Strict anti-cheat system when ?
→ Обратите внимание
→ Лидеры (рейтинг)
| № | Пользователь | Рейтинг |
|---|---|---|
| 1 | jiangly | 3898 |
| 2 | tourist | 3840 |
| 3 | orzdevinwang | 3706 |
| 4 | ksun48 | 3691 |
| 5 | jqdai0815 | 3682 |
| 6 | ecnerwala | 3525 |
| 7 | gamegame | 3477 |
| 8 | Benq | 3468 |
| 9 | Ormlis | 3381 |
| 10 | maroonrk | 3379 |
| Страны | Города | Организации | Всё → |
→ Лидеры (вклад)
| № | Пользователь | Вклад |
|---|---|---|
| 1 | cry | 169 |
| 2 | -is-this-fft- | 165 |
| 3 | Dominater069 | 160 |
| 4 | atcoder_official | 159 |
| 4 | Um_nik | 159 |
| 6 | djm03178 | 156 |
| 7 | adamant | 153 |
| 8 | luogu_official | 149 |
| 8 | awoo | 149 |
| 10 | TheScrasse | 146 |
→ Найти пользователя
→ Прямой эфир
Codeforces (c) Copyright 2010-2025 Михаил Мирзаянов
Соревнования по программированию 2.0
Время на сервере: 11.02.2025 04:04:38 (k3).
Десктопная версия, переключиться на мобильную.
При поддержке
Auto comment: topic has been updated by Axial-Tilted (previous revision, new revision, compare).
what for?
!damn codeforces team should consider it
Yess...Most would even have same IP Addresses
There is no need for it, wait for the contest to get over and it should be fine.
Quoted from this:
You seem to be unaware of how common cheating is even at the high level without a proper cheating system you can go uncaught for months even years , why not just make an app that u need to launch during a contest which monitors every open tab and app on the pc
But who would even play the game if they did that? Someone suggested phone number verification, and honestly that doesn't sound too bad.
u need something better , imagine an LGM using chatgpt to help it do minor stuff , thats about 5-10% improvement (could be even more lgm blog) and they wont get caught this way , also as i stated down below do anything dont be a sitting duck its very clear that cf hasnt been trying new techniques to counter cheaters
I believe they'll try to keep improving the system to better catch cheaters, and after all it's not that much damage done, neither does the cheater gain anything special, just a number on a screen that they know very well they don't deserve.
About your suggestion though, it's sort of unideal, since not all users sit through the whole contest straight without leaving it for a second, also it's still possible to cheat with your phone, so it wouldn't really stop it, it'll just make it harder.
Anyways, try to not be worried about cheaters, the more you think about it the more you feel worse
just like in gaming the cheating is a cat and mouse race , there is no point where u sit for over 8 years without improving your anti cheat system there is no way to stop cheating for ever u have to keep improving your anti cheat because cheating doesnt stop to grow , EAC is starting to die because its not using new techniques in countering cheaters a game like valorant decided that u need an app running in the background and its one of the most played games ever whats challenging about just installing an app and leaving it in the background ?
anyways thats not my point i dont mind it if they dont make an anti cheat app or something but atleast u have to try , for the past few years cheating in cf is evoloving much faster than their plagarisim checker , which if u dont want to try new techniques then a new website will appear and take the sport (which is a good thing since it increases competition and forces cf to try harder) just like vanguard and other anti cheats killed EAC , i wont be surprised if a new website kills cf remember EAC was once the most used anti cheating system and the most loved but if u dont keep improving its game for you
I don't think that's a good idea. A cheater can still use another device to ask ChatGPT then write the code himself.
i cant see them now in the standings, maybe they are banned
Cheating source increase always,cheater get the solutions easily!! Why Codeforces don't take any step seriously?
A bunch of grey bltches. How annoying.
codeforces should be paid so that it can avoid cheaters... or they should give 10yrs of prison if found any doing cheating during contest
Such system is unnecessary
bro wtf is on ur pp
Well I wasn't kidding. During the contest, my team and I have manually scoured submissions and skipped 14 pages of submissions. If a participant have one or more skipped in-contest submissions, then they are flagged for GPTing (or in some cases, other forms of cheating) and will be punished once reviewed a second time.
By the way, automatic system hasn't ran yet. The skipped subs so far are all manual.
As a problemsetter we had a voice chat during contest for skipping cheaters' submissions. Probably funniest experience in CF problemsetting career so far.
As a tester I can confirm they have a livestream exactly.
