Mike, alts and unfriendly comments I know that the following content might not be welcomed and I will lose lots of contribution, but I should write something about such a bunch of chaotic comments under there. I may delete this tomorrow if I find my words useless. ### What happened Several months ago, Mike posted a blog about banning [user:orzdevinwang,2024-10-07]'s alt, [user:zh0ukangyang,2024-10-07], and more alts. Lots of participants, including myself, were angry because [user:zh0ukangyang,2024-10-07] is such a talented coder and codeforces didn't treat the rule of single account seriously at that time. Then more and more alt accounts with high rating are reported, including mine. After that I didn't use alts anymore, and most of high-rated Chinese participants(100% in my friends) discarded their alts and followed the rules. We might think that the rules are not that perfect, but rules are rules. I believe that no one will say "Banning [user:zh0ukangyang,2024-10-07] is unacceptable". But I know what y...

Full text and comments »

WF23 Luxor _(Long blog ahead. Reader discretion advised.)_ I feel we have a rather interesting WF story (note: we did the 47th finals, as the University of Maryland, <b><i>as 2 people</i></b>... explained below), so I figured I'd tell it. Exposition ---------- It actually starts way before WF, too. Our teams were decided on (some form of) merit based on an internal qualifier. My team for the regional and NAC (North America superregional) ended up as {me, [user:Navick,2024-04-24], [user:mr.banana,2024-04-24]}. We did [fairly](https://nasouth22d1.kattis.com/contests/nasouth22d1/standings) [well](https://nac.icpc.global/wp-content/uploads/2023/06/NAC23-scoreboard.pdf) in those. However, for WF, my two teammates only had single-entry visas for the US, meaning that if they left, they couldn't come back into our country. So they couldn't really go with me... Luckily, in such situations, ICPC lets you just replace your teammates. So we held another internal qualifier, and ended up with two ...

Full text and comments »

My strategy at AtCoder & CS Academy There's been much controversy lately about the late submission strategy not penalized by scoring systems of AtCoder and CS Academy. Most of the relevant discussion happened earlier at http://codeforces.me/blog/entry/53431 and http://codeforces.me/blog/entry/53449. I have a lot of thoughts on the topic, so I've decided to share them in a separate blog post. I really like the strategic part of programming competitions. Of course, problem solving is more important. But every contest consists of multiple problems, so there has to be a way of comparing participants which performed better at different problems. There's a huge variety of scoring rules, and I find it truly amusing. The "submit after solving all problems" strategy looks widely attributed to me now, mostly due to [user:Petr,2017-07-24]'s remarks regarding my participation in AtCoder contests. In my opinion, it's wrong for multiple reasons. A closer five-word description of my strategy is "implement after solvin...

Full text and comments »

One final blog about Codeforces Round 1033 and CodeNite 2025 Hi everyone, writing this blog after the recent rollback, getting red back to top of leaderboard as everyone so much desired, and making the overall leaderboard so much better. If you disliked the announcement/tutorial of [contest:2120] or posted a genuine critique because of problem quality or or because of some other situation which we could've improved, then this blog is not for you, I don't have any hard feelings and understand why you did it. The dark side ------------------ First of all, everyone who cheated in the round, I really reprimand you all and hope you all are fairly retributed for what you all made me go through and the deeds you've done. I especially hope all those who run the back alley solution distribution during the contest are gone for good and that they are especially retributed for their wrongdoings. (I don't think we can stop GPT or that it can suffer "retribution", so not sure what to say about it. Anyways, people who use GPT for wrong are responsib...

Full text and comments »

Finding all primes up to half a billion in 83 ms Hi everyone! I was itching to apply SIMD in another problem, and this time my attention picked the following problem: --- **[Library Checker &mdash; Enumerate Primes](https://judge.yosupo.jp/problem/enumerate_primes)**. You are given $N$, $A$ and $B$. Find $\pi(N)$, and print all $p_{kA + B} \leq N$. It is guaranteed that $N \leq 5 \cdot 10^8$ and that the total number of primes to print is at most $10^6$. --- After some tweaking, I managed ([#336744](https://judge.yosupo.jp/submission/336744)) to reduce my time consumption in this problem to **83 ms** and memory consumption to **25 MiB**, which makes it a top-1 submission on the Library Checker. For comparison, the top-2 solution ([#148169](https://judge.yosupo.jp/submission/148169)) takes 172 ms and 15 MiB, meaning that what I describe here is a 2x improvement in terms of the running time. We can also compare it to what appears to be the state of the art of sieving, the [primesieve](https://github.com/kimwalisch/...

Full text and comments »

Winning a Gold Medal at IOI 2026 Two months ago, I posted this [blog](https://codeforces.me/blog/entry/154231) on training for IOI Gold. Despite not feeling particularly good about my chances at a gold medal going into the contest, luck was on my side and I did end up winning a gold in the end! Therefore, I'm writing this blog as a follow-up to my previous blog, as well as to share about some of my experiences and opinions about IOI and training for it. ### Final months leading up to IOI In the two months between posting the previous blog and participating in IOI, I almost entirely stopped doing problems from CF and AtCoder, and instead fully focused on OI-style problems. I had previously done virtual participations in IOI 2016-2023 (and participated live in IOI 2024 and 2025), but deliberately didn't read most of the solutions so I could use them to gauge my progress as I periodically come back to them and upsolved. I worked through most of the problems that had room for improvement, and managed to pass the g...

