11
\$\begingroup\$

I just noticed that two extremely obvious non-answers posted within 20 minutes from each other are both upvoted.

https://electronics.stackexchange.com/a/450345
https://electronics.stackexchange.com/a/450346

I wonder if they are somehow connected or there is a sockpuppet or vote-bot of sorts here. Is there a way for me as a user to search for more instances of non-answers getting upvotes?

Since most users can't see deleted answers, here are a few of the shorter answers that have been receiving an upvote.

I tried this, but with resistor of different values. It didn't worked for me. Did this worked for you?

Hi i want 0 to 1000 volt input covert into 4 to 20 mA output circuit

My mumuso wireless Bluetooth headset is not connected on my device name lava z60. Please inquiry this.

\$\endgroup\$
12
  • 1
    \$\begingroup\$ The thing is, upvoting requires 15 rep. So fresh users can't be upvoting each other, there has to be someone who, at some point, managed to acquire some rep. In any case, searching for non-answers is going to be difficult since those posts get deleted pretty fast and you therefore can't search for them in data explorer. \$\endgroup\$
    – dim
    Jul 29 '19 at 7:56
  • 3
    \$\begingroup\$ @dim After thinking a bit I wonder if it's a "helpful" high-rep user who wants everyone to just get along and be nice and post whatever they want, and thus hands out "compensatory upvotes" whenever they see a downvoted post. This is a problem because it gives users 10 ill-deserved reputation points which they can then use to cause even more disruption.. \$\endgroup\$
    – pipe
    Jul 29 '19 at 7:58
  • 1
    \$\begingroup\$ @pipe It could be, but it seems a bit far-fetched to me. It doesn't add up if we apply Hanlon's razor, which I'm a strong advocate of. \$\endgroup\$
    – dim
    Jul 29 '19 at 8:08
  • 2
    \$\begingroup\$ Maybe non-answers are upvoted because.... "Hey, at least you tried" or "Welcome to EE.SE and since this is your first answer I'll upvote to promote answers", hard to deduce. Are you sure that this is limited to non-answers only? Maybe it's bad answers according to the one doing the upvoting. Or maybe it's only the first answer ever made by a user. \$\endgroup\$
    – Harry Svensson
    Jul 29 '19 at 16:07
  • \$\begingroup\$ @HarrySvensson I'm not sure, hence my question about trying to search for it, I haven't the slightest idea where to start! I understand that there are some ways for users to perform more advanced queries though. \$\endgroup\$
    – pipe
    Jul 29 '19 at 16:16
  • 9
    \$\begingroup\$ Not just non-answers, also badly broken, unresearched, incomplete questions. We've got a joker in the house, or someone who really doesn't understand SE sites and is handing out "thanks for playing" votes - perhaps someone who transferred with rep from one of the non-technical SE sites where they have a tradition of doing that based on collective non-understanding of the underlying mechanism. \$\endgroup\$
    – Chris Stratton
    Jul 29 '19 at 19:31
  • \$\begingroup\$ I am, I imagine, @pipe's idea of a "helpful" [tm] high rep user in this context :-). It weren't me, sir, nor would I upvote such answers, ever. fwiw. \$\endgroup\$
    – Russell McMahon Mod
    Jul 30 '19 at 12:00
  • 1
    \$\begingroup\$ FWIW, I also noticed a couple such upvoted non-answer in the past few months. They weren't obviously connected and they were far apart in time, so I simply said to myself "who's the idiot voting such non-answers?!?" ("idiot" because they were blatant non-answers, not because they were just bad answers) and flagged for "not an answer". Now that you bring it the fact to my attention again, I think maybe we have gone past the "it's a coincidence" stage. \$\endgroup\$
    – Lorenzo Donati -- Codidact.com
    Jul 30 '19 at 23:21
  • 4
    \$\begingroup\$ Since the search for those events is difficult, probably we should keep track of them manually when we notice one. Maybe make a community wiki answer here where we could add links to those answers. If there is something fishy behind them, maybe we can track the problem down and the mods can report the thing to SE staff. \$\endgroup\$
    – Lorenzo Donati -- Codidact.com
    Jul 30 '19 at 23:22
  • \$\begingroup\$ There isn't a moderation system for upvotes... \$\endgroup\$
    – Voltage Spike Mod
    Aug 8 '19 at 17:59
  • \$\begingroup\$ You think that's bad; Last Friday I got 34 of my answers downvoted by some pleb. This follows weeks of getting 3 to 6 regular daily downvotes for seemingly clear and uncontestable answers. This place needs sorting out. \$\endgroup\$
    – Andy aka
    Sep 23 '19 at 7:21
  • \$\begingroup\$ @Andyaka Talk to the mods about that, that is definitely unacceptable behavior, if it came from one user, they might be able to fix it. The other problem is traffic is down, and that makes upvotes go down also. \$\endgroup\$
    – Voltage Spike Mod
    Oct 10 '19 at 17:36
8
\$\begingroup\$

