Why is there a rep system in Codidact?

I largely agree with the answer of Olin Lathrop, but I'd like to put things in a slightly different perspective.

We need a rep system, even if it is not ideal, especially on technical/scientific sites. Why? Because Internet is what it is and no one is forced to reveal its true identity.

Reputation in real world matters, otherwise vocal opinions could swamp sound scientific reasoning outside academic or expert circles, where everyone can evaluate each other statement with confidence and rebut them with sound arguments if necessary.

You say "Your arguments should be your support in a constructive debate, not reputation".

This is overly idealistic and works in real life only between "peers" and with topics whose discussion doesn't need much basic knowledge. That's why, for example, some cunning and intellectually dishonest politicians get away with utter crappy reasoning: people aren't equipped with enough knowledge and competences to understand they are being scammed.

These sites are not only a place for high-level quality discussions about specific topics. They are a place for learning, even for newbies.

Imagine the following scenario.

A newbie (let's call him Lars) asks for an explanation of a difficult topic. Lars is an average 17yo high-school student who is not an English native speaker. Lars could also be a 35yo worker who hadn't got the chance to get higher education in his life, but he's a very passionate amateur. It doesn't matter for this scenario.

Two users, let's call them Bob and Jane, answer with two completely incompatible explanations, both credible at face value.

In real life Jane is an academic with years of experience, Bob is a smart troll who likes to have fun messing with people. Jane's profile faithfully summarizes her resume, and she posts under her real name. Bob's profile is completely fake, not even his name is real, and he boasts about difficult to verify achievements.

How would a newbie know who to trust without the competences to perform extensive cross-validations of information (even assuming he is willing to do so)?

Maybe Lars will trust Jane because she is an "uni prof", but how about the very convincing arguments made by Bob who is a "knowledgeable professional in the field" and happen to be a very clever storyteller.

Without a way for the community to say "Hey Lars, trust Jane, not Bob!" Lars could get very bad information.

And no, just voting on the answers is definitely not enough. Why?

1. Since statistically most people on a site are newbies, votes on wrong but exceptionally well written answers may swamp votes on right answers (plenty of examples on SE network).

2. There is a snowball effect: once a wrong answer gains some upvotes, it's difficult for later answers to recover, even if they are right. The Fastest Gun In the West (FGITW) wins, even if they are the evil ones! (Plenty of examples on SE network).

So rep points are, if implemented "correctly", a statistically meaningful way for the community to say "We trust this guy!".

Moreover, it's a way to say "Thanks you guy for your efforts!". In fact the above scenario is extremely depressing for Jane: she spent half an hour researching sources and trying to dumb down the topic for Lars, then she sees Bob's answer skyrocketing above hers.

I had that feeling more than once on EE StackExchange.

Ok, we are adults an we've got thick skin. You can endure that treatment once in a while. But what happens when Bob-like users almost always win? Well, Jane-like users are completely pissed-off and leave the site. After all, why putting on a community site quality content that no one care about, when Jane could put the same content on her blog and maybe get also some revenues by donations and patrons?!?

No one works gratis to be slapped regularly on the face! Well, unless you are the "Someone is wrong on the Internet"-type.

posted over 3 years ago

CC BY-SA 4.0

3y ago

Lorenzo Donati‭

14 3 63 20

We didn't set out to have a reputation stat, but because we started by adapting code that had it, we started out that way. When we brought up removing it, a few people objected strongly. (See, for example, this discussion.)

We've "nerfed" reputation; unlike on SO, reputation doesn't affect your abilities. It's just a number. Abilities are instead governed by specific, related activity; for example, you earn the privilege of editing directly by having enough of your suggested edits accepted.

We've also made the "usercard", the stats that accompany your name and avatar on posts and in the users list, configurable. A community that feels strongly that rep is important can show it; a community that wants to downplay it can remove rep from that block of stats. I now realize that might be a good thing to try here on Meta; I'll discuss that with the team.

