Editorial has been updated to include a proof that a deterministic solution doesn't exist.

I think it worked out fine because it didn't seem like anyone with a 2^(-29) sol didn't pass, but we probably should have made it 60 bits instead of 30 to make the failure probability much lower than 0.65%.

Unfortunately it seems that we still cannot uphack A-D. Maybe another setting needs to be changed?

lethan3 will apply bitset to win ICPC NAC 2026 ✅

IanDeHaan you put our third teammate in as ufecto but it's actually my goat lce4113

I've had some intended solutions that required enough constant factor optimization to take me ~100 submissions to solve, so I personally wouldn't be in favor of that. I think there are other ways to filter bot accounts that work better (for instance, no more than 50 accepted submissions would work).

Is there any way we could find out which companies are doing it? That way at least we could bring the concern to a larger population, because I think a lot of people who aren't on Codeforces would see large companies ruining hobby platforms for profit as ethically unacceptable.

That being said I have no idea how, and assuming it's not possible, I do think that some action needs to be taken to reduce the inflow of AI submissions. An alternative I might propose is that any account that solves some large number of problems (maybe 25-50 as some other users suggest) within a single day gets banned from the platform.

Thanks for the visualizers! I found them to be quite helpful during the contest.

I was going through submissions to hack, and I received this warning:

"Recently, your account was used to crawl. Please change your password to prevent your account from being used for unauthorized activities."

I can't view submissions anymore. Is there a way I can get this permission back?

Get some of his friends interested! One of the best motivators is a good community. I started at 14 and that's what kept me engaged until I developed strong intrinsic motivation.

It is with great regret and sorrow that I must announce I cannot make it to the in-person edition of BAPC 2025. I tried my best to use what little bandwidth I have to contribute to the organization of this wonderful contest, and I'm happy to say that the organizers really didn't need me at all. They've done a great job putting this contest and event together and I think you will enjoy this edition of BAPC very much. Good luck to the participants!

I get that you're probably gaslighting when you bring up IQ in every discussion about rating, but here goes my take on this.

A few years ago even expert was an impossibility. I had no CS background and little math background, and as you can see on my rating graph it took many contests for me to even reach specialist. When I reached expert that in itself felt like an upper bound. However after almost a year of training after this I made it to candidate master for the first time. You often argue that for a "normal" person like me candidate master is the upper bound, and for almost 2 years it felt like I didn't improve. I kept working at it because while rating is a nice thing to have, the joy I got from solving each problem and trying to learn new things outweighed any sadness that I felt from not "improving" in rating. Eventually I got master.

Maybe I'm at peak right now, or maybe eventually I'll get red, but there's no way to tell at this point. Will I give up? Absolutely not. Do I believe its impossible to get red? Absolutely not. Was it an impossibility even a few years ago? For sure.

TL;DR Just be patient.

Not really an algorithm, but Sprague-Grundy is pretty beautiful :)

Thanks for this blog! What can men do to fix the gender gap in competitive programming? In my local competitive programming scene, almost everyone is male (most are asian males) and even though there are no barriers to entry, almost every new person is male (and asian). I can see why thats very scary to people who aren't in that community (I've had similar experiences in communities outside competitive programming) and I would like to change it, but I'm not sure how I can. If you have any tips I would really appreciate it!

I guess there's also another part of this question, which is whether it's condescending to actively try to recruit girls. I'd like to hear your thoughts on that as well.

I'm not really an expert on this but I think you may also run into trouble with solutions that have amortized complexity. I have definitely seen some problems where I have a solution that looks like $$$O(n^2)$$$ but turns out to work in $$$O(n \log n)$$$ due to amortization—in these cases, I have a hunch that the solution is fast but can't prove for sure. So in submitting that solution I could accidentally cheat if it turns out my hunch is wrong.

I understand maybe you think that everyone should have a full proof of their solution before submitting, which is somewhat reasonable, but I don't really like the idea of having a system where making a mistake when solving a problem could mess up the contest for other people.

Perhaps another important point in this conversation is that Durov is making billions of dollars off of Telegram (including the illegal activities that occur on the platform). If he is profiting off of providing a channel for people to commit crimes, at what point is he a bystander vs. an enabler?

I haven't formed a very strong opinion on Durov's case (and it's also certainly being used as a sort of proxy intelligence war between NATO and Russia), but I think it's important to examine Durov's role with a little more scrutiny.

Minilek (Jelani Nelson, professor at UC Berkeley) is successful enough to have a wikipedia page :)

As a tester

The facts:

• Benq did not make IOI 2024

• myst-6 did not make IOI 2024

• Benq tested BAPC 2024

• myst-6 tested BAPC 2024

Perhaps there are better skill level indicators than IOI...

