25

There is a system-wide plan to test lower voting threshholds. The full details are here on Meta SE. In brief:

  • Reputation required to cast upvotes is changing from 15 to 1
  • Reputation required to cast downvotes is changing from 125 to 1
  • The 1 reputation cost to downvote answers will be removed

I propose that Retrocomputing should not join the low rep voting experiment.

  • Upvote this question to agree that it's a bad idea to join
  • Downvote this question to suggest that we should join the experiment

The powers that be are asking for sites to volunteer. I just want to state for the record here that I am very much against this idea. The numbers can (and maybe should) be tweaked a bit, but allowing totally new users to vote without any other interaction is, in my opinion, a disaster waiting to happen. I encourage others to upvote this and/or to write an answer with more coherent arguments, so that it can be clear where this site stands on this policy.

And if this is the wrong way to do this, let me know. But since they are looking for sites to volunteer for testing, I think it is important to discuss.

18
  • Downvoted due to my understanding of meta-stack convention - downvote to the question is to be against the proposal (not the question) - right?
    – davidbak
    Sep 20 at 17:51
  • 1
    Up vote supports my position (against volunteering to test the new scheme) downvote against me and in favor of new scheme.
    – manassehkatz-Moving 2 Codidact
    Sep 20 at 17:53
  • 1
    oh, well in that case - maybe reword the title to make it clear what the proposal is? that we do not join the test?
    – davidbak
    Sep 20 at 17:54
  • Just saw the title changed. Original title was different.
    – manassehkatz-Moving 2 Codidact
    Sep 20 at 17:55
  • 1
    Edited question to make it clear (at least clear to me).
    – davidbak
    Sep 20 at 17:58
  • 1
    @manassehkatz-Moving2Codidact You can edit the title back if you would like; I didn't realise you wanted the votes on the question to be used like that.
    – wizzwizz4 Mod
    Sep 20 at 18:41
  • 2
    @davidbak That’s why it’s better to phrase the question neutrally, and post own stance as an answer, so that both can be voted on separately.
    – user3840170
    Sep 21 at 11:37
  • 2
    Also, now that the title contains a negative, it’s going to make some forms of answers clunky and/or confusing. (Not to the same extent as ‘Should we not decline to refuse to avoid failing to stay away from skipping the experiment?’, but still…)
    – user3840170
    Sep 21 at 12:27
  • @user3840170 - but that's not the way voting on the meta stacks works at SE. In meta, voting on a proposal question isn't about "showing research effort" regardless of what the tooltip says - it's about for or against the proposal. This is the policy across all SE meta sites. (I wish they would make the tooltip on the vote arrows on meta sites reflect the actual policy, but apparently that's lower priority than other initiatives, like launching a bespoke chatgpt prompt stack.)
    – davidbak
    Sep 21 at 17:30
  • 1
    (see voting on meta - this makes it look like "sometimes" its done to voice approval/disapproval but in fact it's practically universal over at meta.stackexchange.com and also meta.stackoverflow.com, at least for feature requests and also for responding to SE initiatives)
    – davidbak
    Sep 21 at 17:35
  • 1
    Now on the topic itself, the existing barrier if 15/125 is already extreme low. That 15 can be reached even before finishing the tour. Likewise one or two not totally wrong answer - or a C64 related question - will bump anyone easy past 125 points. I do see the merits of having a low hurdle to involve new members, but regarding downvotes I would love to see the threshold increased instead.
    – Raffzahn
    Sep 22 at 19:39
  • 1
    @davidbak Opinions vary about whether poll answers are good. Note that those from the Retrocomputing crowd who contributed to that Q&A seem to be on the “poll answers are good” side of things, and we've done it that way more than once in the past.
    – wizzwizz4 Mod
    Sep 22 at 20:25
  • 1
    As a side note, the most likely to be in favor or lower voting thresholds would be the newest users. Who are also least likely to stumble across RC Meta to voice their opinion here... So there is some selection bias in force, but considering that, for example, there are 153 users with a 250+ rep change this year (so therefore active to at least a moderate degree) and only 8 votes so far, there is plenty of room for others to make their voice known if they disagree.
    – manassehkatz-Moving 2 Codidact
    Sep 22 at 20:30
  • 2
    A corollary is that those most likely to be in favour of lower voting thresholds (the newest users) are the least likely to have figured out community norms.
    – another-dave
    Sep 23 at 20:36
  • 3
    @Raffzahn Using upvotes and downvotes as a means to poll for yes/no answers is a long established idea on Stack Overflow. Downvotes don't mean the same on meta sites. They are usually taken as disagreement with the question or answer, not that it's bad necessarily.
    – JeremyP
    Sep 28 at 8:20

4 Answers 4

Reset to default
17

SE:RC is a community of sorts. It has norms, more or less. It has a consensus, more or less, about what makes a good question, or a good answer.

The 'reputation' requirement for voting on the questions and answers of other members is a proxy (and IMO a reasonably good one) for longevity of membership and for seriousness (I don't much like my own wording here but it'll have to do for now) of interest in the subject.

So: NO to low-reputation voting.

10

15 isn't very different to 1. It indicates minimal participation in the sense you may have contributed a question or answer. If anything, they should be raising the threshold in my opinion.

Down votes need to have a cost. Otherwise, revenge downvoting will become a thing.

Regardless of the above and whether or not I agree with the objectives, I don't think Retrocomputing should get involved in any experiments of this nature. The community is fairly fragile as it is and getting involved in experimental tweaks might be suicidal. If they thought it was definitely going to improve things, they would just implement it without experiment.

This site is not a lab animal to be sacrificed so that Stack Overflow doesn't do something stupid.