We don't yet have a way to show you which of your posts have gotten recent voting. We've kept the rep number in the header (when you're logged in) so that, as a poor substitute, you might at least notice when that number changes. We do intend to do better there; it's on our list, but we're a small team and we haven't done that yet.

Some say that the reputation number is a measure of expertise. Maybe an average rep per post would be, but the total alone doesn't tell you whether the person has a few good posts, a bunch of mediocre ones, or a single hit. On The Workplace over on SE, we often saw a snarky answer on a hot network question skyrocket for the entertainment value. That someone has a couple thousand rep from that single answer doesn't actually indicate any expertise -- but, if people are evaluating posts based on author rep, that person's next answer would look stronger than it should. This is why, in the "usercard", we show the number of posts (by broad type). One of my early designs of that blob of stats showed more of a breakdown, something like "41 answers (39 positive)", but it was cluttered and we decided we needed to look for another approach.

posted over 3 years ago

CC BY-SA 4.0

3y ago

Monica Cellio‭ staff

64 372 2259 828

Part of this question is about making Codidact more Like Slack and Discord, which are not even question and answer sites. The two of those have their own share of problems, and the model that was used for them simply does not fit the requirements of a Q&A site. Social media and chat apps do not encourage quality or contribution toward the site. It encourages sitting at the same screen all day, waiting for some form of feedback from the site. There are many, many things which distinguish the two from each other.

Reputation is a system which(partly) goes against that kind of problem. You will be discouraged for spamming and posting bad content. You will be downvoted for posting poorly supported answers. Although it isn't perfect, it is at the very least, designed for a Q&A site.

posted over 3 years ago

CC BY-SA 4.0

Razetime‭

6 1 31 3

Actually Codidact should support rep better than it does now. Most of the dislike of rep seems to come from a misunderstanding of its purpose. It's not to show who is "better" somehow, but a measure of how much someone has contributed useful content to the site as judged by all the other users.

Your arguments should be your support in a constructive debate, not reputation

Right. Rep has nothing to do with validity of arguments. However, it might give a sense of trustworthiness when other direct metrics aren't available. In that sense it works like reputation in the real world.

Let's say you like reading horror stories, and there are a bunch of book in front of you to pick one from next. You've never heard of any of the authors, then you notice one by Steven King. You've read and liked his works before. You pick that one because his reputation suggests that you are more likely to enjoy that book than a randomly chosen one.

Trust Levels seem to be a better way (established by upvoted answers and the like) but showing a title rather than a flashy number.

But "established by upvoted answers and the like" is this reputation you don't like. Even if you don't want to show it, the system still needs to calculate and track it in this scheme. The only difference is that you want to show one result of that rep score instead of the score itself, but haven't explained how that is supposed to help.

Making it a repo game would lead to lower quality answers and questions as the primary aim would be points, not for spreading knowledge.

Wrong. Gamification means people will want to do those things that they get rewards for. If you've chosen the incentives correctly, then they'll be doing exactly what you want them to do, which is mostly providing high quality answers. Remember, the quality of these answers will be judged by everyone else, so they have to be good to get those points you don't like. Giving out free internet points with some recognition seems like a pretty easy way to incentivize people to contribute good content to a site.

People who want to answer questions (and are knowledgable) really need no 'fake internet points' as an incentive

This is hopelessly idealistic and naive at best. People do things for a reason. There is no such thing as pure altruism. Some people may feel internally rewarded by helping others, but there is always some reward or reason. Few people are going to spend the significant free time you need from them to provide good content with nothing in return. Giving them a little recognition for all the volunteer help is doesn't cost you anything.

What you are basically advocating is that volunteers should come here to contribute their time and expertise for free, and not even be thanked or recognized for doing so.

having a trust system would work pretty well giving them extra privileges

Privileges are different from public recognition. Privileges might be an incentive for some, but isn't going to be as powerful as public recognition for many.