Full text and comments »

Introducing CF Stonks — A Virtual Stock Market Based on Your CodeForces Rating! **2/11/2023 Update**: There had been a month-long hiatus due to the CF name change magic (and my own laziness oops), but everything has been fixed now :D! However, I know a lot of people have gone inactive due to the month-long market closure, so I am hoping to **relaunch StonksCF** in the near future with some new mechanisms (dividends, automatic algorithm trading, etc.). Let me know if you have any ideas! In the meantime, feel free to continue exploiting the way-over-inflated economy and embrace your inner capitalism >:D **Final Update**: We now have our own **DISCORD SERVER**. Join us for the latest information and engage with the community at [https://discord.gg/AeH2GTwkF5](https://discord.gg/AeH2GTwkF5) All the announcements will be there instead of on this page Greetings Fellow CodeForces users! I am proud to introduce my latest project, [CF Stonks](https://codetiger.me/project/StonksCF) (aka StonksCF), which is a virtual stock market based on your CodeForces rating. ...

Full text and comments »

Level up to Specialist. Some tips and tricks! Hey everyone! I finally hit Specialist in the last Div 3 round! However, this wasn't an overnight thing. It has been almost 1.5 years since I started my CP journey. To be honest, my growth was undeniably slow, mostly due to some specific mistakes I made along the way. I’m writing this for my fellow coders who are grinding every day but seeing no results. I’ve been there, and I want to help you break through that wall. I’ve tried to keep this practical and straight to the point. Here are the 5 lessons I learned the hard way: * **Rating is Just a Number (Seriously)** The first and most important step is to change your relationship with that number on your profile. For a long time, I was obsessed with my rating. Even when I was Grey or Green, I put immense pressure on myself to "rank up." This actually backfired; my brain couldn't perform at its fullest because it was too busy worrying about the results rather than solving the problem. Once I started giving less importanc...

Full text and comments »

I tried different warm-ups before contests... **Hey Codeforces,** I wanted to share my experiences experimenting with different warm-up routines that can help improve focus during contests. I believe this topic is particularly relevant for those whose time zones make contests fall during early mornings or late nights. Getting into the right psychological state before a contest can significantly boost your performance! Personally, I usually participate in contests from 10:35 pm to 12:35 am due to my time zone, which makes it challenging to stay focused, especially in the second half since I’ve typically been awake since 7 am on school days. To address this, I decided to try out various warm-up strategies suggested by experienced members of the community. My goal was to find out which ones work best for me (though, of course, this is subjective!). This blog is divided into the following sections: - Tips on warming up from experienced participants. - My experiences and reflections on trying different warm-up combinati...

Full text and comments »

UpSolve.me | Efficient Codeforces Practice  Hello, Codeforces community! First of all, I would like to thank [user:MikeMirzayanov,2020-06-21] for such an awesome platform. It's been almost a year since I joined and ever since I was always in the dilemma which question to solve. It's a commonly accepted fact that one should solve the most solved questions first. Many other recommender websites out there follow the same pattern &mdash; solved count for the win. However, after some days of following it, it kinda felt repetitive. Yes, the ones that everyone can solve would be the easier ones. But how do we determine if it can teach us something new? Codeforces has the problem rating system which gives a more accurate view of which question is easy and which is not. But as with the case of solved counts, rating can't be considered as the sole differentiator. After having analyzed for some time, I found that there are three important factors that decide which problem is best suitable for ...

Full text and comments »

Operating Systems Notes If you are struggling with the right track and a 10-min read for Operating systems from beginning to end, you're at the right place. Even I am not sure where to start, but we would figure it and if you are reading this, I have successfully completed these notes or at least am on right track. Let's start our struggle for OS: ### The Track I found many resources to learn, let's just list all: (Galvin Concise PPTs)[https://www.os-book.com/OS9/slide-dir/index.html]: These were great but I felt that these are a little bit too much, so here are the chapters we would do: 1. Processes 2. Threads 3. Process Synchronization 4. CPU Scheduling Algorithms 5. Deadlocks 6. Main memory 7. Virtual memory 8. Virtual Machines > I am skipping the introduction of OS for now as it was not that important, this is going to be a fast article that works like a last night dose. <spoiler summary="A simple introduction(May skip if you are already familiar)"> An Operating system is like a...

Full text and comments »

Useless CP Information #6: How Much Harder is Div 1 vs Div 2? Hey guys it's your favorite <strike>purple</strike> not anymore CF shitposter here. We're back to hopefully non-controversial useless information lol I think anecdotally we all have this gut feeling that it's harder to get a X perf in a div 2 than a div 3, and in turn harder to get X perf in a div 1 than a div 2. Is this urban legend or a real phenomenon? Methodology ------------------ I took several of the most recent separated div 1 and div 2 rounds where the div 1 round has the same problems shifted by 2 positions. For each contestant in the div 2 round, we'll: - Calculate and record their div 2 perf - Convert to estimated div 1 score by dropping problem A/B, adjusting timestamps on the rest of the problems, and recalculating the score based on the problem values for div 1. (We will discard anyone who solves C/D/E/F before A/B) - Calculate and record their div 1 perf from the new score we calculate. ### Calculating performance (CP lesson time!) Using FFT, our...

