I understand that you are unhappy with my rounds. And I can try to understand what the precedent is.
You think the problems are math. Understand, I don’t like mathematics, but I can solve the problems that are offered to me. There may be other reasons, we have testers for this and they give us all the necessary advice.
About the last round. I apologize to everyone for what happened. We decided to change problem A at the very last moment so that everyone could pass it. Initially, we had a tested problem, but it was more difficult than task A. Problem G was my authorship and there was an error in the author’s solution. I made the weight restrictions greater than int and did not change the data type in one place. This influenced the answer. We asked problem H from other authors and, unfortunately, it turned out to be known. We removed two G problems because they were famous (one was removed because it was in the 1986 Olympiad), but we did not notice this.
I apologize to the entire community for this incident. You don't deserve this and we put a lot of effort into making this round happen. Now this is all history. Happy New Year everyone and I wish you all the best!
Auto comment: topic has been translated by 74TrAkToR (original revision, translated revision, compare)
.
Мои извинения, если мой комментарий вызвал недопонимание. Моя цель была поддержать автора и обсудить возможные улучшения. Всем Хорошего нового года
Answering to rev.1 of your second comment, you literally asked to make this round unrated in rev.4 of your first comment, so in advice I’d recommend you not to take downvotes personally. Your fierce urge to be liked by strangers leads you to confusion and puts you in a foolish position.
so what's the point of losing your contribution because of some comment?
what's the point of contribution at all, brother? these numbers mean nothing, just chill and take it easy, getting positive responses in cf blogs has nothing to do with cp
outplayed.
It's important to support each other, even if we don't always agree. Let's find common ground and build a positive society together
Lmfao bro had a seizure getting all these downvotes
They(problem setters, coordinators ,testers....) anticipated an incredible contest featuring the most challenging math questions, but the aftermath left them with a disappointing farewell for 2023.
OK, then maybe make the round unrated if the entire community is unhappy with it?
I am happy if its rated aswell
I'm happy only if it is rated
maybe 2023 ended well for us
for me too
deleted.
Yes, because finally I'm M, and I didn't even attempt the last 3 problems.
Finally I am CM! I failed to hack people...
I'm happy with rated
same, this round increased my rating by over 100
im happy boi
it's alright
thanks for everything you and others do (preparing problems, testing etc)
"You guys put in a lot of effort" shouldn't be the reason this round is rated.
Agree.
I’ve heard from previous problemsetters and testers, that 74TrAkToR is the kind of coordinator who goes his own way and doesn’t listen to the opinions of problemsetters and testers. If he insists on doing anything, no one can beat him.
I believe that Golovanov399 and Endagorion, as former LGMs, should be able to see the issue of H (if they have seen H). It is said that the setter of H himself also knows that the formula of H is already well-known in mathematics.
Then why, H is brought up?
When I tested this round, there wasn't problem H in the problemset. I also didn't know it was going to be the goodbye round. I wouldn't want H to be here, let alone at this position.
On the other hand, I can't say that the coordinator ignored any of my feedback.
I specifically said H was a known exercise in my feedback.
Agree.Problem H is totally different to those who have learnt in math lessons an those who know little about it.For one they can solve it in less than 5mins while for the other,they probably cannot solve it ever.
Unless they use oeis.After a litle time for thinking they can solve it though they don't understand the meaning of the problem.That's funny.
Agree.
пацаны не извиняются
ахахах, гений
он трактор
Now this is all history!
Now this is all history!
lol
How did you manage to find that some problem was in 1986 Olympiad and at the same time forgot to put pretest of H in oeis?
can you please explain what is oeis as i'm new to competitve programming and i don't know about it
https://oeis.org/
https://oeis.org/A286331
Its a wiki about known sequences
so please distinguish the effort and the end.please unr this trash round.
guys, downvote this instead of the round announcement. That way the right person's contribution gets decreased. There is no way an experienced coordinator misses 2 (3?) googlable problems, a wrong model solution and accepts tasks like D. he just doesnt care about the quality of his rounds
Exactly. It is a well-known trend (at least on the internet) that an apology always garners internet points and the original mistake ends up being penalized.
