# An unusual pattern of unregistered users?

There's probably far more, but these are the ones that I had in my notes, which I noticed mainly based on Q/A they posted:

User IDs between 0-999:
https://mattermodeling.stackexchange.com/users/804/steve-plimpton user804
https://mattermodeling.stackexchange.com/users/950/danish user950
User IDs between 1000-1999:
https://mattermodeling.stackexchange.com/users/1146/roselyn-lemus-martin user1146
https://mattermodeling.stackexchange.com/users/1307/todd-martinez user1307
https://mattermodeling.stackexchange.com/users/1399/alexandre-c-dias user1399
https://mattermodeling.stackexchange.com/users/1376/paul-ayers user1376
https://mattermodeling.stackexchange.com/users/1570/maat?tab=profile user1570
https://mattermodeling.stackexchange.com/users/1507/lee-cronin user1507
https://mattermodeling.stackexchange.com/users/1592/lesya user1592
https://mattermodeling.stackexchange.com/users/1286/lazarus user1286
https://mattermodeling.stackexchange.com/users/1017/ben user1017
https://mattermodeling.stackexchange.com/users/1080/user1080 user1080
https://mattermodeling.stackexchange.com/users/1684/savir user16841
User IDs between 2000-2999:
https://mattermodeling.stackexchange.com/users/2010/sebine-murselova user2010
User IDs between 3600-3999 (notice it's from 3600 not 3000):
https://mattermodeling.stackexchange.com/users/3815/phdstudent user3815
https://mattermodeling.stackexchange.com/users/3619/nishan user3619
https://mattermodeling.stackexchange.com/users/3620/user3620 user3620
https://mattermodeling.stackexchange.com/users/3879/erwinschroedinger user3879
https://mattermodeling.stackexchange.com/users/3914/anya user3914
https://mattermodeling.stackexchange.com/users/3680/anya user3680
https://mattermodeling.stackexchange.com/users/3679/user3679 user3679
https://mattermodeling.stackexchange.com/users/3941/beginner user3941
https://mattermodeling.stackexchange.com/users/3912/m-todd-hess user3912
https://mattermodeling.stackexchange.com/users/3930/richard-m-bertrand user3930
https://mattermodeling.stackexchange.com/users/3981/holy-haptic user3981
User IDs between 4000-4999:
https://mattermodeling.stackexchange.com/users/4000/phy user4000
https://mattermodeling.stackexchange.com/users/4059/user1029384756 user4059
https://mattermodeling.stackexchange.com/users/4171/user4171 user41712
https://mattermodeling.stackexchange.com/users/4389/jeremy user4389
https://mattermodeling.stackexchange.com/users/4725/bewulf user4725
https://mattermodeling.stackexchange.com/q/8434/5 user4740
https://mattermodeling.stackexchange.com/users/4767/alexander-yoseff user4767(*)
https://mattermodeling.stackexchange.com/q/8507/5 user4772
Why extra lines in PDOS plot using xmgrace user4777
• Have I just missed a lot of unregistered users with IDs between 2000-3600 or is there some possible reason why we've had such a high density of unregistered users between ID 3600 and 3999? There have been bugs for user registrations in this network in the past.

1 [This user](https://mattermodeling.stackexchange.com/users/4169/ramesh-kumar) asked a follow-up question, which in it's original form looks like it was from the same person. 2 It's just a new account made by a user on Chem.SE who had their question migrated to MMSE: https://mattermodeling.stackexchange.com/a/6596/5 (you need access to moderator tools to see this link).

How are you searching for unregistered users? I haven't search extensively, but using this search tool, it would seem that userIDs 2292-3500 are unused, or at least a somewhat random search of values in this range returned no results. There may be some in this range that exist as 2292 was created in late March and 3500 was created in early May, and we did have new users join in the interim.

I could look into it more, but my guess is there was a run of spam accounts created over this stretch that existed long enough to get all/most of these userIDs, but they were subsequently mass deleted so no record of the users exists publicly (or for regular mods).

