A gold tag badge requires 1000 votes in at least 200 posts. Could we not fast-track the rights that relate to such badges?

The current rate at which people (especially non-moderators) acquire these badges is not such fast that we get a reasonable amount of (non-moderator) superusers in specific areas of expertise:

gold tag badges Note: year has been calculated from the date as $yyyy+(mm-1)/12+(dd-1)/365$

  • 10
    I don't think we'd get SE to change the criteria for awarding gold tag badges (see meta.stackexchange.com/q/295707/225179 & meta.stackexchange.com/q/69689/225179), or to grant the corresponding privileges on any other basis. But if we want a moderator election I'm sure they'd arrange one for us. – Scortchi Sep 12 at 16:00
  • 5
    It's pretty hard to get a gold badge, unless one answers a lot. I have 3k rep in the [pca] tag but "only" 170 answers... – amoeba Sep 12 at 17:21
  • 7
    The Department of the Obvious wishes to point out that receiving up-votes and writing answers are the sole path to the gold badge privileges. I agree with Martijn and amoeba that it is a long road; I am currently about half-way to gold badges for both the neural-networks and machine-learning fronts. More active voting can help; so can concerted efforts to write canonical answers and close duplicates with the canonical threads as targets. – Sycorax Sep 12 at 17:28
  • 6
    Tangentially, the gold badge privileges crucially hinge on threads being appropriately tagged. Hence, editing questions to tag questions appropriately is very valuable in this respect. Some users undertake to edit many such questions, and are doing CV.SE a great service. – Sycorax Sep 12 at 17:33
  • 1
    I would as well not expect the people that run SE to change the badge system that easily. However, I do think that the question here has a slightly different angle than the two previous questions at the general stack-overflow meta. Or at least there is a new argument which is about the role of tag badges in the management/moderation of the sites. I care much less about the other badges which are just shiny additions for people's trophy cases. – Martijn Weterings Sep 12 at 17:33
  • 2
    Although I believe we will eventually take the more pragmatic road, I hope to see that we can discuss this topic also more freely (not necessarily only pragmatic with the possible restrictions of the SE owners in mind), or at least bring in new ideas or show the current 'public opinion'. It would stop possible useful discussion before it can start if we always assume beforehand that SE is not gonna change anyway. – Martijn Weterings Sep 12 at 17:36
  • 2
    I didn't mean to imply it wasn't worth discussing changes to the badge system; just to point out a path of least resistance. Your graph does make a compelling case, in that acknowledged experts in their fields are taking ages & many posts to be awarded gold tag badges. – Scortchi Sep 12 at 18:46
  • 3
    Note that editing a thread to add the right tag does not confer the ability to use the gold tag badge duplicate hammer. The tag needs to have already been there. – gung Sep 12 at 20:35
  • 3
    It's also worth considering the history of the SE system. It started w/ SO, which really was the model everything was built around. On SO, it just isn't that hard to write 200 answers about R, Python, C++, Java, etc. It's more of an issue to get to 1k upvotes, since upvoting is much lower there, but in general I think gold tag badges are much more doable on SO than on most of the smaller sites. – gung Sep 12 at 20:44
  • 5
    @gung I fully agree about SO. I find it annoying to have enough rep but not enough answers for the gold badge. It means in principle I could quickly write 30 crappy answers in [pca], and even if they all remain with 0 score or even downvoted, I would get the gold badge. This is a weird incentive. – amoeba Sep 12 at 21:40
  • 4
    @amoeba, I'm pretty sure. When I had a gold tag badge, but before I was a mod, there was a situation where someone posted a question & I closed it unilaterally as a duplicate. The OP promptly deleted the thread & reasked the Q, verbatim, but w/o the tag. I edited the thread to add the tag, but couldn't close unilaterally. – gung Sep 13 at 0:44
  • 4
    Also, I actually think you should just go ahead & post another 30 answers to [pca] Q's somewhere, even if the answers are brief & middling. The site would be better off if you had the tag badge. – gung Sep 13 at 0:48
  • @amoeba: note that it's not 1,000 reputation, but 1,000 upvotes you need. The difference is roughly a factor of 10, last time I looked... – Stephan Kolassa Sep 14 at 13:21
  • @StephanKolassa Oh yes, thanks. I do have 3k upvotes in [pca] though :) That's what I meant. – amoeba Sep 14 at 13:33
  • 2
    data.stackexchange.com/stats/query/60719/gold-tag-badges might be easier for next time – Glen_b Sep 15 at 7:50

I think you ask two questions, which I will paraphrase.