Full text and comments »

Is Ford-Fulkerson(DFS) + blocking flow + randomization hackable? Hi guys. When I was doing a problem that basically requires a reasonably fast maximum flow template, one thought came across my mind: Is it possible to hack Ford-Fulkerson(DFS) + blocking flow + randomization? I know that using BFS to run Ford-Fulkerson is much safer, as there is a [test generator](https://web.archive.org/web/20211009144446/https://min-25.hatenablog.com/entry/2018/03/19/235802) (It's for MCMF, which is literally DFS Ford-Fulkerson) that make the basic DFS Ford-Fulkerson runs in $F$ iterations, resulting in the complexity of $O(EF)$ ($E$ is number of edges, $F$ is the maximum flow of the network). </p> However, I just really want to use DFS Ford-Fulkerson, since you can do blocking flow on it, which grants some pretty bullsh*t performance boost on average (and it's trivial to implement). So... is there any close upper-bound on the complexity of DFS Ford-Fulkerson given these optimization? And if my optimization sucked, is there a test generator to blow it up? (Obviousl...

Full text and comments »

So you're a proud ratist? Here's why that's problematic (GRAYtist v1.1 out now) Just kidding, I'm one too. <img src="https://raw.githubusercontent.com/welcome-to-the-sunny-side/graytist/main/public/wordmark.png?v=2" alt="GRAYtist" style="float:right; width:400px; height:auto; margin-left:24px; margin-bottom:8px;"> Introducing [GRAYtist](https://chromewebstore.google.com/detail/graytist/bflnfmllkbhilogfbdilcckpkfganelc), a browser extension that helps you filter out low-quality content from Codeforces. You might say: "B-b-but [user:DNR,2026-06-12], such a tool would surely spam some API calls or use some heavyweight classifier models! My potato PC couldn't handle it, right?!". Wrong. We instead use the principle of ratism ($P(\text{low rating} \iff \text{low quality}) \approx 1$) to optimize away all of that nonsense. <div style="clear:both;"></div> ### Features ##### 1. Filter blogs by author rank and title. - Don't want to see blogs from grays because they're annoying, and blogs from nutellas because they make you feel inferior? GRAYtis...

Full text and comments »

[Tutorial] Quick start in solving problems on GPU + CPU Hello, codeforces! Do you have modern GPU and CPU and want to calculate something with them, but you don't know how to do it? Dont be upset, because you can read this blog and run your first GPU program on your AMD/Intel/NVIDIA GPU. I will describe an installation process and then we will dive into the C++ code. As an example, we will calculate a lot of mazes for [Bug Game](https://codeforces.me/blog/entry/115144) in parallel. For doing this in simple way, we will use library [Boost.Compute](https://www.boost.org/doc/libs/1_80_0/libs/compute/doc/html/index.html), which is a C++ STL-style wrapper over OpenCL library. [cut] ... Installation for MacOS (Apple M1 Pro) ================== As [user:Gosunov,2023-06-10] said, `OpenCL` and `g++` are pre-installed on `MacOS`. Everything that you have to do is install `Boost` with command `brew install boost`. After this you will be able to compile your program with command `g++ -std=c++17 -O3 -I/opt/homebrew/Cellar/boost/1.81.0_1/...

Full text and comments »

[cp-tricki] Welcome to cp-tricki! Hi there! Some long time ago, I and [user:SashaT9,2023-10-01] thought of an idea to create a collection of little tricks useful in competitive programming. We believe that such a website can both boost beginners, and be useful for experts as a reference website containing many example problems, which can be used, for example, to recall something, or, say, make an educational contest. It worth to say that there are already a bunch of blogs here with the same purpose. Our main feature is that we provide not just a list of tricks with their brief descriptions, but a detailed pages for each of them. This project was inspired by [tricki](https://www.tricki.org/) (currently dead) project. I already posted a test blog in order to demonstrate this format [here](https://codeforces.me/blog/entry/118108). We run a website based upon mkdocs (similar to cp-algorithms): https://cp-tricki.github.io/ Now we wrote five more or less simple (but unfortunately unpolished) pages: 1....

Full text and comments »

Shallowest Decomposition Tree Hi Codeforces! Today I want to tell you about a really cool and relatively unknown technique that is reminiscent of Centroid decomposition, but at the same time is also completely different. The most common and well known decomposition tree algorithm out there is the centroid decomposition algorithm (running in $O(n \log n)$). It is a standard algorithm that is commonly used to solve tons of different kind of divide and conquer problems on trees. However, it turns out that there exists another closely related decomposition tree algorithm, that is in a sense optimal, and can be implemented to run in $O(n)$ time in around $30$ lines of code. I have chosen to call it the _Shallowest Decomposition Tree_. This blog will be all about the shallowest decomposition tree. Something I want to remark on before we start is that I did not invent this. However, very few people know about it. So I decided to make this blog in order to teach people about this super cool and relatively unknown tec...