As a tester, how tf did I miss this
as a tester, are you sure you are indeed a tester?
Wait, I thought he was talking about round 1000 since he also problemsetted that one and i tested that one, my bad
credits to you an all of the cf team for what you have done today , the obvious cheaters are a matter of time the real issue are cheaters who use AI for small advantage which are so hard to catch using this technique and i think something new has to happen
Common cry W
As a setter who was sleeping at the time I regret to have missed this goldmine of an experience
So currently they are getting only solutions to individual questions skipped moving them (slightly) down the rankings but eventually (hopefully) you will terminate their accounts?
sooo, (I might be wrong but) i think 'skipped' happens to their entire contest (all the submissions) and not just 1 question even though they only had plag in 1.
which is fair tbh cause cheating should be punished as an whole and not just on 1 question
*I think this way cause i have never seen an account with just 1 submission skipped in their contest and my rank during contest and now is approx. 4k up
Hey, @cry @MikeMirzayano, can you tell us how many days it usually takes for the automatic plagiarism checker to complete its process? Also, in what cases does a solution get skipped due to plagiarism detection? Has the plagiarism check for Codeforces Round 1002 (Div. 2), which took place a week ago, been completed?
How did you even flag my code as GPT? It's ugly as hell bruh T.T
I think GPT can only provide problem-solving ideas, and this cheating method is difficult to detect through a certain detection system. I think the punishment for those who want to achieve higher scores through GPT should be a real-world competition There's really not much we can do.
At present, creating an account is easy, and it can be found that most of the users of GPT are new accounts. Since they want to score quickly, I suggest that the first 6 competitions of the new account be separated from the rankings of other participants
I dont want to sound as racist, but why most of cheaters are Indians?
bro common sense just compare India 's population with other countries there's huge difference obviously the probability of cheater being Indian also increases
What about China, which has roughly the same population size, but not nearly as many cheaters(at least among low rated)
codeforces is not the MOST popular or Most important OJ for them. But in India, Codeforces are most widely used closely followed by codechef ig. Additionally, most of the chinese who does Codeforces are Olympiad participants unlike us, Indians, who started codeforces to get a job in software field (majority).
I am not trying to justify ourselves. I don't cheat. But I do agree that we are obviously on the wrong side, but what to do, this is the reality.
exactly
China: "You, have GeeksforGeeks and CodeChef. We, have OI-Wiki and Luogu. We are not the same."
In the "Top Rated" sidebox of the front page, click on the "Countries" hyperlink and it will redirect you to a page that shows the amount of active users on Codeforces by country. India has by far the most users, $$$22207$$$, while the second place with most users is Bangladesh, with $$$5351$$$, which is only a fourth of India's amount; and China is third place, with $$$4757$$$.
even if you compare it by percentages, the rate is still higher
The same situation applies to Atcoder beginners or CodeChef contests. Also, no rating rollback has occurred for approximately 11 contests now.
Can you please clear my above doubt
WHY NO ROLL BACK?
What is the cheating here? İ can't see...
what for?
Perhaps adding a certain detection system based on (expected strength vs contest performance) could be useful. For example, if a person had stable performance of 1000 before and suddenly gets 2200, it looks pretty unusual. This would likely need to have a certain confidence level, based on # of user contests and dates of latest contests the user took part in.
This doesn't mean a person can't get lucky/unlucky or drastically improve in a short period of time. It is especially true if # of contests is small and/or there is a big time gap between rounds. But once again it's not to say everyone detected by such a system would be a teacher. Rather, these performances look suspicious, so might as well undergo additional checking.
Hello, could someone please tell me when Codeforces typically runs the plagiarism check? For example, how many days after a contest does it occur? Also, for round 1002 div 2 , have the plagiarism checks (and the process of skipping certain solutions) been completed, or are they still pending?
I think GPT can only provide problem-solving ideas, and this cheating method is difficult to detect through a certain detection system.
The punishment for those who want to achieve higher scores through GPT may be a real-world competition There's really not much we can do.
Damn AI for elevating the ignorant to a position they do not deserve.
Random idea : We can also fight fire with fire
Maybe CF can also use AI to detect who uses AI. The training dataset might be the manually skipped submission (as well as the traits of the user who cheats perhaps)
This might lead to false positive, indeed. But if the false positive happens to those who are being honest, then maybe it can be unskipped after they put their convincing justification. And this can also be the AI training dataset to reduce the false positive in the future lol
I dunno, just spit some random ideas lolol