-28

The more I look at Stack Exchange Meta or Stack Overflow Meta, the more I see the effects of the requester's reputation (and posters' ones) on them.

  • Do you have a lot of gold medals and reputation? Your suggestion has a chance to be good and well received!
  • Do you have few copper or silver medals? Await downvotes!

Am I right? Do I exaggerate? Well, we could argue...

I think that hiding the requester's and replying posters' reputation points and medals could be a good thing. Then we would be sure they could never be taken in account during reflection about feature requests.

Because, in my mind, posters previous deeds/achievements should not be taken into account in the case of such study, that should be absolutely neutral, and I think that the appearance of medals and reputation influence you.


When a public service or supermarket opens a suggestion box, it isn't asking those who are putting ideas inside if they are professors, medical doctor or own another ability.

And the service/supermarket is able to study the suggestions immediately, without this knowledge.

So why in Meta stacks do you need to know the reputation level of the posters to do the same task?

Do you believe you could be able to study features requests in the case the submitter and those participating to the debate had no reputation score and medals in front of them?

Like it's done in almost all suggestion boxes?

This factor (reputation) not shown would ensure it is would not be taken into account.

11
  • 21
    people who's been around a lot and have experience in how the sites work are much more likely to suggest good features than new people who simply lack that experience so there is correlation but not because of the cause you're suggesting. Feb 19 at 17:12
  • 3
    Robert Longson is correct: see meta.stackexchange.com/a/44189/377214, section Heavily consider not requesting a new feature if you're new. Feb 19 at 17:15
  • 1
    I'm not new, I have eight year of presence on SO, hundreds of questions among stacks but you see : only 1 silver medal and 10 copper for this meta stack, what mislead you because you take them into account and you're making the shortcut: few medals = newcomer, and you're wrong. If posters with a large number of medals and high reputation are better, their feature-request will be better by their abilities, already, isn't it?
    – Marc
    Feb 19 at 17:20
  • 3
    Here's a counterexample This user has two posts, only one is a feature request and it's massively upvoted. Feb 19 at 18:11
  • 10
    Correlation != causation as has been said. Can you provide evidence that supports your assertion that it's bias on the part of voters and not just not very good quality questions? Feb 19 at 18:27
  • 3
    I wonder if someone could dig up data on this. It may just be one of those "I bought a yellow car and now I see yellow cars everywhere" things... Or it may not. Data will help.
    – Tinkeringbell Mod
    Feb 19 at 18:56
  • 2
    You should be comparing the actual feature requests - how they are written, what propositions they make, and why they were made. Not just the reputation of the people who suggested them. Here is the thing - if Mad Scientist (top person from Oleg Valter's table) decides to propose something bad, e.g., every Thursday people should pay per character in order to post, then their reputation won't actually make that FR be well-received. The actual reason an FR would be well-received is because it's relevant and useful. Which hiding the rep of askers is neither.
    – VLAZ
    Feb 19 at 20:42
  • 2
    How do you propose that this be done? Hiding your reputation and medal count from the question page would be completely pointless, because I can just click on your profile and see it there. As far as I can see, the only way to make a poster's rep completely undiscoverable is to make the question itself anonymous, and allowing people to post anonymously on Meta seems like a great way to incentivise trolling and abuse.
    – F1Krazy
    Feb 19 at 23:59
  • 2
    @Marc - I determine an author has suggested a bad feature requests, based on the actual contribution, it has nothing to do with the contributor themself. You have made an assumption that the downvotes have something to do with you personally instead the bad idea you have suggested
    – Ramhound
    Feb 20 at 0:13
  • @Ramhound If reputation wasn't shown, I could not do this assumption. Because while it is shown, there's always a doubt.
    – Marc
    Feb 20 at 12:00
  • 1
    You are 100% correct in your observation. For example, this post would have got a very different reception had it not been posted by a very high reputation points user (for the exact same post). (Offsite, Jeff Atwood supports it.) Similar proposals by lower reputation points users, before and after, got a very different reception. Feb 20 at 12:09