Penalizing the round authors for something that was the coordinator's job and then the coordinator (perhaps unintentionally) getting sympathy points for the mistake because of the supposed effort that went into the round is absurd.
I haven't yet seen a decent round coordinated by the coordinator in question, by the way.
Where is the apology to marzipan for -142 contributions and a successful top 1 since the end?
it's should be unrated, untested A, wrong G, copypasted H, not enough?
and awful D
Why? U didn't even participated in round?
add genshin to statements next time so ppl wont blame ur problems
Abstraction (a Chinese Internet talking style) finally comes to Codeforces?
That is, you know that H is known, and still use it in the round?
That's not what that sentence means.
See the comment of Endagorion, they do know.
I could be happier if it was unrated ...
It is okay knowing you tried your best , Happy New Year !
it's not 💀
Poorly written statements, googleable last problem, wrong author solution of the pre-last problem, how could this round stay rated?
You forgot googleable F
It's the least of the round's problems... It's like the round was made the day before it
If googleable H is a problem, so is F. Let's not gatekeep hateing googleable problems
The difference is that you need to google one sentence for H, googling F is hard.
Google en passant "partition graph into connected parts"
Holy Hell
New algorithm just dropped
Call the tourist
libobil We need Sashka rounds!
We'll see about that
We need "who doesn't love math" problems
Almost everyone is unhappy with it, and I think it must to be unrated.
Dude, you don't just post a blog saying what difficult you've met and all is over. You ARE actually ignoring what you've made to us, and do nothing but saying you're sorry? If it is, then we are way TOO forgiving.
???
So make this round unrated!
You made lots of bad contests even if you tried your best. Isn't that one reason you should STOP coordinating contests instead of continuing coordinating contests?
Did anyone submit a correct solution to G, only to have it marked as WA? Are you going to retest G? This seems like it alone would be a reason to make the round unrated, unless you claim that it didn't affect many people...
I think it's disingenuous to frame crititicism of your other rounds as "not liking math". There are a bunch of authors and coordinators who love putting math problems and they don't get nearly the same ammount of criticism as you do.
Was there no tester on problem G? Is having a second solution not required to hold a problem anymore? I think you should go through some training from other coordinators and learn how to prepare a round before doing any more, given the history of your rounds.
I believe that we will see the history again in the near future.
I'm sure that these tasks is not easy job, but please unr this round. Everyone will enjoy happy year.
Ok let me believe you are caring about your rounds and you tried hard (yeah actually too much hard) for them and other bul***it stuff you said.
CAN YOU EXPLAIN WHY ARE YOU DELETING COMMENTS THAT CRITICIZE YOU TOO??? link
Bro played the victim card. This is all history Lmaoo
Why is he still a coordinator? — Firstly NP problem in round 846, now a mistake in author's solution and plagiarism in one single round
Apology accepted!, now make it unrated? I guess?
unrate pls
What a convincing blog "this is all history".
world war 1, world war 2, covid19 virus & ... is all history too so lets forget them dear people!!!
I wish i could also say Happy New Year but it is not happy until this is rated
Maybe you made lots of efforts. Maybe you do not want such incident happen. But all above can not be the reasons that this round will still be rated.
Imagine how bad history would be if worst round with incorrect problem would be rated.
Crocodile tears!? Is it you?
I'm sorry about the -ve contribution
chill out
I could mash a better contest in a week without knowing solutions for my problems.
chill out
https://codeforces.me/contest/1916/submission/239626221 Weak system tests as well, this solution is hackable. Make this round unrated. Edit: Sorry guys, it's alright. I missed something.
Nah, it's not.
sorry my bad, missed something.
I just tried to hack a very similar solution and bruteforced all test cases with overflow and numbers from the divisors of 2023.
The last sentence sounds pretty dumb... What do u mean by you "put a lot of effort into it"? Does it seems like a good excuse to pardon your muff?
Will the round be rated or not??
