Last validated CoC


#1

Can someone from CG (@_CG_Thibaud or @_CG_SaiksyApo I presume), after a natural investigation, can treat the following COC and remove them and have, at least, a serious discussion with their(s?) validator(s?).





Btw, as @eulerscheZahl said “Thank you very much for this demonstration of how vulnerable the approval system is.”


#2

One can add https://www.codingame.com/contribute/view/477643f7d3d676e2c04a7dd918b67f4bab20, https://www.codingame.com/contribute/view/4777cbe991fbd0f60680e83400d5c5ab7fb7, https://www.codingame.com/contribute/view/47791682739b26d7325f5dac4a537c8018e1, https://www.codingame.com/contribute/view/47666b39d453fcf8921300f4cf14c0cb54dd, https://www.codingame.com/contribute/view/4771201a367a4d524f315306c168d854b02c
to the previous list.

I really think, it’s time to seriously modify CoC validation system.


#3

these clashs are :

  • not fun
  • not interesting
  • with bad tests
  • wth bad statements

but who cares, no time to check, validated as soon as they appeared by friend/fake accounts …
if i could, i had removed them as quickly ( and banned all these account ) … ( i can … but i won’t because they’ll be stupid enough to resubmit )

it is rottening clash of code as we said it was going to happen …
cf Ian Malcolm: I hate being right all the time.


#4

These clashes are trash !


#5

+1
I really like CoC, actually it was the first reason I created my Codingame account.
And it’s really nice to have new puzzles from time to time

But these are just bad clashes. Like the one in Reverse mode, with always 0 as output in samples, that expects 1 in validation cases. Seriously?

There are plenty of CoC topics that are just perfect, from easy to really tough - nobody will miss these…


#6

I’ll look into the contributions and accounts today. Thank you for reporting it.


#7

I’ve removed moderator rights to jafar_03, Firdavsihas, Nazarov_Shohrukh, _ismoil, Ibrohim, Dark_hack, A.Safarkhon. I’ll contact them on Thursday to send a last warning before ban of their account.

On the 10 contributions reported:

  • 4 have been removed by the moderation bot already
  • 1 should be removed soon (once it gets enough notes)
  • 2 have OK ratings. They could benefit from a few updates (validators mainly).
  • 1 has too few ratings to judge even if it seems a bit easy (80% mean score; 124 players)
  • 2 are WIP

#8

The point here is not to let a bot do the job but MANUALLY remove those contributions since they, most probably, have been validated by fake accounts and without any real review process ! CG must really act against that kind of XP farming and have a real brain storming on the most global problem of contribution validation.


#9

XP “farming” (we’re talking about 10x3x15 = 450 XP spread over 8 approvers here) won’t be an issue anymore since these 15 XP will be removed.

To prevent group of friends to bypass the moderation system like they did, we could:

  • change the conditions to become a CoC moderator and raise the threshold of 50 CoC played to 100 or 150. Or add an minimum level too. Or add a minimum number of CoC victories.
  • add a delay of 1 day during which no approval/refusal is possible

#10

May I suggest a mix of those. Something like:
lvl 20, 100 CoC played and at least 50 CoC won facing real users (not a duplicate account nor a bot)

Concerning the delay, it won’t solve anything IMHO


#11

Not for the one who’s submitting the bogus CoCs.


#12

WOW ,this is shock for me !!!
I’ve never seen like this cool problem !!

https://www.codingame.com/contribute/view/48152bfac1a030b4f62489808331002056eb


#13

The author has been banned and the clash deleted.


#14

The validators are banned too ?


#15

These moderators are not moderators anymore but I haven’t deleted their account. I’m giving them a chance (I’m always trying to give people a chance).


#16

have or haven’t ?

Btw, the contrib is still present in the accepted contributions page. Are you sure you have removed the contrib ?


#17

wups, edited. Yes, the clash is not available anymore in the pool of exercises but the contrib is still visible.


#18

not a bad clash … maybe i should have validated it …
but the approve comment “I approved this contribution because it’s made by akai Sadi” … :disappointed:


#19

There is still one problem with it (and I don’t have permission to modify already approved clashes and correct it myself):
The validators are always much longer than the visible tests. The default code is generated by read s:string(256), which isn’t sufficient for validators having a length > 900.


#20

I know that i don’t touch CoC validation so maybe i’m wrong.

But how the f*ck is this not a bad clash ? Validators are not even close to visible tests. There’s no chance anyone found the solution during a CoC. Just, no … no chance. How am i suppose to guess this code. Even with validators there’s no chance i’ll find it.