Just because you comment 1e9 times on the same blog does not mean your contribution will magically increase from -30 sir

As the author of J, you just validated my entire problemsetting career (after it was completely destroyed by Benq who solved G in 3 minutes 56 seconds in testing). Glad you liked the contest!

We had a lot of fun organizing and hope you enjoyed it!

BucketPotato orz. I hope everyone enjoys the problemset!

I believe this is the original "work hard bro" comment. There could be an older one though, I'm not sure.

1. Get master ( same as last year :( )

2. Complete my goal of 200 2400+ rated problems without editorial (over 2023 and 2024)

PurpleCrayon and GusterGoose27 will tie with cnnfls_csy to win IOI 2023!

I've participated in LIT for the last two years, and I highly encourage others to try it out! LIT CTF in particular was a really fun introduction to CTF contests (with the added entertainment of rickrolls, among us references, and obligatory codetiger orzes). I also really enjoyed the standard round from previous years, as the problemsetters worked hard to make high quality problems.

I look forward to participating again this year!

I've noticed a lot of comments like "problem F barely affects anyone" or "we just should remove F" or "CMs and above should have solved D anyway so just removing problem F would make it fair". I don't believe that my personal experience lines up with these comments/opinions.

I solved A,B,C relatively fast (within 25 minutes), and at this point there were less than 100 solves on both D and F. I read D, noticed a solution, and then quickly checked the solve counts of each problem (I noticed that the solve counts for F were increasing very quickly; when I checked, problem F had around 90 solves while problem D only had 70-80). As a result, I chose to hold off on implementing D to try and implement F (perhaps this is my fault, but it is what I did).

In some way I was very unlucky, because I noticed a square root decomposition solution quickly (which ended up being too slow), which I thought could pass the constraints (since $$$q \cdot \sqrt{n} \approx 5 \cdot 10^8$$$ and the time limit was 4 seconds). The solution involved the observation that either $$$|r - l| \leq \sqrt{n}$$$ or the answer was at most $$$\sqrt{n}$$$. I thought this would be fairly straightforward to implement, so I first tried it with $$$O(n\sqrt{n})$$$ memory (which did not work). I got unlucky again when I realized that the answers could be computed in a different order to use only $$$O(n)$$$ memory, so I spent some time implementing this, thinking it could work.

At this point there were about 40 minutes left. There were about 150-200 solves on F and maybe 400-500 on D (I don't remember exactly), but since I had already implemented a solution and I saw 200 solves, I assumed that it might still be possible to implement the solution in time (also, solving D at that point would have given me way less points than solving F, and I already had a negative predicted delta). Perhaps this was my fault, but with the incorrect solution and the solve count, I was tricked into believing that I could solve the problem.

By the end of the contest, I had reimplemented my solution, but I realized that it involved an extra log factor from a segment tree that I used to calculate part of the answer. At this point I had about 15 minutes left, so I thought about how I could reduce this. I couldn't think of anything, so I ended up not solving the problem. From about 90 minutes remaining to the end of the contest, I felt like I was close to solving F, so I didn't even go back and attempt to solve D.

At the end of the contest, I felt that I was pretty close and I just needed to practice more and upsolve the problem (I was ok with getting -70 delta). I was fine with the result even after realizing that the problem had appeared before. However, after learning that the older problem's link was posted in the round announcement, I no longer felt that the contest was fair. The contest no longer gave an advantage to people who practice solving harder problems, but instead gave an advantage to people who share code with each other during the contest. And somehow, I was caught up in the middle and got the worst of the outcomes of this (-70 delta and false hope that I could solve a problem that I probably couldn't solve given 3 weeks).

For me personally, removing F and recalculating rating changes would probably not have an impact on my rating change. Perhaps not a lot of people are in this situation, but I wanted to make others aware that it could happen as well. I think that the best decision would be to just make the round unrated rather than to try and recalculate the ratings after removing a problem (or worse, to not do anything at all).

This looks like a really high quality contest, I look forward to participating in it! I hope the novice round will be fun (and that I will be able to AK).

Thank you to the authors for putting together the problems!

I was looking at submissions for problem C to see if my solution could be hacked, and I found these two solutions: https://codeforces.me/contest/1661/submission/153183418

https://codeforces.me/contest/1661/submission/153182822

They are identical, but from different accounts; both were submitted during the contest window (clearly cheating). I was wondering if there is a way to report these accounts/submissions; will they automatically be taken care of since the submissions are identical, or is there some other procedure that should be followed?