  1. Can CV.SE get special dispensation to change gold badge requirements?

The only people who have the power to do this are the folks who run Stack Exchange. You can raise it with them, but I doubt they would agree, just because they won't want to open the floodgates and take on management of dozens of different domain-specific modifications.

  1. Does CV.SE need a moderator to specifically moderate the growing number of questions in or ?

I think there's a compelling case to be made here; the interest in these topics is growing. We have some dedicated users who answer these questions, but it is sometimes the case that threads bearing these tags are left to languish in the sense that their are unanswered, or answered poorly, because they are unanswerable in their breadth or lack of clarity.

Or, the same questions are asked over and over but are not closed because it is challenging to identify the duplicates, or the closure effort stalls at 3 or 4 votes, just shy of the 5-vote threshold. Stated another way, the specialist knowledge required to recognize and close as duplicates and questions is not present among a quorum of users who are active in the closure queues. (This analysis assumes that the other reviewers are voting to "Skip" rather than register "Open" or "Close" votes; this isn't necessarily the case, though. Perhaps only 2 or 3 other users agree with me in the votes which fail to achieve quorum, in which case the VTC queues are working as intended!)

None of this is intended as a criticism of our current moderators. I believe that our current moderators are knowledgable, fair-minded, and even-tempered. I only mean to suggest that there is a small gap in moderator coverage pertaining to these particular topics.

A moderator would have unilateral power to close threads and mark as duplicates.

My understanding is that for a moderator election to happen, the SE website merely needs to make a request to the SE community team, and an election will take place.

  • 1
    I did not really ask those two questions explicitly/directly, or it was not my intention. I do however agree that these questions are part or the issue that I raise (the fact that the mechanism of golden tag badges and the privileges associated with them are not doing their job as CV.SE). Machine-learning and neural-networks are one thing. Principal component analysis, which amoeba mentioned, is another thing (although in that place the issue is less problematic due to the topic being more well known among users and moderators). – Martijn Weterings Sep 12 at 17:58
  • 2
    Possibly I should have stated those questions more explicitly on machine-learning and neural-networks specifically. However somehow I wanted to frame the question more in this general way about the gold tag badges and the mechanism how SE provides more privileges to other uses. I feel we should at least go past this route before we think about moderators, which I feel is more like a work-around and not really like a solution to the underlying problem. – Martijn Weterings Sep 12 at 18:01
  • @MartijnWeterings and of course we could have another election and the community could choose someone with zero knowledge of machine learning and neural networks. – mdewey Sep 13 at 8:09
  • 1
    @mdewey I do not think that the community that asks for an election, because they feel there is a need of moderators with different expertise and specifically machine learning and neural networks, will choose someone with zero knowledge of machine learning and neural networks. another thing is whether the election is gonna have a lot of choice anyway. – Martijn Weterings Sep 13 at 8:14
  • 1
    @MartijnWeterings well you have more faith in elections than I do although perhaps your country is more sensible than some anglophone countries I could mention. – mdewey Sep 13 at 12:40
  • 2
    @mdewey that is a very cynical approach to elections. I do not think that elections here would go in the same way as elections for politics. I think that it is not possible, or at least not likely, for an insufficiently competent person to become moderator here. (although that doesn't mean I would be in favor of more moderators, I guess that superusers would be enough) – Martijn Weterings Sep 13 at 16:03
  • 5
    From my observations of previous moderator elections, my view is that the candidates have ample technical knowledge. The main consideration that I make is that we want moderators who are fair-minded and even-tempered; this is the larger part of how I make my votes. For me, at least, it would be a shift to look towards the moderator's technical knowledge as the primary consideration; currently, I look to interpersonal skills first. – Sycorax Sep 13 at 16:08

More rights associated with silver tag badges

One possible option to increase users with more closing rights would be to use the silver tag badges as well. (although it requires to change the website)

This would make sense because it is in line with the intentions and argumentation for the introduction of rights for the gold tag badges but with a different metric/focus (currently, StackOverflow has awarded 7.9k gold tag badges, CrossValidated has awarded 127 silver badges).

