Disclaimer: the first post has been deleted a few minutes earlier. Probably I should not post this during contest.
Facts: this is a large chat group in China(you can see there are ~2k members) with lots of IGMs and LGMs. Most of them are middle/high school students. In the earlier post, the author replied:
I am so impressed by this comment. Furthermore, 1F is clearly a classical problem. But since there must be someone discussing this right now, I believe that they can give the better explaination. One year ago, I was informed that posting "contest is not very good" after testing is forbidden, well... I am speechless this time. Let's enjoy the original post.
Translation:
aaa_Pigeon: It feels like there is quite a large skill gap among contestants on Problem E.
[zh1221_qwq deleted a message.]
aaa_Pigeon: The standard solution is actually pretty short.
Translation:
aaa_Pigeon: I’ve already stopped caring about cheating.
aaa_Pigeon: As long as it doesn’t affect me making money, it’s fine.
Translation:
aaa_Pigeon: Isn’t this problem very easy?
aaa_Pigeon: I feel like at least ten people will solve every problem.
Translation:
aaa_Pigeon: How is it easy? Damn it.
Oken: [Censored]
aaa_Pigeon: tiger and cdx didn’t even solve every problem in Division 2.
aaa_Pigeon: I think some contestants are simply too strong.
aaa_Pigeon: And there are also a bunch of AIs.
Translation:
aaa_Pigeon: Fortunately, the standings weren’t distorted.
Dang it.
Wow he just says it right there, crazy how much hate I was getting for exposing something so trivial for anyone that has eyes.
It's out fault for the reused problem, but I don't think my massage brings any useful information during the contest.
Well, considering that the standings were affected by the time penalty (and that Benq's strategy of solving F before E1/E2 paid off), knowing which problems are easy and which are hard seems to be really useful.
Standings without the WA penalty:
Benq: 10432
Kevin114514: 10288
Benq had a 144-point advantage despite taking half an hour longer than Kevin114514 to solve all the problems.
As long as the messages do not provide any advantage — yes.
That's what I think. You can even see that im saying a problem both "easy" and "hard", this is complete useless.
Looking at the editorial's code for E it seems he has a funny definition of "short"
the std has 150+ lines，dovyou think it‘s short？Can you guess the solutioj by that？
trash contest
Please, learn how to speak English before you say these words, thank you.
The most arrogant, immature, stupid author forever.
I don't think it was useful information in any way or form, but as a rule of thumb and in lieu of fairness contest authors should have the decency of not talking about their contests, also since this served no purpose the author should simply not have posted this, the only reason i can imagine is that the author are children who felt cool for some deluded reason. and the author has shown time and time again that they have no regard for the rules of codeforces and should simply be not allowed to have the authority to write contests or be testers till they develop the maturity and the ability to A) own up to their errors and B) have the required professionalism.
also in what world was the standard solution of E short, TF.
now because of $$$AI$$$, one can't even be sure that authors are of a certain skill level anymore, so stuff like this might be the norm in the future
aaa_Pigeon2 is so notorious! ❌
Hard to believe several people banned for cheating were allowed to set a Codeforces Div. 1 round—and even reused an existing problem.
This is a homework problem for undergraduate students in THU a few years ago:
In fact, my free Gemini can find a source with some comments that is enough for me to make up the solution (but Gemini is somehow intellectually disabled; it could not work out the problem itself with the comments).
I did send this to the coordinator, and they told me they were considering replacing this problem. However, at the last minute, they said they would not change it.
I was confused and didn't expect so many honest participants. Did administrators manually ban many of the AI solves?
There are dozens of rounds with cheater problemsetters, this is not new
Damn it，how can he say that " I’ve already stopped caring about cheating.As long as it doesn’t affect me making money, it’s fine."
That reminds me of how innocent and kind-hearted I was back when I was a Chinese middle school student.
what can I say,anybody who can say this should not be allowed to prepare for contests
It is quite redeemable given they're likely still teens.
And [censored] lmao.
No, I don't think so. As a young person myself, don't take the responsibility of running a CF round if you can't take it seriously. These authors are tarnishing the trust we have in all contest authors. I hope this round doesn't become the reason CF says you have to be at least 18 to submit a round now or something dumb like that.
Yeah we need problems written by not only ACMers but OIers to enhance the diversity of this community, so there shouldn't be an age limit of the authors.
But some authors like aaa_pigeon2 are pushing this to a bad edge.
I mean it is very normal for a teen to be arrogant or something else. So I don't get furious at their chat at all.
Okay, so I am a friend of aaa_Pigeon2 plus a tester for this round.
I guess I should point out various funny things (as a friend lol):
You guys are getting baited lol. He mentioned to us that he was deliberately farming downvotes, and you guys actually thought he was legit trying to explain.
Regarding problem F, he literally told us long before the contest, "Dang, a lot of my problems are becoming research paper problems". He just didn't think everyone read the paper lol.
