CoC - new contribution system


#1

Hey dear Codingamers,

Here is my suggestion for a new Clash of Code contribiution system.
First my reasons for changing it:

So now my suggestions:

  • level 10 as minimum (after level 10 you have basic coding knowledge)
  • if the puzzle is approved - a last check by somebody level 20+

I think there must be a change but i’m open to suggestions in comments.
(And sorry for my bad english :wink:)


#2

Thanks @BlackBird1for bringing this topic up. We’re aware of the current issues and looking to solve them.
I’ll gladly read other suggestions too.


#3

Thanks for your reponse and i’m glad to hear your are working on it!


#4

Your suggestions don’t actually tackle problems 1 and 3 as listed


#5

I did not say i know a answer for every problem i just wanted to open this forum for suggestions for a better system. So if you know something improving the system post it here.


#6

We’ll analyze the ratings of CoC puzzles to possibly find a pattern there. Then, we’ll take actions to:

  • remove bad clashes
  • either add a moderation rule according to the pattern that was found
  • or implement a bot that will automatically remove a clash that has too low ratings.

#7
  • Clearly, level20+ is too low in the BlackBird1 proposition.
  • A system to remove bad clashes would be nice.
  • It should be nice to allow anyone to see both already accepted and refused contribution (if possible with a tag system) in order to prevent too many duplicates and resubmission.
  • Concerning the validation itself. The minimum requirement IMHO for an acceptation should be to have coded a solution in a language which is not the contributor one (which implies in particular to hide the contributor solution or at least not to allows someone who have seen the solution to be authorized to vote on a contribution). Such a restriction is only for acceptation: some contribution have just no solution w.r.t. to the statements (cf https://www.codingame.com/contribute/view/2147bc78b75e38ba723144b4dd4358ec7578, https://www.codingame.com/contribute/view/21495dbd907e35353c4686056ece71146887).

#8

Yes, contributions should probably be visible to all CodinGamers.

We’re also thinking about removing the XP for moderating puzzles and CoC.

I also agree on having solved the puzzle to validate it.

We’re aware of the current issues in the moderation system. It’s clearly not perfect and we’ll work on improving it.

However, I’d like you and other veteran players to keep being polite and nice in your discussions.

If any complaint should be voiced, it should be to us, CodinGame, not to new players merely using an imperfect system. Some players are quite young and enthusiastic. I believe that they, most of the times, don’t realize what they’re doing.

@Boulet @Niako @Stilgart @JBM @dwarfie (and others) … I see the good work you’re doing in moderating and I thank you for it. Please be patient :heart:


#9

Concerning the form, that’s the second time you seem to accuse me to be unpolite and not nice in my discussions. I would like you to point me (in PM if you whish or in public I do not care) any case where I have been unpolite.

Concerning the real problem, we can try to have here a constructive discussion to solve this acceptation/refusal inflation.


#10

One problem is that the solution is visible to others, don’t forget it ! ^^ Moreover, we can see tests & validators… I can’t think to a solution here… even a secret validator can be given in private message…


#11

Not exactly since I propose to hide the solution. I agree that this is not the best possible solution (only the best I found writing a message). Maybe also hide validators too before having solved all tests or something like that. The goal is not to prevent active cheating (that’s CG role not our) but to prevent to fast clicks (pro or cons).


#12

You don’t ensure here that the solution from the submitter is not hardcoded :confused:


#14

Just your friendly reminder that not seeing the solution prevents from editing the submission.


#15

I remember a community puzzle (already approved) where I passed all IDE tests but failed one on submit.
After some frustration I decided to have a look at the validator only to find out that there was an extra whitespace at the end of a line, making my input parser crash. It worked fine in other languages that have a different way to read numbers from stdin (like cin in C++ instead of line.Split().Select(int.Parse) in C#).
In such cases not showing the validators is an issue in my opinion, as these can be faulty too.

And for a reverse-only clash the problem setter’s solution can help to figure out what’s going on.


#16

To your last point with the reverse puzzles: Especially in this mode I think solution should be hidden because without a solution you really have to guess what you should do. With code it’s always easy to know the question so it’s not objective anymore. If you have questions you can ask them in comments.


#17

Of course I try those (especially reverse) without looking at the solution before.
But there might be an easy reverse puzzle where I just don’t see the obvious solution.
Or it’s completely random, as the Kanban one mentioned above.


#18

I guess that reverse puzzles you can’t find a solution to after some time thinking could be refused. (after checking with others in the comments probably).


#19

One easy solution to stop very fast acceptance of CoC is to froze acceptance of problems for 24 hours. That way more people could see the proposal, who actually has an idea of standards. And honestly I don’t think any of these people would just randomly refuse a problem.

The other thing is sometimes even the person who proposed a problem not sure how well it will go, and very open to ideas, but it gets refused/accepted before any change could have been made.

For the acceptence of CoC, haveing solved 50 of them is like an afternoon time of work, definitely not enough as previously mentioned. Unfortunately level and rank not necessarily correspond here, and I’m not sure there are not enough lvl 30+ people who have a strong and reliable judgement of problems also have the time to do it.

Thank you in advance CG team, I would no want to be in your place to solve this problem, and make everyone (or at least most people) happy.