I just hate the problem B, I came up with the idea using half an hour but still don't know why. And C-E for me is just fine, but others i just don't know why. Also, as a Chinese programer, participating this contest until this moment without sleep, expecting gaining some joy from it, but it came out to be an upset result, just feelng painful. Hope next contest could be better.
https://codeforces.me/contest/1916/submission/239633251 this solution is hackable maybe !
I love googleforces
orz
Why is it rated?
Everyone is unhappy with round(sry, everyone except few people that googled H), and u say this is all history. U guys litteraly took people's 2 hours. Make round unrated cuz its crap.
Everyone is unhappy with round(sry, everyone except few people that googled H), and u say this is all history. U guys litteraly took people's 2 hours. Make round unrated cuz its crap.
i see allot of people crying to make it unrated cause they lose rating
This test can hack solutions that received an AC verdict for problem A on system tests that uses int rather than long long. The system tests are weak.
1 4 1 293 879 1253 2023
In problem B , b shouldn't be <=1e9 , because if b can be 1e9 then x will not be <=1e9.
I have sent a clarification for this issue : "How b could be 10^9 with the fact that b is the largest divisor and not equal to 10^9 and x must be less than or equal 10^9?"
And the answer was : "Read the problem statement".
Now enjoy this submission : https://codeforces.me/contest/1916/submission/239725756.
And try to be clear next time.
I really don't understand what you're complaining about. The statement says that $$$1 \leq a<b \leq 10^9, x \leq 10^9$$$, all test cases have $$$1 \leq a<b \leq 10^9, x \leq 10^9$$$ and the tests aren't weak. The statement could've said $$$1 \leq a <b \leq 10^{1000}$$$ or just $$$1 \leq a < b$$$ and it would be equally accurate.
I am complaining about laziness in reply. The answer to the clarification could be "NO". You are right, I am just asking what is the point of putting a<b<=1e9 when you can just put 1<=a<b .
Yeah, you absolutely right, statement should be: ($$$1 \le a < b < 999999999$$$). It would be much more better! (no)
Where do you complain about problem during contest?
Open the Dashboard of the Contest --> Ask a question ---> Choose the problem you want a clarification for ---> Write --> Send
I had nothing to do with this round, but please don't use clarification messages to complain to authors about their problems during the contest. They have lots of other things going on and when you do this you're delaying the amount of time it takes for them to answer real clarification requests.
"Now this is all history."
Bro pulled a "I'm sorry, but fuck you everyone"
Ходи оглядывайся
why rated if some people couldnt hack and that's an obvious disadvantage?
Someone won't be a setter in 2024, as i can see
To follow up on many of the comments above, I wanted to summarize what I view as some preparation errors in the round and to give constructive feedback to coordinators in the hope that we can avoid similar errors in the future.
In my opinion, the original post (quote below) mischaracterizes why rounds you've coordinated have been viewed unfavorably:
I believe that (aside perhaps from the arguments against D), the criticism of your rounds is motivated not by the topics of the problems but by mistakes in preparation and problem selection. The coordinator is ultimately responsible for ensuring that problems are prepared well, low-quality tasks are rejected, and rounds are reasonably balanced. (As an example from another round you coordinated, CF round 907 F was not suitable for its position and belonged around the D2D range.) For this round in particular, there were several significant errors, most of which were relatively simple and will hopefully be easy enough to avoid in the future.
Below, I discuss my criticism of the problems in roughly decreasing order of significance.
H: By far the most major error imo; this problem should have been rejected immediately for being too well-known. The biggest problem is that the hardest steps can be bypassed immediately by Googling; this page is easy to find and gives a formula that is relatively straightforward to implement (at least, once you're willing to assume division by zero modulo 998244353 won't be a problem). Perhaps this problem might have been suitable for an offline contest, but it doesn't fit well on Codeforces, especially not at position H.
G: The error in the judge solution here is confusing to me. Was there not a tester solution, or did the tester's solution have the same bug? The obvious fix here is to hold strictly to the requirement that there should be multiple independent tester solutions (and these solutions should be rerun when limits are changed to ensure correctness).