Lorenzo's idea of a community Wiki to track them probably isn't a bad idea. Here are some I've spotted, so feel free to add:

\$\endgroup\$
12
  • 4
    \$\begingroup\$ The fact that all of these answers are deleted pretty much demonstrates this to be a non-issue, baked into the cake of SE tools and policies already. \$\endgroup\$
    – Scott Seidman
    Jul 31 '19 at 13:52
  • 9
    \$\begingroup\$ @ScottSeidman while the deletion demonstrates there is a mechanism for cleanup, the fact that these are getting upvotes (including in cases of literal SPAM) is itself evidence that someone is abusing this site. \$\endgroup\$
    – Chris Stratton
    Jul 31 '19 at 15:48
  • 2
    \$\begingroup\$ What is the problem here really? Bad questions and bad answers, or someone upvoting them? - There's also a slight problem, upvotes and downvotes are anonymous, and here's a problem where someone was downvoting instead of upvoting. If any moderator can find out who it is, then that means that it is not anonymous, this means that moderators have been lying -> lose some of our trust. \$\endgroup\$
    – Harry Svensson
    Jul 31 '19 at 17:50
  • 3
    \$\begingroup\$ Let me rephrase what I think that the problem really is here. - To me this just looks like brain tumor in a patient having a nosebleed. Do you A, treat the tumor (bad questions, bad answers). Or do you B, focus on the nosebleed (someone upvoting the bad questions and bad answers). You can remove the nosebleed all you want, but the underlying problem is still there. If you however remove the tumor you'll also remove the nosebleed. \$\endgroup\$
    – Harry Svensson
    Jul 31 '19 at 18:13
  • 3
    \$\begingroup\$ This is a virtual ball on a metaphorical lawn. \$\endgroup\$
    – Scott Seidman
    Aug 2 '19 at 14:34
  • 2
    \$\begingroup\$ At the point of someone upvoting spam this probably should be escalated to whatever level of authority allows seeing who was doing the voting. (Not sure if moderator tools have this ability, or if they need to get someone working for Stackexchange to do the digging.) \$\endgroup\$
    – Dan Is Fiddling By Firelight
    Aug 5 '19 at 20:41
  • \$\begingroup\$ @DanNeely: While moderators can see some limited information on "anomalous voting patterns", I'm pretty sure they can't see enough detail to investigate this. Unless, of course, somebody has already programmed the script that tracks voting patterns to report "multiple upvotes on answers that get deleted as spam/offensive" or something similar as anomalous. \$\endgroup\$
    – Ilmari Karonen
    Aug 8 '19 at 20:03
  • \$\begingroup\$ I'm assuming all the links in this answer are going to non-answer answers, I'm not 100% sure since I can't see deleted answers. Chris Stratton said that questions got upvotes too, "Not just non-answers, also badly broken, unresearched, incomplete questions.", so shouldn't there be some links to those questions? - We're making a list of bad questions and bad answers, imagine if we had a better system that could detect comments in answers and whatever that makes an answer a non-answer. Then this list wouldn't be here. \$\endgroup\$
    – Harry Svensson
    Aug 9 '19 at 16:27
  • 2
    \$\begingroup\$ @HarrySvensson Correct, purely non-answer-answers. You can at least see that they're answers from the URL (…/a/…). I don't think linking to questions is a good idea because that's much more subjective. No one can claim in any way that the answers listed in this community wiki are answers so there can't be any discussion about that part. \$\endgroup\$
    – pipe
    Aug 10 '19 at 9:25
  • 1
    \$\begingroup\$ @SamGibson I'm not sure it is useful to maintain this. It's gonna grow and grow, but will it enable anyone to find out the cause at some point? \$\endgroup\$
    – dim
    Oct 11 '19 at 11:56
  • \$\begingroup\$ @dim, Hi, If the consensus is to stop adding more examples, then of course I will stop adding them. "will it enable anyone to find out the cause at some point?" I have no idea of the inner SE workings of what is/isn't possible, with the evidence of these examples. It wasn't me who started collecting them. Perhaps we do have enough evidence now, but if we stop adding examples, how can we see if this anomalous voting pattern is continuing? Anyway, as I said, if there's a consensus to stop adding examples, then I'll stop. Thoughts? \$\endgroup\$
    – SamGibson Mod
    Oct 11 '19 at 15:17
  • 1
    \$\begingroup\$ @SamGibson I don't really know. I just think that, most probably, nobody will ever have a look back at this, and that you may have better things to do (but it's not up to me to judge). \$\endgroup\$
    – dim
    Oct 11 '19 at 15:20
2
\$\begingroup\$