• "How are you searching for unregistered users?" I've been tracking them in a personal .txt file. I find it disturbing that we've had 11 unregistered users pop in the in the last 300 (3600-3900). I see that some of them seem to be people with their real names (and LinkedIn profiles), and that Anya accidentally created two unregistered accounts, and the example from today (holy-haptic) has a registered account at Physics.SE, but the batch from 1000-1999 also has people with real names (Todd Martinez, Paul Ayers, Alexandre Dias, Lee Cronin). Too many unregistered users in the 3600-3999 batch. Jul 20 '21 at 22:26
• @NikeDattani I think a lot of it is just people accidentally creating new accounts. user3620 is the same user as Nishan, both unregistered. Anya actually made 3 accounts including 3679. So about half of that 11 comes from 2 "real" users who just got confused by the account making process. And I haven't seen anything suspicious about the other unregistered users.
– Tyberius Mod
Jul 21 '21 at 0:18
• Between 3600-3999, the following unregistered users have made posts under a pseudonym which can't be matched to any real person or other registered SE account: phdstudent, erwinschroedinger, user3679, and beginner (4 of them). Between 0-3600, I only know of the following: Danish, MAAT, Lazarus, user1080 from the semen analysis question (4 of them). In a group of 7.5% we have 4 cases, which is the same as in a group of 92.5%, does the density of cases in the 7.5% group not seem unusual? Furthermore the early cases like Danish (user950) would have had little motivation to make fake accounts. Jul 21 '21 at 14:48
• @NikeDattani With small numbers involved and the site only existing for a year, it's tough to say what's a normal amount of unregistered users. We might expect the rate to go up as the site gets more active. The older numbers also may have shrank, as there has been more time for those users to have become registered. The general profile of these newer unregistered users also seems similar to prior ones: a user makes an account to post one question/answer and then basically disappears.
– Tyberius Mod
Jul 21 '21 at 15:32
• You say that as the site grows, so will the number of unregistered users, however I'm looking here at the density of unregistered users, over the range of 300 users vs 3600 users. You say the older numbers may have shrank, but this is contrary to my observation: 100% of the unregistered users in my database (in which users were added to my list, on the day they were created, and never touched since then) are still unregistered users (in other words, 0 of them became registered later). When you say the "general profile" of these newer cases seems similar to prior ones, what do you mean? Jul 21 '21 at 15:50
• @NikeDattani I think the rate would increase overtime as well. As the site gets more larger, I think the rate of new users will increase and consequently the rate of unregistered users (probably even more so as people find the site through Google and ask a one off question, rather than finding it organically and joining long term). By general profile, I just mean nothing has seemed weirder about these newer unregistered then the old ones. The questions and answers they have added haven't seemed spammy and they generally haven't stuck around afterwards. If they were trying to build up rep...
– Tyberius Mod
Jul 21 '21 at 16:03
• for some destructive purpose (e.g. get enough rep to downvote, vote to close, flag and then target users with these), they would presumably post more to earn rep and return after their initial posts. For someone to inflate their rep with say sockpuppet upvotes, they would need to have a lot more rep before they could do anything harmful.
– Tyberius Mod
Jul 21 '21 at 16:06
• It's also worth noting that SE overall has a problem with registered spam users. The barrier to register is very low and if you look through our user list and other sites on the network, you will find dozens of registered users that were just created to promote some company or service. SE has safeguards to mitigate the impact of these (e.g not counting these users towards certain statistics and not listing them with search engines), but if you are just going off the total user ID, you might get a skewed picture of how many real users we have.
– Tyberius Mod
Jul 21 '21 at 16:14
• Perhaps the rate would increase, in this case it seems to be a sharp increase from 3600-3999 vs 0-3600, so it really raised some flags for me. About the other things you're saying, I think you might find this conversation from approximately 1 hour earlier, to be interesting! Jul 21 '21 at 16:42