Full text and comments »

2026 Turkish Team Selection Problems Are Now Public! Merhaba everyone, After a high quality Turkish Div. 2 round, I'm happy to announce <a href="https://qoj.ac/category/770"> what is mentioned in the title</a>. Problems were written by <a class="rated-user user-red" href="/profile/tolbi" title="one who has written 2 basis construction problems">tolbi</a>, <a class="rated-user user-red" href="/profile/mychecksdead" title="one who accidentally scammed the EGOI scientific commitee for ideas">mychecksdead</a>, <a class="rated-user user-orange" href="/profile/erkam" title="one who got his problem idea from playing too much NSS">erkam</a>, <a class="rated-user user-orange" href="/profile/alperent" title="one who set up tight limits to ensure there was no lack of death threats in the contest room">AlperenT</a>, <a class="rated-user user-orange" href="/profile/weobe" title="one who finalized her statement 5 minutes before the start of the contest">Weobe</a> and <a class="rated-user user-unrated" href="/profile/erray" title="one who created...

Full text and comments »

Seeking Guidance on My Journey to Candidate Master !! ##### Hey Codeforces _:D_ I’m Harmit [user:H_a_r_M_i_t,2025-04-10] Khimani, and I’ve been steadily working on my competitive programming skills with the goal of reaching **Candidate Master**. I’ve made decent progress so far, but I know there’s still a long way to go, so I’d really appreciate any insights to help me level up faster. [Here’s my profile](https://codeforces.me/profile/H_a_r_M_i_t) if you’d like to take a look.   I’m looking for advice and would love feedback on a few things: - Speed and Accuracy Under Pressure: - How do you stay fast *and* accurate when time is tight and pressure kicks in? - Consistency in Contests: - How do you deal with rating ups and downs? What helped you stay steady? - Solving 1600–2000 Rated Problems: - What’s your approach to consistently crack problems in this ...

Full text and comments »

Way of problemsetter *The mystery of creating rounds unveiled by author of the 4 of them! The guide for preparing the round without horror-fiction and pastorals!* Thanks for [user:RodionGork,2014-06-13], who push me to writing this post and thanks for [user:lucAbalo,2014-06-13], who push [user:RodionGork,2014-06-13] to pushing me to writing this post. Also, many thanks to [user:RodionGork,2014-06-13] for the translation of this post to English. ##1. Inventing problems It's hard to advise anything on this point. There's no some standard way of creating problems &mdash; and if one have existed we'll have only "complicated algorithmic exercises" instead of "problems". People who participated in Russian Code Cup do know well the tasks of that sort. To cook good and interesting problem there should be some **idea!** which came to your mind (and later to minds of participants of the contest). Even the simplest problems of `A` level for the `Div2` should have some **idea**. So here is the war...

Full text and comments »

First RED-INDIAN AI Cheater? **NOTE: RED INDIAN IN THIS BLOG DOESN'T REFER TO THE NATIVE AMERICAN RACE. IN THE CONTEXT OF THIS BLOG, RED-INDIAN REFERS TO AN INDIAN WHO CHEATED HIS WAY TO RED, THAT IS, GRANDMASTER.** The person I am going to talk about here is none other than the most recent Indian Red Coder [user:arnabmanna,2025-04-27], who achieved this feat in the last Division 1, [contest:2097] after solving 4 problems and getting a Global Rank of 62 and India Rank 1 in the contest, giving him a whooping 119 rating boost. Such a jump not only made him red but also made him surpass 2 existing Red Coders [user:cerberus97,2025-04-27] and [user:invertedwinger,2025-04-27] in all time India leader board, making him the 5th highest rated coder in India and 582nd Globally. His progress seems very inspirational (for a certain group of people who follow his footsteps) since he achieved this feat with just 425 problems solved for all time, 74 of which happen to be 800s. Not just this, he went from Newbie to Gr...

Full text and comments »

Asymptotic expansion of harmonic numbers Hi everyone! In this blog, we will learn how to find the asymptotic expansion of $H_n = \sum\limits_{k=1}^n \frac{1}{k}$. #### Motivational example In the recent Meta Hacker Cup, [problem C](https://www.facebook.com/codingcompetitions/hacker-cup/2024/round-3/problems/C) of the round 3 reduced to finding the sum of $\frac{1}{a+bk}$ for given $a$ and $b$, over all $k \in [L, R)$. Now, [putting it into Wolfram](https://www.wolframalpha.com/input?i=sum+1%2F%28a%2Bbk%29+for+k+from+L+to+R-1), you can see that $$ \sum\limits_{k=L}^{R-1} \frac{1}{a+bk} = \frac{\psi(\frac{a}{b}+R)-\psi(\frac{a}{b}+L)}{b}, $$ where $\psi(z)$ is the [digamma function](https://en.wikipedia.org/wiki/Digamma_function). It is sufficient to solve this problem, as you can use boost or scipy to find it. But I feel like it's long time to satiate my curiosity on how it can be computed without knowing the digamma function in advance, and this blog is focused just on that. #### Harmonic numbers F...