FWIW, a google search for "stack exchange meta bad upvotes" yields dozens of hits, dating to the birth of the platform, most leading to conversations much like this one. My favorite answer is https://math.meta.stackexchange.com/a/15145

The help center points out that reputation is an approximation. Efforts to make it more than that seem misguided.

To a certain extent, xkcd captures this beautifully:

enter image description here

Of course, this is hyperbolic, and SE does need to convey the accuracy or correctness of answers, but you won't be able to convince me that mechanisms for that are seriously broken by the observations in this thread.

\$\endgroup\$
21
  • 2
    \$\begingroup\$ It's not the mechanism that is broken, it is one or more users who are (to use your word) "broken" in that they are either acting in bad faith or with poor understanding of the mechanisms of SE. \$\endgroup\$
    – Chris Stratton
    Aug 2 '19 at 17:53
  • 1
    \$\begingroup\$ ... and the tools to deal with them are more than in place. There's the simple downvote to balance things out (like has been employed on my answer here, and I've applied to others in the the thread), there's the "not an answer" flag, and there's whatever tools the platform has to sniff out sock puppets. I don't believe a spurious upvote will have ANY impact on any user that sees an answer so treated. Again, no problem to be solved. \$\endgroup\$
    – Scott Seidman
    Aug 2 '19 at 18:35
  • 3
    \$\begingroup\$ Again, the problem is user misbehavior not site mechanisms. You are talking about addressing the symptoms, and everyone agrees that there are countermeasures. What you are still failing to understand is that this particular page in meta is about the observed issue itself, not the countermeasures to it with which we are all familiar. It's not a bug report or a demand for action it's an observation of fishy behavior. \$\endgroup\$
    – Chris Stratton
    Aug 2 '19 at 18:43
  • 1
    \$\begingroup\$ I agree with you that the site has already mechanism in place to deal with random blatantly "bad" upvotes. The problem I see, and I think that is what worries other people, is that I suspect that we are dealing with some user(s?) that acts in a non-random way. As I already mentioned in some other comments, in the last few months I've seen this behavior a bit too much to be considered just "random noise" form ignorant users. ... \$\endgroup\$
    – Lorenzo Donati -- Codidact.com
    Aug 4 '19 at 11:31
  • 2
    \$\begingroup\$ ... What I fear is that some user is up to some kind of "low-threat vandalism" which maybe could escalate if not looked into. Until I saw other people complaining about that, I just thought I was a bit too paranoid, but now maybe we have some clues. My point is: if there really is some specific user that is abusing the system on purpose, it is worth to act proactively before he finds ways to do more damage or sets an example for other "jokers" out there. \$\endgroup\$
    – Lorenzo Donati -- Codidact.com
    Aug 4 '19 at 11:31
  • 1
    \$\begingroup\$ ... The site has no specific facilities to help community differentiate between random bad behavior and willful bad behavior of this kind, and that's ok. This doesn't mean we shouldn't try to track down the "culprit(s)" and report them to "authorities". In short, background noise is normal and not worth worrying about too much, active jamming interference is not. \$\endgroup\$
    – Lorenzo Donati -- Codidact.com
    Aug 4 '19 at 11:31
  • 1
    \$\begingroup\$ Lest I give the wrong impression: I don't want the mechanism for dealing with bad upvotes to become stronger. I just suspect the currently observed behavior could be some very specific user (or maybe few users acting together or maybe on their own). So mine is a call for specific action, not a call to reform the site usual policy and infrastructure. I'm not calling for a site-wide witch-hunt! The point is: "keep an eye on this apparently willful behavior and take note about it" so that if it turns out more virulent than it is we have the information log needed to report to SE staff promptly. \$\endgroup\$
    – Lorenzo Donati -- Codidact.com
    Aug 4 '19 at 11:40
  • 1
    \$\begingroup\$ Perhaps the evildoer also jaywalks. \$\endgroup\$
    – Scott Seidman
    Aug 4 '19 at 11:47
  • 1
    \$\begingroup\$ BTW, I fear that sock-puppets detection algorithms try to detect people trying gaming the systems, not doing this sort of "vandalism". I stand to be corrected, but I doubt SE staff has programmed the system to detect a single user that regularly and willfully upvotes non-answers of low-level users. \$\endgroup\$
    – Lorenzo Donati -- Codidact.com
    Aug 4 '19 at 11:48
  • \$\begingroup\$ I keep picturing Bill Murray in Caddyshack, destroying a golf course to get rid of a gopher. \$\endgroup\$
    – Scott Seidman
    Aug 4 '19 at 11:52
  • \$\begingroup\$ "Perhaps the evildoer also jaywalks." well, in Italy you could get fined for jaywalking and also get penalties on your driving license, if you have one (even if the "crime" has been done not while driving). :-) \$\endgroup\$
    – Lorenzo Donati -- Codidact.com
    Aug 4 '19 at 11:55
  • \$\begingroup\$ While it's not me, what would happen if I started giving participation upvotes? Which user agreement would I be violating? \$\endgroup\$
    – Scott Seidman
    Aug 4 '19 at 12:12
  • \$\begingroup\$ google.com/url?sa=t&source=web&rct=j&url=https://…. There are upvote pattern detectors \$\endgroup\$
    – Scott Seidman
    Aug 4 '19 at 12:18
  • 1
    \$\begingroup\$ @LorenzoDonati What is the problem here? Someone upvoting non-answers, or non-answers? If you can tell that an answer is a non-answer, then surely some machine learning can be taught that as well. \$\endgroup\$
    – Harry Svensson
    Aug 5 '19 at 20:20
  • 1
    \$\begingroup\$ "Engineers don't make efforts to make "good enough" better than it needs to be." Fair point, indeed. \$\endgroup\$
    – Lorenzo Donati -- Codidact.com
    Aug 5 '19 at 21:17
-3
\$\begingroup\$

