Hello, I'm Sanae.
In ufc+8 zone, I watched the entire contest at midnight. Too many cheaters emerged and we were busy banning them. I noticed WA submissions on 2228B - Remilia Plays Soku, 2228C2 - Cirno and Number (Hard Version) and TLE submissions on 2228D - Sanae, Cross and Color, which was intended. Soon after the end of the contest, I released the editorial. The bad news was, the complaints flooded the website and my contribution fell from +91 to -92. After reflecting on the issues, I was anixous and blamed myself. Now, I make an apology to point out the problems to avoid them next time. Sorry, everyone!
- Too many cheaters emerged. Comment 1
- The problems stay in the queue for a very long time bothered contestants.
- B is easy but many contestants misssed the corner case. And C has many corner cases. Comment 1, Comment 2, Comment 3
- The time limit of D is too tight. Comment 1, Comment 2, Comment 3, Comment 4, Comment 5, Comment 6
- Problems are too hard. Comment 1
- The samples are weak. Comment 1, Comment 2
- The contest causes a lot of rage. My heart is broken. (;﹏;) Comment 1, Comment 2, Comment 3
- C>>D Comment 1, Comment 2
- E is AI-solveable. Comment 1
- E takes too much time and F is so easy. Comment 1
There are still problems unmentioned, but they are enough. I take 4 most seriously. Although 2 seconds can be enough for the best solution, some correct solutions are killed, which obey the principle of problem setting. During testing, D was accepted by different solutions and even $$$O(n\log n)$$$ solution can run within 1.1s. Given that, I didn't modify the TL. $If I had set the TL bigger, the contest could reach its expectation.
I'm sorry for the weak sample. The $$$n=2$$$ in B bothering everyone is intended, but I should have put it in samples at CF. But I didn't even realize C has many corner cases. I set $$$a$$$ up to $$$10^{17}$$$ to reduce the diffculty. In the original version, $$$a$$$ could be up to $$$10^{10^5}$$$. Although the corner cases are the necessary parts in the problem, I missed them. Sorry again.
We tried our best to capture the cheaters, and you can point them out on your own. I'm sorry for that. They are easy to find, though.
Indeed, E should have more scores than F, since it takes more time. I have done the wrong evaluation for the difficulty.
I don't want to show CNOI style initially and make the problems easier and easier. I only hoped that problems would be interesting. Some are said to be standard, but I still believe they have the value of thinking. A is easy. B is easy but have corner cases, which is intended. C is not so easy and happens to have many corner cases. TL for D is too tight. Score of E is too low. F is ok but misses the $$$O(n)$$$ solution. I have made so many mistakes even after so much testing. I let contestants down, failing to keep my promise. (T_T)
I will not author problems any longer until I'm mature enough. I'm guilty and carry this regret from then on. I hope my mistakes serve as a lesson. I have let you down. Sorry...
May love persist?
Dude, you don't have to apologise for this. Downvoting a round because of your poor performance is not a healthy habit.
Though I got a lot of Wrong Answer from A and B.And spent a lot of time on C instead of D.I still think your contest is good.Thanks for your contest!
please don't stop authoring problems. people are mad when they perform poorly. sure, the contest quality was not great but please keep authoring problems and please don't feel guilty for trying.
To the Codeforces community
I hope you see what you are doing by downvoting rounds whenever you don't like anything about it, without any regards to the problemsetting team.
Whenever you downvote a round, the authors of the round is less likely to write another one. Soon enough, there will be nobody willing to write rounds for Codeforces. And you all will be wondering what happened.
Only downvote if there is a serious issue. You doing bad isnt a good reaosn to downvote.
Yes, downvotes can hardly be a pleasant thing for the most people.
But how else do you suggest learning from mistakes, other than through negative feedback? Would you argue that authors are more likely to pay attention to the criticism and take it seriously if they mostly receive upvotes for their poorly prepared contest?
However, I think there also might be a problem (that is, a way for improvement) with how criticism is filtered and managed by setters. Sanae listed 10 points and for at least half of them they can't take any blame for. Say, how can an author take blame for cheaters unless they leaked the problems?
Other than that, I think C and D were somewhat mediocre. Both are dull implementation and attention exercises.
For me the contest wasn't the worst in terms of its quality. I encourage Sanae to give it another try in the future, with a work on mistakes done.
the point is that the "mistakes" are just people bitching for no reason, which again, hurts problemsetters for no reason
I can't agree that there is no reason, but yeah, I think some of the comments indeed have no value. That's why I said about filtering.
I think the most important reason why this round was voted down is that the weak sample was overlaid with queueueueueforces. According to my research, many contestants were very angry when they submitted problem D and found that C1 and C2 had both received a wrong answer on pretest 2 predict
If there weren't such a long queue at the beginning of the round, this round shouldn't have been downvoted so severely.
I know I don't have any rights to write about the contest, but ... umm... isn't problems are interesting when they're tricky? aren't they part of your learning journey? aren't you here to LEARN? if not, sweetheart, codeforces isn't going to make your CV heavy. Try doing different real life projects for high paying jobs you're targeting. But I can understand when people are angry because they might have missed the chance to reach next rank by little. idgaf
Plz don’t apologise, it was a great round except the queue issue which was not your fault. People just have their own excuses to downvote and blame the authors and their problem setting over their own bad performance. I really like the problem C.
The problem was pretty good for D and quite intuitive .... Only the tl was a bit tight rest was on spot ..... can anyone tell me why O(n) solution — first performs bad than the O(nlogn) solution second
For what it's worth, thank you for the contest.
Problems do have corner cases. Rather than hiding my head in the sand and asking for cleaner problems, I prefer to acknowledge the existence of tricky cases as well. I usually strive to find a most general solution, with the least number of corner cases. Also, I teach my students to strive for the same. It helps in the long run.
In problem D, my issue was input speed. In problemsetting, personally, I try to avoid ones where input speed is important, unless the whole problem is explicitly about constant optimizations. Nevertheless, such problems do exist, and I learned a bit of new stuff while dealing with problem D.
Problems do contain math. I didn't solve problem E. Also, I don't like it: statement is too formal and arbitrary to my taste. However, it's a personal opinion, and I'd hate to impose it on other problemsetters: if you do like such problems, great, go on setting them. Also, it's interesting to read the tutorial, even more so since it shows three approaches.
Accepting constructive criticism is important. However, don't let the popular opinion completely override what problems you set. If you, coordinator, and testers agree on something, most probably it's OK to try.
Yes, D is not good enough. By the way, why E is formal and arbitrary is to make the statement clear and neat. In the original version, I wrote a short background and deleted finally.
the problems in this round were not bad at all. i especially enjoyed upsolving d, but i guess weak samples and a long queue led to a great many frustrated downvotes. as for me, the weak samples were the final push to incorporate stress testing into my workflow. so don't be too discouraged, everything has a bright side! cheers!
An update to downvote on codeforces
- MikeMirzayanov circa 2026
You shouldn't mind those comments — they have never made a round yet so they don't know how hard it is. Your contest is great, and I love it. Let's just don't talk about the fact that I lose 60 ratings because of your contest. NVM, there're nothing here.