Full text and comments »

The case of slow printf and scanf I know what is the problem ================== _Where we are going we need all the courage we can get_ When I first read the comment that our shiny new gcc 6.2 C++14 compiler has slow printf and scanf I immediately knew what I would do the next weekend. The new MinGW switched to custom implementation and I didn't need anything else to tell what the problem is. _Of course_, their format string parsing is too slow, they copy data around countless times, no file buffering is used and they have enough levels of indirection to make finding the real code feeling like a treasure hunt. What else could be the problem after all? Keeping this thoughts in mind I patiently waited for a weekend. When I finally opened MinGW source and searched for printf function, I was pleasantly surprised. The real implementation was only a couple hops away and soon enough I was looking at the heart of printf code. MinGW did try to conceal it a bit though by naming it mingw_pformat.c. But a ~100Kb source fi...

Full text and comments »

Busy but Master #2 Alternative name: "Road to Master when you don't have time". Episode 1: [link](https://codeforces.me/blog/entry/133458) Episode 2: September 2024 > See you in a month when I'm purple. Just unlucky?:) Two months of grinding Codeforces and I'm still not purple. It might seem like a failure, but if I look closely, I can see significant improvement compared to my performance in August. ### Practice results I feel like in the second half of the month, I've lost my ability to think properly, feel so dumb. Maybe it's just an illusion because I’m solving 2200-2400 rated problems instead of 1800, but still, it feels strange. I motivate myself by reminding that this must be a transitional stage, after which I'll see a big boost. University has started, and while it hasn’t taken up much of my time yet, I already spend ~full,2024-10-04 saturday for it. Plus, I spend a few hours a week chilling on social media, so compared to August, I probably have only about 70% of that ti...

Full text and comments »

jaiveer_017 is a cheater: 1491 to GM Perfection in One Contest? [user:jaiveer_017,2026-05-17] is clearly cheating. I was looking at the standings for today’s contest and found something absolutely ridiculous. A user currently rated 1491 (Specialist/Newbie level) somehow pulled off a Grandmaster-level performance, solving A, B, C1, C2, D, and E1 in less than two hours with barely any penalties. Solving D and E1 that fast with almost zero wrong answers is completely unnatural for a 1400-level player. But the real proof is in the source code. If you look closely, it's incredibly obvious that these solutions are either copy-pasted from a Telegram group or generated by AI mid-contest. ### Suspicious Evidence #1: The Shifting Template If you look at his submissions for A, B, C1, and C2, he uses a standard copy-paste template. But suddenly, in submission D, he randomly adds `#include <cstdint>` and `using ull = unsigned long long;` to his standard boilerplate. Why would you change your base template in the middle of a contest just to solve D? ...

Full text and comments »

Codeforces Lite: Enhancing Your Codeforces Experience ## **Codeforces Lite: Your Ultimate Companion for Codeforces** Looking for a **Dark Theme**, a **Code Runner**, or a smoother Codeforces experience? _**Codeforces Lite**_ is here — a Chrome extension already trusted by _**1700+ active users**_ to level up your competitive programming journey. ### **Key Features** - **Built-in Code Editor with Code Running and Submissions** Write code directly in our editor to save time and avoid switching tabs. Run your code _**Free of Cost**_ using integrated **Judge0** support. Prefer manual submissions? Just paste your code in the editor and submit _**FASTER THAN EVER BEFORE**_. - **Testcase Parsing** Testcases are auto-parsed into the editor and verdicts like _**AC, WA, TLE**_ are displayed right where you run the code. - **CF Dark Theme & Enhanced UI** Enjoy a clean and intuitive layout with a dark theme designed specifically for Codeforces. - **Customizable Templates** Store your own templates and benef...

Full text and comments »

January Report Hi everybody! <img src="http://assets.codeforces.com/images/tired.jpg" style="float:right;margin:0 1em 1em 1em;"/> Have you spent the January as efficiently as we have? Here's a small list of innovations the users can see: - Now not only the red coders can become coaches. The coders with a yellow handle are now accepted as well, provided that they have taken part in an official Codeforces round no earlier than 6 months ago and have taken part in at least 30 rated Codeforces contests. This innovation roughly doubles the number of potential coaches. So we invite you to check out your new role in the Gym! - Now it’s absolutely easy to add a Polygon contest to the Trainings. All you need to do is to copy the data from contest.xml from the Polygon system contest interface to contest.xml in the sandbox folder in the Gym (use FTP). - We’ve added protection against losing data in the Codeforces and Polygon projects’ forms. Now the textareas have a semitransparent ‘Drafts’ button i...

Full text and comments »

Meet the team — what it's like working at Grakn Labs My name is Vlad and I want to tell a story about how I ended up in the Grakn Labs. Before joining, I already worked for a couple of years in Kyiv and right after the graduation from Polytechnic Institute I moved to London to my current position, but it was a long path. I was never sure if this whole experience is useful in any way for the real world, as a software engineer, up until joining Grakn Labs. ... a lot of attention during the recruitment process is paid to problem solving skills. We have algorithmic problems as part of our interview process and now we want to find our next engineer here on Сodeforces ... **— When did you start competitive coding?** _When I was in the ~7-8 grade, my teacher showed us a website with problems and we got very competitive about it. After that, I was participating in math and informatics competitions up until the end of university_ <center> <img alt="graknworker" src="https://codeforces.me/predownloaded/c3/8a/c38add6c2e42afea7760c6945...