2 Answers 2

7

Here is a list of top 10 users with posts on Meta Stack Exchange as of the time of this writing:

User Posts Years Total Score Whois
Mad Scientist 177 11 6518 Long-standing active community member
Pekka 104 12 3968 Long-standing active community member
Gilles 'SO- stop being evil' 130 11 2884 Long-standing active community member
Monica Cellio 71 10 2497 Former mod ousted by SE, one of the founders of Codidact
ChrisF 42 12 2404 Mod on multiple sites of the network
Shadow Wizard Hates Omicron 185 11 2383 Long-standing active community member
Lance Roberts 164 12 2144 Long-standing active community member
cletus 39 12 1870 Long-standing active community member, all FRs made back in 2009
Manishearth 103 9 1854 Long-standing active community member
Shog9 33 12 1832 Revered former CM

Notice something about the list? All of them are extremely active community members with an enormous number of posts that have been around for about a decade. Some of them are also former employees of Stack Exchange. They know ins and outs of the network better than anyone else and are thus able to propose changes in a way that is positively received by the community at large.

Now, let's take a look at the bottom 10 users:

User Link Posts Years Total Score
Evan Carroll 7 12 -217
JMP 3 6 -114
Dmitri 4 11 -95
Ben Madison 3 2 -92
peterh 18 7 -91
Monika Michael 1 9 -86
TechnicGoblin5R 2 2 -86
IndRaj95 2 8 -83
danny117 2 8 -80
Paulo Roberto Rosa 4 8 -80

One thing that immediately jumps out about them is a much lower number of posts and a much more diverse user age — which shows that to suggest negatively received changes, you do not have to be a new user.


As for the correlation between reputation and reception, I think you are interpreting the relation the wrong way: do not forget that every upvote gives the user +10 reputation points, meaning the better a feature request is received, the more reputation the user will end up having, not the other way around.

This explains the illusion of higher reputation users' requests being warmer received than those with lower reputation — they have the rep because they posted positively received requests (as well as made tons of other quality contributions to the network).


Finally, let's take your post here as an example: you made a feature request without supporting your claim with data (SEDE is available for everyone to leverage) and then proceeded to make conclusions based on unverified observations. This is the real reason why your request is getting downvoted, not your current reputation level.

For your benefit, please, read the following stellar guides on participating on Meta and making feature requests (the latter made by one of the top 10 users from the query above, the famous Shog9):

8
  • 4
    You should not be calling out users by name.... Feb 19 at 21:07
  • 2
    @DavidPostill hm? That's just data samples with aggregated scores, no one's calling out anyone here. Feb 19 at 21:31
  • My topic isn't "Are members with high reputation good or not?", it's: "We should hide the reputation level of posters in feature request questions to allow a better study of the request, because knowing it might impact your answer to it". Therefore, as an illustration, you all explained here why you are taking much care about that reputation level. This is why I think it's important to hide it, sometimes, because you are looking driven by it. And this is an obstacle to your neutrality during study. A neutrality expectable in order focus on the subject of a request only and nothing else.
    – Marc
    Feb 20 at 4:25
  • 4
    @Marc "because you are looking driven by it" you've offered no actual evidence towards this conclusion. So, you want this implemented because maybe it looks like this to you. That's not a good justification for a feature request.
    – VLAZ
    Feb 20 at 7:27
  • @VLAZ No, it's only that making this information disappear would ensure it wouldn't be taken into account, that's all. It's evident.
    – Marc
    Feb 20 at 12:16
  • @Marc it is not even evident that this is taken into account, let alone taking it out ensuring anything. Seconding VLAZ here, a feature request, especially one that proposes to take something out, should be substantiated with data and impeccable arguments - of which we, unfortunately, heard none (hence the score of the post) Feb 20 at 13:23
  • Note that it does not deny the existence of a phenomenon called "familiarity", bit has nothing to do with rep, but rather with a username. F.e., I know Glorfindel, and, if they make a controversial FR, I will be more willing to give them a benefit of the doubt than to that of a user I know nothing about and who just popped into existence yesterday. But regardless of how I feel, the next step will be reading the content of the post, and if the former suddenly decides to make bogus args, and the latter provides a coherent request, my votes will distribute accordingly. If you want a... Feb 20 at 13:52
  • ...completely unbiased interaction, as F1Krazy mentioned, you need the posts to be anonymized as in blind review, but that is in direct contradiction with the terms of CC-BY-SA license all our contributions are licensed under (not to mention that blind review is temporary and is applied before a scientific work is accepted in a journal). Feb 20 at 13:56
4