I quote: "But during the contest, someone pointed out that 1F had previously appeared in a research paper." It is himself lol
Translated image text:
aaa_Pigeon: Definitely replace H, mainly because there exist similar research papers
Guany: I don't think it is necessary, because we trust high-skilled players won't Google for papers
Yeah this is incredibly out of touch to literally know it's a paper problem and then proceed to assume people won't Google (which is literally allowed and encouraged on CF btw) (And also AI generate the editorial)
"Yeah goodbye 2023 H is OEISable but we trust that nobody will check the sequences in OEIS!!"
Guany is actually the coordinator, so I feel like I’m taking a bit too much of the blame for the duplicate 1F.
Maybe I should be more decisive.
It isn't even prohibited to Google it bro, what are you saying? They are expected to google it.
So, he didn't even want to explain, which means that, he admitted that those messages are "important"?
Also, you used the word "lol" a little bit too much lol.
Not really; obviously, he doesn't care about his contribution. There is no doubt that aaa_Pigeon does what he likes to do most, which is watering groups (sending messages). Why do you care whether the messages are "important"? He himself doesn't even care, and I find the whole topic quite meaningless.
But, his action may have violated some CF rules.
I don't see any information regarding the communication between the contest writers (and other people) during the contest mentioned in the rules. Moreover, the messages don't really affect much about the problem since the messages only expresses aaa_Pigeon's POV of the problem, so I think that it is completely legal to post messages about this.
Regardless of the value of the sentiment Pigeon had towards solving the problem, I feel like given hes a contest writer, anything he had to say that wasn't in problem should've been either publicly announced, announced solely to testers/coordinators, or not announced at all. I feel like if what he had to say truly didn't have any bearing on the problem, it could've gone into the initial announcement as a small remark however it didn't and instead went to a select group of people
do you really believe his words? i dont think so.
I think we should ban aaa_Pigeon2 from setting problems/contests.
If the author is freely chatting about the contest in public chat while its ongoing, i think its only reasonable to suspect he gives friends/teammates/colleagues hints and solutions before/during the contest too.
aaa_Pigeon2 does not deserve being a problem/contest setter
a monkey on a typewriter can write something original but a pigeon can't?
Meanwhile, we kept trying for a month to host a round alongside this contest. Until we couldn't get it done with Codeforces in the required deadline and ended up just posting the contest in Gym section.
But a cheater author was able to host a round.
In this case I believe CF generally doesn't like hosting official mirrors except for large contests like ICPC and wants to keep rated contests in the CF format (~6 problems that rapidly increase in difficulty)
We had that :D
Sorry did I misunderstand? Your contest was more ICPC style 14 problems arranged in random order.
Exatcly, so we choose a subset suitable to make a Div2. problems were more Codeforces style.
Ah I see. Yeah that's unfortunate. It seems like generally queue can be quite long (I remember OtterZ saying they had to wait like half a year), surprisingly we got to prereview stage very quickly and it seems like so did 1112.
hmmm reading your comment I can say that you are going to author a round. Interesting :)
You can orz KAN and Um_nik for helping us get such quick pre-review. We might get to coordinating stage soonTM
Great news! Now I'm looking forward to it.
So did you propose a div.2 or (div.1 and div.2)? Should I lose rating and become CM to participate rately? lol
It will be div 0. Problem A will be 3000, problem B will be 4000+, problem C will be to pretrain an LLM on CF (you will get 8 seconds TL) that obtains better loss than claude fable 5, and D & E will be NP-hard.
Here I'll go ahead and leak problem D because I'm orz like pigeon. It is output only. Find the two prime numbers that multiply together to give
(We will try to make it div1)
Did someone just leak the next contest?
Guys this leak be true I was one of the coordinators
Am i right?
Dang I was not expecting someone to get it so fast. did you run GFNS or did you just know
You call this NP-hard?? =))
" or did you just know "
Come on, who doesn't know the legendary meme number 2012143212464846358953306572921606106425603248617529551804241356732859022591581928055658 87258342852981765672067333154878543232100367780170768382852936529 and it's prime multiples. I mean, who doesn't know those?
You don't need any fancy algorithms to find such small factors though. Hard factoring problems need to have both primes be large.
It's normal , when you're factorizing at "theortical" complexity of $$$O(\sqrt{n})$$$ , that's on the assumption you don't know nature of prime factorization , but given you know it's already on form $$$pq$$$ where $$$p,q$$$ are primes , then finding $$$p$$$ is enough , which in this case is just $$$67$$$ so nothing fancy on this.
try harder ;)
I meant to ask if they just run trial division / GNFS or if they just knew that one of the primes would be 67
Ironic
Lol. We know the context.
To whoever is watching: I gave everything I had. My value is self-authored , no spectator holds the pen.
aaa_Pigeon2 has even leaked something about the problems even before the contest. See at this blog.
he just ruins every single contest and gives everyone a hard time >:(
aaa_Pigeon2 could just apologize and this would end. But he didn't...