirty analysis. Andrew has performed a more systematic one, based on the flags that he has raised, which suggests my claim is false. Only being questions that Andrew flagged is its own source of bias, so I wanted to do another analysis.

Today, I tried to do a systematic one based on all deleted answers, using the Stack Exchange Data Explorer. However, my assumption that I could see the deleted questions if I was logged into the Data Explorer as a mod was false.

So, instead I ran a query of the newest deleted questions that reference the holocaust and manually walked through it until I got bored, which turned out to be 16 questions. [Disclaimer: Not having a fixed target before collecting the data is a potential source of bias.]

I stopped the clock when the question was closed (if it was closed before it was deleted). I bolded the ones higher than 2 hours.

Based on the number of bolded entries, I agree that Andrew's criticism of my "2 hours" claim is fair and correct; I thank him for challenging me on this. I withdraw the claim, and apologise for being misleading.

  • only mods can see the list with deleted questions. And the one question that was deleted 48 hours later was deleted because I destroyed that user at that time.
    – Mad Scientist Mod
    Feb 11, 2019 at 7:13
  • @MadScientist: *scratches head* Can't 10K rep users see deleted questions? What am I missing? [Destroyed the user? That explains it. Thanks.]
    – Oddthinking Mod
    Feb 11, 2019 at 7:41
  • 1
    @Oddthinking For your search query, I'm getting "Note: only content you own is returned when searching for deleted content". I can view deleted questions/answers if I have a link to the question, but I can't search for them.
    – tim Mod
    Feb 11, 2019 at 7:55
  • 1
    @tim: Thank you. I didn't realise that. I am disappointed, because I thought 10K users could keep a watchful eye on mods and what they were deleting.
    – Oddthinking Mod
    Feb 11, 2019 at 8:06
  • 3
    Maybe it’s only 2 hours but feels like longer! Feb 11, 2019 at 11:02
  • Maybe more like: "+10k users can "donate an eye", occasionally, if pointed towards such a deleted post?" – Without someone raising awareness (eg on meta), the deletion thing is quickly hidden from almost all views Feb 11, 2019 at 22:06
  • 2
    This question about the Sandy Hook shootings is the furthest I've ever seen a troll get on this site. Despite getting a little bit of reputation (maybe as high as 30 or 40 points), OP decided to post "answers" to the question and some comments on my answer until their account was deleted (and probably also with new accounts until the question got protected)—4 troll answers total. All that is left is a (heavily edited) question that doesn't look all that trollish and a semi-high-voted answer by me, which is not really that bad of an outcome.
    – Laurel
    Feb 12, 2019 at 20:59
  • 1
    I've done my own survey of how long such questions last - it is different from what you described in this answer. Feb 23, 2019 at 0:39
  • @AndrewGrimm: tell us more. I didn't research this systematically and am ready to be down to be wrong.
    – Oddthinking Mod
    Feb 23, 2019 at 5:06
  • 1
    @Oddthinking Sorry - I posted it in the question. Feb 23, 2019 at 5:15
  • 3
    @Oddthinking I added examples from my flags to OPs question as well. Taken together, the average is about 3 hours, and a non-negligable amount of questions lasting 5 hours or longer. That's not as bad as it could be, and it's not the fault of mods or the community (we all need to sleep and work and have a life), but I do think that it is longer than it should be. [it may well be that the average is way lower, because a lot of questions may have been closed by mods in a short timeframe, and we can't flag those; but that would likely mean that it's (much) more than one question a month]
    – tim Mod
    Feb 24, 2019 at 18:30
  • So, does your edit indicate that you changed your conclusion? "I'm leaning towards No."
    – fredsbend
    Mar 5, 2019 at 4:00
  • 2
    @fredsbend: It weakened how much I am leaning, so I am more on the fence, but ultimately this is a weighing up of several factors: I don't think my opinion is worth a lot here.
    – Oddthinking Mod
    Mar 5, 2019 at 4:05
  • I'm not 10K, so I can't see any of these. I think if we're going to argue "minutes of exposure" as a useful metric, we also need the total time frame. So this would be 16 posts lasting 4258 accumulated minutes (70 hours) over X days/weeks/months. If X is significantly high, perhaps perspective is that we don't have a problem. What would be significantly high enough?
    – fredsbend
    Dec 13, 2019 at 23:34
  • However, in theory, this breaks down, because if the total volume of questions on this site multiplied by say 100, presumably so also would this issue. That would make it 1600 posts lasting 425,800 minutes (7100 hours) over the same time frame. This metric looks worse for bigger sites. We should perhaps also consider site total posts over the same time frame as well, perhaps broken into percentage.
    – fredsbend
    Dec 13, 2019 at 23:34

I'm starkly against giving any topic special treatment unless there's a good reason to do so. "It's a sensitive topic, denialism is offensive", etc. is special pleading, not a reason to do this.

Technical possibilities aside, I'd be in favor of this if targeting this topic (any topic, really) helped lower spam and trolling posts by a significant amount. We'd first need to know how much of our time is wasted on moderating these kinds of posts. tim was kind enough to take some time looking through his flagging history, just as Andrew Grimm (OP) did, and they discovered a fair amount of such questions, but we'd need to compare that to all or a random sampling of our trolling/spam posts that weren't automatically caught. If I was going to put a number on it, I'd say at least 10%, but more reasonably 20%. That is to say, 20% of our trolling/spam posts that aren't automatically caught need to be on a single topic before I would support topic specific auto-actions.

To get this data, I'm sure we'd have to depend on SE staff or the moderators. SE staff, being as clever as they are, may have already implemented such things on other sites. Perhaps they have their own protocol before implementing these things.