There is probably a small effect here, but I doubt it is as large as you suspect. It is likely more relevant on the main sites, not on meta. I think hiding reputation in general is interesting, MathOverflow for example doesn't display reputation on user cards at all. But I see no reason to special case Meta here, if this is changed or adjusted it should be done for all sites.

The biggest advantage high reputation users like me have is the experience on how Meta works, and knowing what the community thinks about certain topics. Along with a lot of practice on how to write good meta posts. I can usually estimate pretty well whether a particular feature request will be received well on Meta or not, simply from experience. But I can't make an unpopular topic popular, even with all my Meta reputation. See for example my lowest scored meta post, I fully expected this to be heavily downvoted before I posted this. This is a topic where I disagree with the Meta consensus, which is reflected in the score.

In the end scores for feature requests on Meta don't really matter. The only thing that matters for them is getting the attention of someone from SE and convincing them to put it to discussion internally or ideally even champion the issue within SE. Having a high score helps a bit for getting noticed, but it doesn't really help on its own to get SE to actually do something.

The most reliable way I found to get feature requests implemented is to propose things SE is planning anyway ;-). The roadmap isn't driven by Meta, SE decides that own their own. The feature requests that get implemented are often small things that don't take that much work, or are things SE decided that matter to them. Sometimes a meta post is the trigger to get them to notice an issue or give them an idea, but my understanding is that very often SE puts an issue on their roadmap independently from meta and only then searches meta for any previous posts on this. And for this score doesn't really matter, but arguments and details do.

6
  • If a public service or supermarket opens a suggestion box, they don't need those who put ideas inside to specify that they are professors, medical doctor or own another ability, to be able to study their suggestions. So why in SO would you need reputation, medals to do the same task? You think that the reputation of the poster doesn't really matter to set your judgment, but all the argumentation I read until now is that "It's essential". Could you study features requests in the case the submitter and those participating to the debate had no reputation score and medals in front of them?
    – Marc
    Feb 20 at 8:35
  • @Marc "but all the argumentation I read until now is that "It's essential"." where exactly did you read that? As far as I see, everybody who commented or posted an answer here claimed the opposite. That it's not reputation that makes a good feature request, a good feature request leads to more rep. And that more participation on Meta leads to better feature requests because of knowing how and which feature requests to post.
    – VLAZ
    Feb 20 at 12:23
  • @VLAZ You are arguing that high reputation level ensure more chances to create a good suggestion. I don't deny it. But what I'm writing is that not knowing the reputation of the posters ensure that their reputation won't a factor taken into account during the study time of that feature request.
    – Marc
    Feb 20 at 12:31
  • @Marc "You are arguing that high reputation level ensure more chances to create a good suggestion" no - the exact opposite in fact. There is no lottery to it. It's a skill.
    – VLAZ
    Feb 20 at 12:34
  • @VLAZ Read the comments again. A skill, yes, but that you are relying to the reputation level of the poster that gives his experience in it. And this is why seeing this reputation affects you during the study time. It's a factor that could disappear with interest during that time. You understand that if in front of each poster was written his political obedience or the religion he his following, if any, it could also affect neutrality? I think that the appearance of reputation scores and medals tend to affect judgment.
    – Marc
    Feb 20 at 12:39
  • @Marc Which comments? You've not actually said where you got the idea from. I've read them and summarised what I got: that people who post good content tend to get good rep. You're claiming that good rep leads to good content. Point out where is your inference coming from exactly.
    – VLAZ
    Feb 20 at 12:42

You must log in to answer this question.

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