A: Pretests excluded int vs long long. Of course, there's a school of thought that suggests pretests should be weak and allowing such hacks is a good thing, but at this point the majority of the community seems to prefer strong pretests, and in general it's assumed at this point that int vs long long is always included in pretests. Indeed, the Codeforces problem preparation rules specify that a test against integer overflow should always be included in pretests. This was probably an oversight that could have been avoided by preparing this problem further in advance rather than switching to a new task at the last minute.
F: Another problem that's much too easy to Google. Searching "partition biconnected graph into two connected subgraphs" got me to a StackOverflow comment about st-numbering, and googling st-numbering gave me a free solution.
E: A 1s time limit seems unnecessarily tight to me. In a problem like this where many contestants will use lazy segment trees and sets to achieve solutions of the desired n log n complexity, the time limit should be raised so as to be insensitive to the constant factor (accidentally accepting some n log^2 n solutions doesn't seem like the end of the world to me). I think 3s would have been more suitable.
The last few comments below are less significant; one is a matter of taste and the other is a minor clarification to a statement. Neither meaningfully harmed my experience in the round.
D: In general I don't like problems where the easiest solution path is to write a brute force and notice a pattern on small cases. This is a matter of personal taste, and I don't think this problem is objectively bad (I would not have complained about it if there weren't many other issues with the round).
B: The statement could have been made more clear by explicitly stating that a and b are the two largest divisors of x other than x itself.
As a result of these issues, and particularly because both of the problems relevant to my rating level (F and H) were largely tests of Googling ability, I had a negative experience in the round (despite gaining rating). Below, I suggest some specific ways to avoid these issues in the future:
Obviously, these guidelines are not sufficient to guarantee perfect contests, but I believe that following them would have prevented a lot of the negativity surrounding this round and several past rounds you've coordinated.
In any case, thanks to the authors/coordinators for their efforts organizing the round. Obviously, the contest was not as successful as the community hoped it would be, but the organizers still undoubtedly put many hours into devising and preparing the problems, and hopefully everyone involved can build from this experience in order to propose more successful competitions in the future. (Also, I should emphasize that the contest was not a total failure: I quite enjoyed C, for example, and aside from the preparation issues noted above, A, B, and E were pretty good for their positions.)
Finally, an appeal to the other critics of the round: let's stop downvoting/flaming the authors, please. Most of the setters were preparing this as their first Codeforces round, and as a result mistakes are inevitable (preparing a round is obviously an incredible amount of work and nobody is going to get it exactly right on the first try without outside support). Rather than blaming the authors, we should thank them for their effort; then, we should better organize round coordination so that these errors are caught prior to the start of the round.
I believe that a lot of these issues were a direct result of this round being the goodbye round. There is no way that the organizers didn't know about these issues. It is mentioned above that some testers even told them that H was a well known problem, I also refuse to believe that the coordinator didn't know about the rule of having at least 2 independent solutions for every problem. I believe if this was a regular round it would have been cancelled or at least postponed, but because this was the goodbye round cancelling was not an option. So Mr. 74TrAkToR, decided that rather than telling Mike about this, he would just risk it and pray that everything goes fine, or at least looks like that this was something he was unaware of.
I know that the decision if 74TrAkToR continues coordinating or not is not mine. However I do have a suggestion for the codeforces team that will probably prevent something like this from happening again in the future.
It would be nice if codeforces keeps about 10-20 tested problems of various difficulties as a reserve. And if something like this ever happens again, i.e. when we have a round that has to happen and there are problems in the round with known issues, they can just be substituted with one of those problems in reserve. I think this will help a lot and it is not that hard to maintain.
Definetely the most controversial contest of the year.
Universally hated =/= controversial.
It was'nt universally hated,I think that are more people complaining because of drop rating than because the quality of the round.You can find some blogs and comment doing the opposite for the same reason.
actually, it's some people supporting it because they don't want to get it unrated and lose the rating they gained rather than the opposite
Actually,most of those who think the contest should be unr don't even have the ability to solve E.They just want to find an excuse for their loss,isn't it?