Disclaimer: This may be gaming the system to an unacceptable degree. It seems like a grey area to me, so I offer this for your consideration.


If there are particular tags for which we:

  1. do not have a gold badge member and think it would be beneficial to have one,
  2. have a member who's clearly on the path to achieving it, and
  3. is limited by upvotes rather than by number of answers

We could, as a community, choose to seek out and upvote the member's answers on that tag (if we think they are helpful, of course). This is leaning on the scales a bit by specifically visiting a set of questions and using our voting power. It's a form of bias, but I think it's worth considering because it is biasing our attention rather than our assessment of quality.

A few votes a day by many of us for one member would probably achieve the goal in days to weeks.

  • 4
    Be aware that some vote bulk up/down-vote actions are automatically reversed, and other kinds of misbehavior around voting can result in permanent bans. I don't know much about the policies around this, but I would tread carefully. – Sycorax Sep 13 at 23:21
  • @Sycorax Thank you, I agree it's borderline and don't plan to undertake anything unilaterally. I bring it up here for discussion. – mkt Sep 14 at 5:38
  • 2
    I would feel uncomfortable about voting for posts which lie outside my range of interests even in a good cause. However, if @Sycorax were approaching the reuqired number of posts and short of rep would it work if one of us awarded a bounty on one of the answers (like the one about why my NN does not work)? Does anyone know if bountied reputation counts for the tag? – mdewey Sep 14 at 11:28
  • 4
    @mdewey, it's not rep, it's the net upvotes on answers for the tag. – gung Sep 14 at 12:04
  • 2
    +1 As I have argued elsewhere, high-rep users should generally upvote more on CV, even if an answer is not as elegant as it can possibly get, so upvoting more on posts of other high-rep users who tend to give good answers sounds like a good thing. – Christoph Hanck Sep 25 at 12:06

Adding to other answers, we need to employ tags better. For example: What is the difference between a neural network and a deep neural network, and why do the deep ones work better? does not contain the tag . It could, and in my opinion, should. The answerers did not receive points toward a badge answering a question that's is obviously about a part of machine learning.

  • 8
    Ooooh boy, the machine-learning tag is a bit of a mess... I agree with the sentiment that tags need to be correctly applied, but I think the machine-learning tag has its own issues which should be addressed separately; I do not think that it is a good example for tag hygiene because its usage widely deviates from its description in the tag wiki (suggesting that either the wiki or its usage need to be revised) – Sycorax Sep 13 at 16:46
  • @Sycorax Well, I mentioned machine-learning because it is mentioned by OP. I see not guidance in the tag wiki though. – Firebug Sep 13 at 17:25

We should not need to get more people with more rights.

The rights for gold tag badges originates from an idea that came from StackOverflow. That site has currently 80 times more questions per day (6.6k vs 79 q/day) with a moderator team that is only four times larger (28 vs 7).

On Cross-validated we should be able to manage the moderation of questions sufficiently without technological solutions (changing the website). If it takes too long for questions to pass trough review queues then 'to give certain people more power' would be more like a work-around, rather than a direct solution. In principle we should aim to have consensus decisions rather than a few judges.

  • 2
    Admittedly, only 6 of our moderators are regularly active these days. Also, as far as closing questions is concerned, moderators are not the only party involved. SO has 200x more users who may participate in the closing review queue (~52k) than CV (261). – Jan Kukacka Sep 20 at 13:47

You must log in to answer this question.

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