Full text and comments »

IQ is Pure COPE, Here's Why (Useful CP and Life Information #5?) Background ------------------ On CF I think you can tend to see people attributing good rating to IQ. This is often associated with hopelessness when people are stuck in gray/green. I think it's the perfect cope &mdash; it's the easiest thing to say that completely explains why you're 1100 and your friend is orange, and, bonus, if anyone tries to prove you wrong, you can just claim that they are privileged and intelligent. You may think I am 2000 and thus out of touch to be able to comment on this topic, which _may_ be true. However, I was also once a noob who got 28/1000 on a USACO silver round. And I was pretty overweight as a kid and sucked at sports and always thought of myself as not having the genetics to be an athlete, but [I now hold the squat, bench, and deadlift 18-and-under state records in my powerlifting federation](https://pa.liftingdatabase.com/records-default?recordtypeid=72&categoryid=603&weightclassid=5). (Probably not for long though lol) A lot of this i...

Full text and comments »

Schroeppel and Shamir's algorithm for subset sum or a cool idea to divide in 4 Hello, codeforces! Today I decided to write my first English blog on codeforces and present an interesting optimization for the subset sum problem. I found out about this technique in my high school and decided to explain it properly in the blog on codeforces. ##Problem statement Given an array $a$ of size $n$ and an integer $C$. Find a subset with distinct indices $\lbrace i_1, i_2, ..., i_k\rbrace$ so that $\sum_{j = 1}^{k} a_{i_j} = C$. ##Simple solution We can use here a meet-in-the-middle technique by dividing an array in 2 equal parts, putting all subset sums from first half in any structure like hash map, iterating through subsets in second half, calculating their $sum$ and trying to find $C - sum$ in our structure. This solution requires $O(n \cdot 2^{n/2})$ time and $O(2^{n/2})$ space if we greedily calculate sum of every subset with a loop. This can be improved down to $O(2^{n/2})$ time complexity if we find all our sums in another way: let's initialize a...

Full text and comments »

Editorial Codeforces Round #181 (Div.2) ### [problem:300A] In this problem you just need to implement following algorithm. Split input data into 3 vectors: first will contain negative numbers, second positive numbers, third zeroes. If size of first vector is even move one number from it to the third vector. If second vector is empty, then move two numbers from first vector to the second vector. This solution works in $O(n)$. [Аuthor's solution](http://pastebin.com/8aVCeRYx) ### [problem:300B] Input data represents a graph. If there is a connected component with at least $4$ vertexes, then answer is $-1$. Every connected component with $3$ vertexes is a complete team. Other teams are made from $1$ or $2$-vertex components. If amount of $2$-vertex components is greater than $1$-vertex answer is $-1$. Otherwise match $2$-vertex components with $1$-vertex. If there are some $1$-vertex components left then split them into groups of three. This algorithm works in $O(n+m)$. Also you could implement $O(n^4)$ solut...

Full text and comments »

Being rated is motivating. I don't know how did this happen, but my rating just jumped up 300 points. this is just after the DIV 2 round I gave some 2 hours ago. I didn't even solve any problem but just attempted one (the easiest of them) and in that too failed, due to a time limit exceeded error. I felt a bit down after that. But I was solving other problems. This has now boosted my coding engines. Now, I'll work twice as hard. You may see this as childish celebration and a noob getting excited over something which is so common but to me its a first. I really don't know any coding (in comparison to the average coder of Code forces ) but this is motivating. To be in the rankings, no matter how low; It shows that maybe I'm on the right track. So cheers to everyone whos working hard. Keep trying! Edit : I see some people in the comments have a bit of discomfort/unsurity about me getting a 360 jump. Maybe thier consternations are well placed and that made me wonder as well about things at ha...

Full text and comments »

C++ programming tutorial for beginners - part 1 Hello guys, welcome to my blog. I am DumDum and today we are going to do our first beginner tutorial on C++. Today we will start with the most beginner stuff, printing a line in C++. I will just copy and paste my template, wait a second... ~~~~~ //This is a comment, in a comment you can write whatever shid you want to write, the compiler won't complain #include <bits/stdc++.h> #include <limits> using namespace std; typedef long long int lli; #define prnt(a) cout << a << "\n" #define prnt2(a, b) cout << a << " " << b << "\n"; #define forp1(a) for (lli i = 0; i < a; i++) #define forp2(a, b) for (lli i = a; i < b; i++) #define forf(a, b, c) for (lli i = a; i < b; i += c) int main() { ios_base::sync_with_stdio(false); cin.tie(0); //YOu see, commputers are dumb, they don't understand human language. So we can't print //letter directly. But don't lose hope, their is a trick to make computers write our language //Let's start by p...