The problem with upvotes on useless postings is not confined to spam or machine generated nonsense, it also extends to entirely useless posts that add nothing but noise, yet seem to garner "thanks for participating" upvotes from one or more users who do not understand that upvoting indicates an answer is useful.

For example,

https://electronics.stackexchange.com/a/451314/2246

Give Delay HAL_Delay(300) and check

At best that is a comment, but even as a throwaway guess it is pointless - substituting one small 16-bit number for another actually present in the question will not yield a different result, and there is not even any argument made for why the proposer thinks it should.

\$\endgroup\$
3
  • 3
    \$\begingroup\$ I'm still not seeing a problem. This is why users get down votes as well as up votes. Where particularly egregious, theres a "not an answer" flag mechanism. If anything, I suppose there's an argument to be made that deleting answers can be added to the moderator tool privilege to unload mods. \$\endgroup\$
    – Scott Seidman
    Aug 2 '19 at 17:20
  • 1
    \$\begingroup\$ The problem is with the misuse of upvotes. In neither this case nor the others mentioned on this page is it an insurmountable problem, but it is a case of one mechanism malfunctioning in the user base, even if that malfunction is being compensated by other mechanisms. \$\endgroup\$
    – Chris Stratton
    Aug 2 '19 at 17:22
  • 4
    \$\begingroup\$ It's not like only one person is allowed to vote. Poor upvotes and downvotes are well accounted for in the stack exchange model. People post answers, and community members tell the community how they feel about them. There's no guarantee that any individual vote has real meaning. In fact, new users can't even see anything other than the total! As problems go, this one doesn't hold up to a Pareto analysis. \$\endgroup\$
    – Scott Seidman
    Aug 2 '19 at 17:28

You must log in to answer this question.

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