2
  • 5
    I fully agree with the first two paragraphs. And while I do not see RC.SE as fragile, I would as well appreciate to not introduce more issues than we create on our own :)
    – Raffzahn
    Sep 28 at 11:32
  • 2
    There are already mechanisms against revenge downvoting. I won't go into too much detail, but cost-free downvoting would not make such behaviour particularly hard to detect and reverse. (Note: downvotes on questions are currently free.)
    – wizzwizz4 Mod
    Sep 29 at 18:36
4

I think this experiment would be bad for our site. However, the site is small enough that we could clean up after it goes wrong, and we have enough expertise to mitigate the harm in the meantime. If the experiment reaches a positive conclusion, we might just have low-rep voting. (This will hopefully not be sprung on us, but you never know.) So if it really is as bad as we expect, it might be better to find out early by participating in this experiment.

Alternatively, we could propose a different experiment to be run on this site. If there's something we're more enthusiastic about trying, and it's not too hard to implement, I'm sure they'd let us.

8
  • I'm sure they'd let us Really?
    – manassehkatz-Moving 2 Codidact
    Sep 20 at 17:17
  • 2
    Probably. I mean, something like “10k users should get a double upvote” might not fly, but they'd change most of the numbers on this page if we asked nicely, and I think they'd consider other things too.
    – wizzwizz4 Mod
    Sep 20 at 17:19
  • 3
    I'd like to run an experiment here where we left long comment threads around. (I mean we tend to do that, but sometimes they do get moved to chat where they die.) It's just the way this particular stack works. Not to unfairly generalize from me to all of you - but I don't think it's just me that's a garrulous old geezer contributing useful and relevant information in the form of comments. (Though maybe it is.)
    – davidbak
    Sep 20 at 17:50
  • 1
    As a moderator I'm wondering if you see interesting behavior from low-rep visitors to retro - i.e., newcomers to the site. The obvious is that a lot of newcomers don't really understand the limits of "retro" and ask questions inappropriate for the site. And that includes the edge-case of asking about old video games (discussed here recently).
    – davidbak
    Sep 20 at 18:13
  • 1
    @davidbak It's definitely not just you. Unfortunately, there are many problems with using comments that way (not least that the presence of certain keywords allows drive-by deletion by unprivileged users): sometimes the chat migration feels like a lesser evil. Though… if we ask really really really nicely, we might be able to get some comment-related features cribbed from Discussions (a recent Stack Overflow Collectives thing).
    – wizzwizz4 Mod
    Sep 20 at 18:14
  • 1
    I've noticed that sort of behaviour a bit. Moderators won't have particular insight into that (we mostly handle flags, meta, and talking to the company… and, in a past life, tagging), so I'll leave that question open to others. (Or you could start a new meta question, if you actually want an answer / detailed discussion.)
    – wizzwizz4 Mod
    Sep 20 at 18:16
  • 1
    @davidbak re comments - agree; there is often much value and certainly much of interest in the discussion. The Q, in this garrulous old geezer's opinion, is merely the prompt.
    – another-dave
    Sep 20 at 23:50
  • 1
    @davidbak It's not your suggestion, but your (and another-dave's) comments finally got me to post Do we want Discussions functionality on Retrocomputing?.
    – wizzwizz4 Mod
    Sep 21 at 21:35
2

Removing the barrier entirely is a recipe for disaster, but we should trial reduced barriers to voting. Here's a specific proposal.

  • You should be able to upvote if:
    • you have 15 reputation; or
    • you have the [• Teacher] badge on main or meta; or
    • it's an answer to your own question (and, obviously, not a self-answer).
  • You should be able to downvote if you can upvote.
  • Downvotes should cost nothing, unless you have an answer to the same question (or post one within 12 hours), in which case there should be a 1 reputation penalty for downvoting.

Where parts of this proposal are difficult to implement (e.g. cross-site badge detection; retroactive reputation penalty within 12 hours), they need not be implemented for the trial.

5
  • I don't see how Teacher makes that much of a difference - that will already get you 10 points, so I'd leave that out. But I am concerned about potential for abuse, so I'd rather see at least 25 points (maybe even 50) for Downvote, though perhaps 25 for Up or Down might work. Of course, right now someone comes in with 100 if they've been active elsewhere (which should at least show they have some general SE experience) and can then immediately Upvote but not Downvote - with this plan they could Up or Down vote immediately if they came in with that magic 100.
    – manassehkatz-Moving 2 Codidact
    Sep 20 at 20:26
  • 1
    @manassehkatz-Moving2Codidact If you think the magic 100 is a concern, there might be scope for adding 100 to the thresholds for users with the association bonus (as is done for protected questions). That wouldn't work on meta, mind.
    – wizzwizz4 Mod
    Sep 20 at 20:27
  • Not worried about Meta.
    – manassehkatz-Moving 2 Codidact
    Sep 20 at 20:27
  • 1
    @manassehkatz-Moving2Codidact The motivation behind the “Teacher” requirement is two-fold: you can get rep for asking or editing (neither of which necessarily indicate domain knowledge, so we perhaps shouldn't reduce the threshold to 10 rep), and you can get upvotes on meta (which suggests you probably at least get the concept of curation). Might not be sensible, though; one potential issue is that badges don't go away, but maybe that's a good thing here.
    – wizzwizz4 Mod
    Sep 20 at 22:47
  • 1
    If nothing else, I support removing the downvote rep cost, perhaps with the provisos mentioned. I realize downvoting can be abused, but think a blanket cost for using it isn't nuanced enough.
    – Jim Nelson
    Sep 25 at 19:14

You must log in to answer this question.

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