Full text and comments »

OCPC 2025 Summer season — wrap Hi everyone! <center><a href="https://ocpc.camp"><img src="https://ocpc.camp/images/ocpc_light.svg" height="100px"></a> <a href="https://icpc.foundation/"><img src="/predownloaded/43/08/43086529a28ccb5c9f2f2eec560ffa85e09c6881.svg" height="100px"></a></center> Just recently, the last mirror of the summer season of OCPC 2025 has concluded. I would like to take this moment to share some of the details and statistics on how it went with the community &#128522; [cut]<br> ### ICPC award Okay, this happened last chronologically, but it's probably most relevant, so I will start with it! <center> <a href="/predownloaded/43/df/43dfc0edaaa29db59b2b5ba6afbd4f033c8b535f.jpg"><img src="/predownloaded/43/df/43dfc0edaaa29db59b2b5ba6afbd4f033c8b535f.jpg" height="200px"></a> <a href="/predownloaded/b9/25/b9256e789a31ab2d1b3858bfa984ea1341dbc6c5.jpg"><img src="/predownloaded/b9/25/b9256e789a31ab2d1b3858bfa984ea1341dbc6c5.jpg" height="200px"></a> <a href="/predownloaded/c5/21/...

Full text and comments »

Codeforces contest results and Data Analysis Introduction ============ In the recent [user:tourist,2017-08-01]'s [blogpost](http://codeforces.me/blog/entry/53457) about a new strategy for some types of contests, there was [an interesting comment](http://codeforces.me/blog/entry/53457?#comment-375339) from [user:ftiasch,2017-08-01]. Apparently, some participants try to artificially boost their rating with the following strategy: solve problem C, if it goes fast and nice &mdash; submit and continue with the contest, otherwise don't submit anything. The assumption is, other problems in the contest are created by the same author, so if one performs good on problem C, they are more likely to perform good on other problems. In the comment thread, some people expressed doubts about this strategy. Reading their comments, I wondered: is there any point in discussing this with groundless arguments, when there is large volume of historical data of contest results, and we can just analyse them to understand if the strategy really wo...

Full text and comments »

10x Your CP Workflow: A Smooth Bash Script to Automate Parsing, Testing, Stressing and many more # 10x Your CP Workflow: A Smooth Bash Script to Automate Parsing, Testing, Stressing and many more Hello Codeforces fam, Over the past few months, I've been building and using a personal tool that **completely automates** my competitive programming workflow. From **parsing** problems to **checking solutions**, **stress testing**, **debugging**, and even **validating edge cases** — this tool has made my life 10x easier and smoother. And now, I’m excited to share it with you so that you can boost your productivity! ## What is this tool? It’s a **template-driven Bash script** designed to integrate with [Competitive Companion](https://github.com/jmerle/competitive-companion) and streamline everything — whether you're solving problems on **Codeforces, AtCoder, LeetCode, CodeChef**, or anywhere else. With just one command, you can: - Parse problems - Compile and run code - Switch between fast and debug modes - Stress test using a brute-force checker - Validate solutio...

Full text and comments »

Machine Learning Based Rating Predictor ### TLDR; I built an enterprise grade machine learning model to predict a user rating 6 months in the future. The model has mean absolute error of 65.15 (that is, $ E \left [ \left | \hat{x} - x \right | \right ] = 65.15 $). You can use it here: [https://fbrunodr.com/predict-codeforces-rating](https://fbrunodr.com/predict-codeforces-rating) ### Motivation: Check this thread: [https://codeforces.me/blog/entry/143626?#comment-1282206](https://codeforces.me/blog/entry/143626?#comment-1282206) Before going forward with this post I have to admit a pretty important thing: I did not do what I promised, as I did not build a foundational model on top of codeforces data. Reasons: 1. Takes to much time to train on my personal laptop (or money to rent gpus, which I am not willing to expend for a toy project). 2. I still almost went down the path of finetuning some feature extractor model (such as [this one](https://huggingface.co/Qwen/Qwen3-Embedding-0.6B)), but then I r...

Full text and comments »

levsivko (1508 rating points over 5 rounds) — the most brazen cheater of the season? Hello everyone! I'd like to bring the community and platform coordinators' attention to the account of user levsivko. In just 5 rounds, this participant shot to the top and achieved a rating of 1508, earning the status of "Specialist." At first glance, this seems like a great start for a talented newcomer. But if you open his contest history and look at the results tables, you'll see a classic, stupid, and completely shameless timeline of buying solutions from Telegram channels or using third-party help. This account's behavior completely contradicts any human logic for solving Olympiad problems. Let's examine all of his "achievements" step by step. **Evidence 1**. Educational Round 192 (Div. 2) &mdash; A classic "failure" of Problem B after failing A. Look at his results in Ed. 192: Problem A: Complete failure. The participant makes one attempt, receives a Wrong Answer (-1), and completely abandons the problem without even trying to fix it. Problem B: Successful submission (+) at ...

Full text and comments »

Motivation and Tips for Beginners! Are you a beginner in the exciting world of competitive programming? Are you just learning the basics and foundations of competitive programming? Are you ready to face the exciting yet challenging obstacles of competitive programming? Look no further! In this blog post, we shall dive into some valuable tips and tricks in order to help you get started on Codeforces. Whether you're aiming to improve your problem-solving skills or climb up the ranks and surpass Benq, these tips will set you on the right path for success. First and foremost, practice is key! Regular activity on Codeforces will improve your coding abilities significantly and expose you to unique algorithms and data structures. Start with simpler problems and gradually progress to more challenging ones. It is always important to understand the logic behind each solution and analyze the efficiency of your code. Next, familiarize yourself with common algorithms and data structures. Algorithms like binary search, sorting,...