In fact, the flaws pointed out by the participants in this contest, such as the original problems, wrong pretests, and google-able problems, are all objective issues, not subjective evaluations. These flaws exist, but contestants with lower scores prefer them to be noticed, while contestants with higher scores prefer them to be covered up. In fact, in most contests, these flaws don't exist, no matter what the participants say.
You are right.But unrated will result in fliping the mood of those rank 20+,and have little help with those who rank 20-.I think it is really a more serious flaw instead.
Well, I'm afraid I can't agree.
In my opinion, some participants are affected because of the authors' negligence, that means the contest is not fair. If a contest is unfair, it has to be unrated.
The mood of the participants with 'rank 20+' is because the contest is rated now. It's mistake of the CF managers that it's rated now. We should take measures instead of letting the mistake go on, right? =)
Well,I think both of us cannot agree with each other:) so it is meaningless to go on discussing.Just see what Mike will do.Whatever the decision is,I will accept it.Whatever if is rated or unr,I will thank the contest of providing me with courage and confidence,although it will be disappointing for me to participate in an unr contest as I have few chances to do it.That may be the reason why I seem to be selfish.
And thank you for discussing it with me.orz KHIN
True. Thank you too :)
I solved E.
Update: I cannot agree that
100+ participants passed problem H1 & H2 in contest, and nearly 200 for H1. Those people can get a performance of 2600+ easily, and, most of them cannot solve E in contest too. This, objectively, completely disrupted the standings of the top 200+ participants.
You know, Codeforces is for competitive programming, not for google programming or competitive OEIS.
It might not be a big influence on Div2 participants, but at least, the Div1 part of this round should be unrated (delete problem H, and then re-calculate the ratings of Div2 participants).
I think that more than 90% of the people are just caring about rating instead of they experience in the round.
Don't bother to apologize. Just quit.
Almost 4000 downvotes and still isn’t unrated? Seriously?
It means there is a bug in the standard solution of problem G, right? Is it fixed now?
should be fixed before systests because someone somehow passed
Oh I see. So the participants who tried to solve G are affected by the WRONG pretests, if I understood correctly?
The pretests of a problem are WRONG and some participants are affected. Isn't this enough to make this round unrated???
actually no, the standard solution is correct and fast but is $$$O(n^2)$$$ in complexity. traktor failed to prove why it's that fast before the contest but the problem is still used, despite that, the tests are all correct(but maybe weak?)
upd: sorry for the wrong information, he missed the
long longin his stdbut then how the hell did all the others get FST, they certainly deleted all the pretests (maybe except test 1).
as far as i know, there's now an alternate $$$O(n\sqrt n \text{polylog}(n))$$$ approach by djq_cpp.
I believe there also exists a $$$O(N\log A\log^2 N)$$$ solution.
Let's apply centroid decomposition, so we only have to consider two paths from the centroid to one of it's children. Let the gcd from the centroid to verte $$$i$$$ be $$$g_i$$$ and length by $$$l_i$$$. The answer is at least $$$g_i\cdot l_i$$$. Now, we want to merge every pair $$$(g_1, l_1)$$$, $$$(g_2, l_2)$$$, which contributes $$$\gcd(g_1, g_2)\cdot (l_1 + l_2)$$$.
WLOG, we assume that $$$l_2 > l_1$$$. What does this tell us? We know that we want to do better than $$$g_2\cdot l_2$$$, and that the best we can get is $$$\gcd(g_1, g_2)\cdot l_2\cdot 2$$$. However, if $$$\gcd(g_1, g_2)$$$ is not $$$g_2$$$, then it is at least half, meaning that it is not better than $$$g_2\cdot l_2$$$. This means that $$$g_1$$$ must be divisible by $$$g_2$$$ and that $$$g_2$$$ is the gcd.