I know from personal experience on EE.SE (I was #1 with about 280 k rep when I left there in 2018), privileges weren't much of a driver. I found some privileges useful, but didn't use others much at all. When I got a new privilege it was more of a surprise, and wasn't what I was aiming for. My thought was more "that's cool, maybe that will be useful some day".

Rep, on the other hand, made things competitive. We couldn't have Andy or Spehro with higher rep than me. I mean, geesh, that would tear up the fabric of space-time and end the universe as we know it. Can't let that happen.

while not signifying that they are all-knowing.

Right, and rep doesn't do that. It's a measure of appreciated and useful contributions.

posted over 3 years ago

CC BY-SA 4.0

Olin Lathrop‭

46 122 342 122

I agree with the notion that reputation numbers cause more harm than good, at least in general and at least in the long run.

I was not convinced by any argument I have read so far to why its "sorely needed" here in Codidact.

People ask questions and give answers about fields of knowledge that they have want to acquire knowledge in / be grasped as understanding in, respectively; so if the main motivator for such operations was points, then Wikipedia articles and community websites (say, MediaWiki support desk, Drupal forums, etc.), weren't so rich as they are with data given by volunteers.

What is repo/reputation?

Reputation is just number. Any user can know how active you are in Codidact by seeing your repo.

Why reputation is needed?

As @Monica Cellio answered is a next-generation academic platform for physicists and astronomers, in#answer-281378) repo doesn't affect your ability. While it does in SO. Actually I was very active user in SO and Unix SE. Sometimes I ask question right there not every time I got answers. Sometimes lot of user answers "simultaneously" we get in trouble that which answer I should try(It happens only when user is completely new to that technology or knowledge). So, obviously I will choose a professional's answer. By seeing the reputation we can know who is professional. That's why reputation is needed.

Look reputation isn't useless. Sometimes lot of user asks poor quality question. No one wants to review on a poor quality post. And, I personally like the reputation system of Codidact rather than SE (SE has the feature for answer not for question) cause, when we get downvote our reputation decreases but, if we remove that post than we get our reputations again while it doesn't have in SE. While user asks a poor quality question/answer we mark that as negative. There's another function I like there's differences of downvotes and upvotes. That's why we can know how helpful the post is by seeing the votes. You may say that your question was about reputation not upvotes and downvotes. But, your reputation decreases or increases depending on quality(upvotes and downvotes) of post that's why I added information of it.

Not only Codidact or SE has the reputation system most of Q&A sites have the features. I am adding some Q&A sites.

1. Kali Forums
2. Linux questions
3. PhysicsOverflow
4. Physicsforums
5. Manjaro Forums

There's lot more. I didn't add Quora and, Reddit cause, people uses them as Social Media (I was a Quora user also).

Reputation doesn't decide character or integrity.

The way to gain a good reputation is to endeavor to be what you desire to appear ~ Socrates

It takes 20 years to build a reputation and five minutes to ruin it. If you think about that you will do things differently ~Warren Buffett

Worry about your character, not your reputation. Your character is who you are. Your reputation is who people you are ~ John Woden

A good reputation is more valuable than money. ~ Pulilius Syrus

I remember I was bargaining with a Mathematician about a simple Math. Then, I told him that I am correct even, I showed him some articles in Quora. He replied,"I don't believe Quora cause, Quora is public place Q&A site. Anyone can answer there." Then, I didn't reply to him. I decided to look at upvote and downvote then, I noticed my answers have less upvote than, his. That's why upvote and downvote is very useful. But, you question was why there was reputation system. Actually, when we write bigger answers in Q&A site. And, if we can't get anything from them. Than, we don't feel good. But, when we see that we can gain reputation by those upvote and downvotes we feel happy. Even, when someone downvote (negative reputation) on our post we think that we should update our question. That's why reputation is very useful.

I completely changed my mind now. But, I want everyone to see what I said earlier.