Unregistered Users (in Theory and in Practice)

There's been past discussion (e.g. here) about whether MathOverflow should allow posts from unregistered users. I am fully on board with the apparent consensus that the answer is yes, because of the potential for a registration requirement to deter valuable contributors.

At the same time, we seem to be currently under some kind of an assault from a user who, over a very short time period, has created dozens of unregistered accounts (mostly with different user names) for purposes that are unclear but might be nefarious (I'll elaborate on this below). This leads me to ask two questions:

Question 1: Would it be feasible to limit the number of unregistered accounts coming from a single IP address (say to one or two or three)?
Question 2: Would it be desirable to limit the number of unregistered accounts coming from a single IP address?

The above two questions stand on their own regardless of the cluster of unregistered accounts that inspired me to ask them. But I'd like to elaborate a little on why I believe that's a cause for concern:

At the time I last posted about this, the user in question had approximately 13 accounts. Since then, they've proliferated like hydra heads. They mostly focus on a single theme (i.e. the purely topological properties of schemes), and the user continues to avoid engagement with responses of all sorts, including those seeking clarification. (See here for example.)

It seems to me that:

• The sheer number of these questions suggests that the OP is investing very little effort in them before posting. It looks more and more like his (or her) policy is to instantly post every question that enters his head.
• The lack of engagement with others who are trying to be helpful suggests that the OP doesn't actually care very much which questions get answered.
• This has led me to wonder whether the OP's goal is to use MO as a machine to produce a paper (or a PhD thesis?), not caring much which particular questions get answered as long as there are enough answers to compile into a manuscript.
• The above speculation might also explain why the accounts are all unregistered and (mostly) under different names, making it marginally harder for a referee or a PhD committee to tie them all together.
• On the other hand, I am aware that I am a relative of a chimpanzee and hence programmed from birth to be oversensitive to apparent attempts to breach social norms. So maybe I'm way off base here.
• Nevertheless, it seems that something pretty weird is going on, even if I have no idea what it is.
• So again----although Questions 1 and 2 stand on their own and require no opinions on the current situation, I will add Question 3:

Question 3: Are concerns about the current situation a good reason to adopt the proposal from Questions 1 and 2?
• 1. Yes. 2. No. 3. No. If a number of questions are posted publicly with the result that more mathematics is considered and created, and the basic mission of the forum is fulfilled, then we should not care who posts them. Ultimately, it does not matter about the responsiveness of the poster so much as the responsiveness of the community. Since the information provided is widely accessible, I could try to make a thesis out of it if I were so inclined. Would you be as concerned if this were a syndicate and not an individual? Gerhard "Would Cite MathOverflow, Of Course" Paseman, 2019.06.02. – Gerhard Paseman Jun 2 '19 at 15:48
• @GerhardPaseman: I am mildly annoyed that I took the time to answer one of these questions when I could have spent my time thinking about a different question that someone actually cared about. One sometimes answers a question in the hope that the answer will interest someone. These particular questions mostly seem unlikely to matter much to anybody but the OP. Now that it appears that they also don't matter to the OP, I feel like I wasted a (very small amount of) my time. I am presumably not the only one. – Steven Landsburg Jun 2 '19 at 15:56
• Your feelings and impressions are important. So you should talk about those. Would you feel you wasted time if someone else engaged with you about your answer? Gerhard "I Feel This Is Important" Paseman, 2019.06.02. – Gerhard Paseman Jun 2 '19 at 16:14
• I believe concern 3 (let the MO community write your thesis/paper) is not unique to unregistered users. I can remember registered users using MO for asking slowly evolving questions on one and the same topic over and over again and not doing anything else here. – Christian Remling Jun 2 '19 at 23:31
• regarding question 1, I am not sure how exactly are the IP addresses registered but at some large math departments (at large state universities in U.S., for example) there can be potentially more than 3 people who at some point of their careers wish to anonymously contribute to MO. In the current situation, it seems that only questions are being asked while the valuable anonymous contributors tend to give answers; this distinction might be useful. – user74900 Jun 3 '19 at 13:51
• regarding question 3, I am not convinced we should care about the nefarious intents of the anonymous user. In the past, Anton Geraschenko has stated, I think, that MO is not responsible for somehow banning people who cheat on homework, it is only responsible for maintaining its own quality standards. If they do publish a plagiarized paper or thesis and you happen to notice it, you can in private capacity notify the relevant university but I do not think that MO as a community has to care (given Anton's earlier stance on a similar issue). – user74900 Jun 3 '19 at 13:54
• you can deal with the questions individually, if some of them are really low-effort to the point of being incomprehensible, just flag them (or what is the usual procedure?). No new ideas needed here. If you feel like there is a consistent stream of really bad question, a suspension on the IP address might be instituted, I think that is possible in terms of technical means. – user74900 Jun 3 '19 at 13:56
• I agree with you that one special feature of this case might be that even if the anonymous user happens to ask a reasonable question and they happen to receive a reasonable answer, they won't accept the answer and thus the system is going to send this question to the front page sometimes. Though I personally think it is the weird behaviour of the Community user that is at fault here (never really understood why they built the software this way, it also causes some issues with the big-list questions I think). – user74900 Jun 3 '19 at 14:00
• @cardinal Actually, as soon as one of the answers has score at least one, the question is no longer bumped by community user. It is not necessary that the question actually has an accepted answer. Some links with the basic info on community user can be found in the tag-info. There was also an older question on this meta that touches on purpose of bumps by the community user: Auto-bumping of unanswered questions. – Martin Sleziak Jun 3 '19 at 15:02
• I will just add that the linked discussion was specifically about registering for asking questions (not posting answers). While it is true that registration might scare of some users, it's probably still better than if they later find out that they no longer have access to their account. Probably I should have stressed this aspect back then, too. (On Meta Stack Exchanges you can find more details about cookie-based accounts.) – Martin Sleziak Jun 3 '19 at 15:29
• Wouldn't it mostly solve the problem without doing any significant harm if posts of new or unregistered users would be shadow banned until successfully having passed the "First Posts"- / "Low Quality Posts"- and questions also the "Close Votes" review queue? – Stefan Kohl Jun 6 '19 at 10:46
• What is an "unregistered account" with a user name? I'd have guess that being "registered" means having a user account. So having an unregistered account with a user name seems oxymoronic. $\qquad$ – Michael Hardy Jun 9 '19 at 2:48
• @MichaelHardy you see Michael, when you register you have to confirm your e-mail by clicking on a confirmation link. Unregistered accounts are the accounts created when you are not logged in any registered account and ask a question or post an answer. You have to pick a username when doing so, and afterwards you can confirm your e-mail, so your unregistered account becomes registered. So no oxymorons here. – user141498 Jun 9 '19 at 7:43
• regarding 3, it seems to me that the activity in question involves bots, and not only real people typing questions... – Dima Pasechnik Jun 12 '19 at 16:20