We have that $$$g_1 = c\cdot g_2$$$ where $$$c$$$ is some positive integer. Let's try to restrict c. We want to do better than $$$g_1\cdot l_1 = g_2\cdot c\cdot l_1$$$, and the combined value is $$$g_2\cdot (l_1 + l_2)$$$. So we have $$$g_2\cdot c\cdot l_1 < g_2\cdot (l_1 + l_2)$$$ ⇒ $$$c < \frac{l_1 + l_2}{l_1}$$$ ⇒ $$$c < 1 + \frac{l_2}{l_1}$$$. To maximize the bounds for $$$c$$$, we have $$$l_1 = 1$$$, which gives $$$c\leq l_2$$$.
Let us consider a $$$(g_2, l_2)$$$ if $$$2\cdot g_2\cdot l_2 > \text{ans}$$$. We use $$$O(l_2)$$$ time to enumerate all possible $$$g_1$$$. First, let's (virtually) compress all edges $$$(i, \text{parent})$$$ if $$$g_i = g_\text{parent}$$$. This means that all paths from the centroid to a leaf have a size at most $$$\log A$$$ (with $$$l_i$$$ remaining the same). Let's pick a vertex $$$v$$$ on the condensed graph say the closest considered ancestor as $$$u$$$. We have that $$$l_v > l_u$$$ and that $$$g_v = \frac{g_u}{c}$$$ where $$$c\geq 2$$$. We want that $$$2\cdot l_v\cdot g_v > g_u\cdot l_u = g_v\cdot c\cdot l_u$$$ ⇒ $$$\frac{l_v}{l_u} > \frac{c}{2}$$$. Now, the worst case is when $$$c=2$$$, but if we consider one subtree at a time, it can only happen once since we if there are multiple children that reduce the gcd by half they are "merged" (note that we do not actually merge the vertices but we store the largest $$$l$$$ for each $$$g$$$). So, we have one chain from the root to a leaf that has size $$$\log A$$$ and therefore uses $$$N\log A$$$ operations. Then, we must have an edge with $$$c\geq 3$$$. Now, we repeat the same process but we know $$$\frac{l_v}{l_u}\geq \frac{3}{2}$$$, meaning that this can only happen $$$\frac{\log(N)}{\log(1.5)}$$$ times, meaning that we consume a total of $$$N\log A\cdot\frac{\log N}{\log 1.5} = O(N\log A\log N)$$$ time.
Edit 1: I think there may be some inaccuracies with the proof in the last paragraph if you have different $$$c$$$ diverging from a vertex, but I think a better way to phrase it is something like this: first, take out the chain with $$$c=2$$$ using a factor of $$$O(\log A)$$$. As for the size, we know that $$$\frac{l_v}{l_u}\geq \frac{3}{2}$$$ meaning that $$$l_v - l_u\geq \frac{l_v}{3}$$$. This means that if we consider adding edges as a cost, when we spend $$$O(l_v)$$$ time we expend $$$\frac{l_v}{3}$$$ edges, which proves the sum of $$$l_v$$$ to be proportional to $$$N$$$; this still has to be multiplied by $$$O(\log A)$$$ since we have to account for taking the $$$c=2$$$ chain out.
At each centroid decomposition level, we need $$$O(N\log A\log N + N\log A)$$$ for calculating each $$$(g_i, l_i)$$$ and enumerating every pair. Therefore, the total time complexity is $$$O(N\log A\log^2 N)$$$.
239751483
Please let me know if I made any mistakes.
Cool, thanks for the bound writeup, I think my solution was essentially the same.
The problem is not to unrate or not. If we just simply unrate rounds and rounds, and doing "You are wrong. Here's why" to keep on rating rounds and rounds, we will just be like using mosquito swatters, showing dissatisfied attitude only when we participate in a round and after that.
To be honest, I personally don't want this round to get unrated (as my high rating increase), but I have no objection to unrating it because this is not my real programming level at all. But it reflects, that, you are just casually picking problems, casually preparing problems, casually immersed in your own feeling about how good this problem will be appreciated after contest. As a result, the topped participants on the list (but more than the top, the situation can be seen all over the list) are casually selected.