On a subjective note, as a regular and often daily reader of this site, I don't perceive that we get a lot of holocaust denial trolling ...

  • 2
    Treating bad-faith posts containing Holocaust denial differently from, say, bad-faith posts about the moon landings is not special pleading per se. Stack exchange treats “offensive content” (along with spam) as a reason to delete content, and also to punish the poster more severely, so the community has decided that offensive content is something to treat differently. If I was explicitly advocating that we should treat Holocaust denial differently from Armenian genocide denial on the grounds that Holocaust denial is offensive, then that might be special pleading. Mar 1, 2019 at 2:32
  • 1
    I understand that you concede that, in absolute terms, there is a significant amount of bad-faith Holocaust denial posts, but you question whether it is a significant proportion of problematic posts. I can’t give you precise figures, but I’d say it constitutes a majority of successful “hate speech or spam” flags I generate, and a non-trivial proportion of all flags I generate. Mar 1, 2019 at 2:43
  • 2
    @Andrew Regarding offensive content being deleted and perpetrators punished, that's not what you're advocating, as that's already done. You're advocating that there be automatic actions only for this topic, precisely because you find it so offensive. That is special pleading. - I haven't looked through my flags, but we seem to have the opposite experience.
    – fredsbend
    Mar 1, 2019 at 2:53
  • I have just added to my question an analysis of what proportion of problematic content consists of holocaust denial trolling. Mar 2, 2019 at 0:45
  • I’ve added my analysis answering this objection. (Edited version of an earlier comment) Mar 3, 2019 at 1:49
  • 1
    "To get this data"… I agree on your impression regarding ratio of questions. But the data/transparency angle is crux. Mods seem less interested in doing a hard analysis on this. Staff seem to be pressurable into taking action if a twitter mob shows up, but not when users start to 'have ideas' (as per the usual voting on metaSE). Even if a few more users like Andrew start analysing their flags and share the data somewhere, it'll never approach the level of insight staff could have&share on this. I find it a gripe that this resembles so often 'security thru obscurity' in principle. Mar 5, 2019 at 2:41

Seems to me that questions on that topic are the lesser problem, often being 'just' a tedious nuisance if well-intended, and really quickly shot down if on that level of easy recognisability that would be handled by regex, review queue or mod-action.

After all, this would be the site on SE to handle this. Holocaust denial is a widely held belief, hence notable and in need of address (if it's not a duplicate).

Questions shouldn't be seen as reflective of the belief system of the asker. They sometimes are, which in this case is really not good. But as the default position it seems quite wrong to assume so.

That said

Has the Holocaust been exaggerated?

seems like quite the evil frame to set up, including personal belief system of the unwelcome kind. Going by views it seems to have made HNQ at some time (would it be worse were they coming from Google?).

The most problematic point to observe are the answers given.

And how they are handled by the mods, the community, foreign SE visitors.

In the example above about 'exaggeration' it looks to me that it attracts really bad answers. We should ask ourselves here whether that handling of answers can be improved?

On historySE the issues with answers of that calibre are as per current meta not of large enough interest to the community at large and too many do not recognise perfidious argumentations.

That is a problem much bigger than a daily troll trying to push his dreck.

On SkepticsSE a flag of mine on an answer to another question that was going into the exact direction of 'this is denialism and distortion' was declined with:

declined - please leave politics out of skeptics –

So I just downvoted this answer, that still has a vote count of +2/-7, in other words a net-rep-gain. (The fact that the flag-declined-comment is somewhat nonsensical should be clear, at the latest evidenced by having this post, at all)

Both of these experiences leave me with a preliminary conclusion that this problem keeps boiling, on a systematic level.

This should not be read of an accusation of MetaSkeptics readers, but for educational reasons in general, the following links need to be read again, and again in many cases:

United States Holocaust Memorial Museum article on "Holocaust Denial and Distortion"
Ask Historians Reddit – ELI5 - Holocaust Denial
Wikipedia: Holocaust denial

Especially the 'softcore denialism' and 'distortion' aspects are already in dire need of downvotes (and even deletions) –– in answers. These downvotes are much too rare on these attempts in answers. This distortion can happen even inadvertently, and is consequently hard to spot for many if not most users here on SE.

In summary, I don't think the questions on this are a problem here and now.
But the answers –– and how they are handled, even here –– *are, way too often!**

  • 1
    On the “exaggerated” question, there’s only one awful answer which hasn’t been deleted, and it’s been downvoted into oblivion. I don’t think I’m properly understanding what you’re saying. Feb 12, 2019 at 11:45
  • @lang Regarding this, I disagree. I think the question and answers are fine. I think the voting on them represents well what reality is. In other words, you'll get a real and accurate answer on that page.
    – fredsbend
    Feb 13, 2019 at 18:13
  • @lang Ok, and is the holocaust a special topic here, or does everything else stand against this metric also? "relativists and denialists" are not useful terms here. The dogmatic defense of any history is more concerning to me, regardless of how tasteless any denial of it might be.
    – fredsbend
    Feb 13, 2019 at 22:03
  • 1
    Alas, denial isn't even in question on that page, so you have to invoke "relativism", a term I've never seen used in conjunction with the holocaust. And the takeaway is you are no longer taking reasonably. The take away is that you have an exact position and all others are not just factually incorrect (the purview of this site), but are morally wrong and should be stamped out (a dogmatically aggressive approach that I think is more tasteless than the denialism itself, and is certainly not the attitude of a respectable skeptic).
    – fredsbend
    Feb 13, 2019 at 22:03

You must log in to answer this question.

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