I have given more than 10 problems before (I estimate that 2500 people in all solved at least one of them) on some website and I can give some suggestions about it. At the first stage you were in fact giving shit problems. You saw a conclusion somewhere, or you discussed a lot of situations, and then just used it as your own problem. However, when you some time came up with a better problem, you will understand the feeling of what a problem recognized good is based on and whether an approach is interesting. Motivated by this, you will produce much better problems (it also requires something like inspiration and your ability to make hackive datas before contest). You can rate your own problems with quality (but better not difficulty, I have never predicted my problem as hard as what contestants felt when facing it). Thus there will be a cooperation between problemsetters and testers, which will help each other if it goes into a beneficial cycle ———— help when rating problem difficulty in thinking of a solution and implementing, and even worse, if your problem is an NP-Hard or something like that.
So I'm wondering, if all your problemsetters are at the first stage, giving just shit and average-leveled problems, and none of the testers have ever taken notice of it? Or as a coordinator, are you simplying ignoring all the feedbacks and believing you have the sharpest eye in the world? I want to know your time spent on solving problems D, E, F, G, H1 and H2 and I wonder if you really have the ability to solve all the problems without referring to the editorials, and even rate the problem as where it will be put? Come on bro, from several past experiences, I think you should learn. Don't just be an ostrich which sticks its head to the sand and avoid meeting problems and leave your butt to be wiped by Mike! And if your coordination are so perfect, why do you insist on changing A before the contest? Have you ever distinguished what would be the main problem not checked and may happen during contest?
btw, I am very confused why there is such a long queueueueueue for even div 1+2 contests in codeforces. Are coordinators' efforts wasted in shit & non-original & not-even-cared-by-problemsetters problems like yesterday? Yes, I use this word to stress, not only you, but also the problemsetters, don't care about the feedbacks and even have no face-to-face communication with others. Have you experienced writing a 20KB generator and stay up late to modify the datas? Have you experienced the joy after an entire contest when seeing different solutions from backstage pass (or not) exactly as intended? And, you, as a coordinator, should always remember, with great power, comes great responsibility. If you don't care, he doesn't care, now it will be us contestants caring about our rating.
If this comment is deleted due to violation to some rule, I will definitely not make a second comment on this issue. Good Luck, 74TrAkToR.
YOU ARE RIGHT. HERE'S WHY.
orz.
orz.
You are right and you deserve GM!
So,why this contest is still rated? Wrong G,copied H. I also think D is boring. Do you think brute force is interesting? I spent much time to think about maths. Some people even passed H1&H2 just by OEIS. It's really unfair because some people can't cheat themselves,but the others can. And people who cheated themselves got higher rating instead. And this problem is also well-known in university students. I can't believe all the testers didn't notice this problem. So,how did H pass?
So,why this contest is still rated? Because you are 74TrAkToR? I heard of this contest before,everyone who knew something about it told me not to participate in the contest. I myself also had been f**ked by you before. I realize your contest isn't worth to trust now.
My biggest mistake is not to trust my friends but to believe you wouldn't f**k us any more. And you? You don't have any mistakes because you are 74TrAkToR. You messed up a lot of time before,and you messed up again at the end of 2023.
I don't think G is the main problem, and it seems that djq_cpp already gives a $$$O(n \sqrt n)$$$ solution to solve it, although it is very difficult to implement and the constants are huge. I think the bigger problems are F and H. F is the classical problem for bipolar orientations, while H can be found even on OEIS. Also I don't think D is a good fit for Codeforces, the problem is entirely a math problem, and most of the way people solve it is to list all the cases and then look for patterns. D is not good but ok. But based on F, G and H, I think it has to be unrated.
I also think G isn't the main problem,so I only mentioned it in one sentence. I agree with you,but he is 74TrAkToR,he won't listen to us.This contest won't be unrated as well.
So why did this round rated?
Please, learn from other coordinators about how to coordinate a round well. And, things like changing problems can be done much earlier, don't delay it to the last moment. This is not history. If you don't make any changes, the history will only repeat itself.
Since testers have already indicated some problems in feedback, why are you ignoring these issues as coordinator? I don't think this is a round "put a lot of effort". And this round is not your first trash round, you never seem to take it seriously.
It's RATED??? What a big joke!
You are wrong,here is why.
What is history ? It was just yesterday! I couldn't even solve your last moment problem A.
It's okay bro, mistakes make us better !
:// why so many downvote
FH are all original questions, which is not good
It shoud be unrated. I feel like I wasted a lot of time because of this round. And...... where is Tutorial ??? This is a bad ending of 2023.
1916F - Group Division
Why is hack gets Unexpected verdict
Is there something wrong with the std ?
Good morning. Is it still rated? :D How can I change my voting???
You are wrong. Here's why.
You did mistakes and wasted our time ,ok, maybe we will forgive you for wasting our time , but why this round still rated ! Why we should pay for your mistakes!!
None of his mistakes were related to you. You had terrible performance and there is no one else to blame for that but yourself. You guys who hand't touched G and H decided that you can just jump on the bandwagon and keep spreading hate towards him. Just get better, lol.
Actually they were , problem a has a weak pretests because he changed it in last minute, and b was a terrible Problem . But even if problems like h,g and d I hadn't touched them, i still have the right to claim that round shouldn't be rated even if i didn't participate at all, he did mistakes and he should pay for them not the contestants. I don't spread hate towards him , he is just a terrible coordinator, every contest with his name on it has many mistakes and bad quality problems , so he is the one who should get better,lol.
I don't like that there are problems with weak pretests either, but it's acceptable in codeforces, as there is a whole hacking system dedicated to it. While this round should definietly be unrated because of mistakes in problems H and G, you guys who weren't affected by these problems shouldn't be complaining here as if your performance was terrible because of the bad coordinator, it was perfectly normal round if you consider problems A-F.
Yup, problem D is amazing :)
Amazing or not, there is nothing wrong with problem D.
Funny how I got downvoted by people who aren't capable of solving problems like D.
you're correct
I firmly boycott 74TrAkToR, even a OO won't attend any contests held by him.
Пацаны не извиняются.
After reading this post I think this round's quality is just like some contests on Luogu
true dude
i love what sort of an animal you are
Username:rafi1321974 Contest link: https://codeforces.me/contest/2007 Problem Link:https://codeforces.me/contest/2007/problem/D problem:In the D number problem I write code myself in Python but there show a error in 7 test case.So then i write code in c++.But it see that now my code is same to someone .But i write code in my own now i give my documents and logic.You an check my problem
Problem Logic Explanation The provided C++ code solves a graph traversal problem using Depth First Search (DFS) applied to a tree. The tree consists of nodes, each with a weight of either 0, 1, or 2. The goal is to compute a result based on the distribution of node weights. Problem Breakdown: ------------------- 1. Input: — The number of test cases is provided. — Each test case has: a) A number of nodes and the edges connecting them (forming an undirected tree). b) A string where each character represents the weight of the node: 0, 1, or '?'. The '?' character indicates a node weight of 2. 2. Node Weights: — 0 and 1: Regular node weights. — 2: Special weight, which needs specific handling. 3. Goal: — Perform DFS traversal starting at the root node (node 1). — Count occurrences of the node weights (0, 1, and 2). — Identify leaf nodes (nodes with no children) and update the frequency of their weights. — Apply special counting logic for nodes with weight 2. 4. Special Count: — If a non-root node has weight 2 and children, it contributes to a special count. DFS Logic: ----------- — Mark the current node as visited. — Recursively traverse unvisited neighbors. — If a node is a leaf node (no unvisited children), update the frequency of its weight. — If a non-leaf node has weight 2, increment the special count. Result Calculation: ------------------- — The result is based on the weight of the root node (node 1): a) If root node weight is 0: Result depends on frequency of nodes with weights 1 and 2. b) If root node weight is 1: Result depends on frequency of nodes with weights 0 and 2. c) If root node weight is 2: i) If counts of 0 and 1 nodes are equal, the result is influenced by special_count. ii) If counts of 0 and 1 are unequal, the result depends on the larger frequency of 0 or 1 nodes and half of 2-weighted nodes. Summary: --------- The code parses the input, performs DFS to count node types, and calculates a result based on the balance between nodes with different weights, paying special attention to nodes with weight 2.