Full text and comments »

South African Camp/Olympiad CMS Page Has it always been your lifelong dream to flex on the South African IOI team? Or maybe you've run out of problems to solve and want to solve more? Or maybe you just want a self-confidence boost from solving some somewhat easy problems? Well, you're in luck! South Africa will (hopefully) be using a CMS page open to the public for all future olympiads and camps. Simply go to https://saco-evaluator.tk/ and register for an account to solve our problems and/or flex on the South African IOI team by getting perfect scores in under an hour. Currently, there is only 1 "contest" with 6 problems running, but more will be made in due time. Solutions can only be submitted in C++ or Java right now, but Python3 support will probably be added later in the year. The scoreboard can be found at https://saco-evaluator.tk/rws/. (The scoreboard takes a while to update rankings, so don't be alarmed when your name doesn't show up on the scoreboard at first)

Full text and comments »

Wrong Answer on Pretest 2! Hello coders! I am sure all of you would have faced a "Wrong answer on pretest 2" at some point of time during your codeforces journey. It is a pretty frustrating feeling when you see your ranks (or ratings) dropping by a significant amount only due to some edge cases which you could not figure out during the contest time.. But the question that then comes to one's mind is : "Am I that dumb to not have figured it out then?.. and is there a way to make sure it doesn't happen a second time?" Well the answer is both a yes and no. So if you're someone struggling in the range between 1000-1400 (or even lower) then you might find this blog helpful! Actually, for the past few weeks this has been a major concern for me because either during contests or even while practicing I used to get a lot of "Wrong answer on test 2" and when I used to check for the case where I went wrong.. Most of the times it used to return something like "786th token doesn't match". Now how on Earth am I to figure...

Full text and comments »

The Virtual STEM School Saga _“When the system is broken, it’s not cheating to expose it. It’s duty.”_ &mdash; [user:Technoblade_never_dies,2025-06-26] Some of you probably remember that account [user:Technoblade_never_dies,2025-06-26] — the one who used to post blogs exposing cheating groups in online STEM stuff from Pakistan. Some people said it was just drama, but let’s be real, a lot of us knew he wasn’t making things up. He had screenshots and timestamps. For a while, it actually felt like someone was doing something real, not just pretending everything’s fine. Eventually the account just went quiet. Maybe he got tired of it, who knows. But the problems? they didn’t stop. they got worse. Now it’s not just shady circles in one place. Cheating is happening in Indian coaching centres, bangladesh Discords, telegram bots handing out live contest answers. Some groups literally exist just to boost ratings for uni apps. It’s not random anymore, it’s a whole system. Codeforces tries to catch what it can, but ...

Full text and comments »

We need leaderboard for negative contributions You see, having a negative contribution of 35 in codeforces is the best/worst thing I have ever done in my life. It's quite easy, every time I make a post about "how to become a red coder", my contributions pile up a little. Now, I am not content with just -35, I envision to take it much further. Currently, my aim is to become a LGM in negative contribution by surpassing the likes of great GMs and LGMs like [user:Sparky_Master_WCH1226,2022-12-01] and more. Now using the same formula results in diminishing returns as stated by a great economist, and getting to that height takes a lot of talent and effort. It's a shame that there's no achievement trophy for it. Good and evil are like two sides of the same coin and one shouldn't be looked down at than another. There should be a leaderboard for people like me, like maybe a "Bounty Hunter Board" or something, whenever you see 'em downvote 'em. A thing like that might boost our morale and efforts. By the way, how to become a red coder?

Full text and comments »

PSUT JCPC Qualification 2026 Tutorial Hints are mostly related, so you have to think and get some observations before understanding the next, make sure to understand the code before you submit it as well, "Most problems in life are due two reasons: act without thinking, think without acting" ~the,2026-08-02 poet,2026-08-01 Big thank you to the testers for improving this problem set beyond what it was, [user:the_seal,2026-08-02] , [user:Rayo,2026-08-02], [user:samsoom,2026-08-02], [user:lazlooz,2026-08-02] The contest exist in the group so make sure to join https://codeforces.me/group/ppRciMeJFg Some of these problems were inspired from other problems across the internet, "If I have seen further, it is by standing on the shoulders of giants." ~the,2026-08-02 poet [Problem A: MrMoon creates Lucky Number Seven Command](https://codeforces.me/group/ppRciMeJFg/contest/707652/problem/A) ================== ### 7 [Problem B: Moon creates seal backflips](https://codeforces.me/group/ppRciMeJFg/contest/707652/probl...

Full text and comments »

2018 CMUT BeihangU Contest, Editorial This editorial corresponds to [contest:102114] (stage 5), which was held on Aug 6th, 2018. Moreover, this problem set was also used as Jingzhe Tang Contest 1 in Petrozavodsk Winter Camp on Jan 30th, 2019. **This post is now finished**, in which I try to elaborate on notes, solutions and maybe some data generating. Alternatively, you can refer to [an old published material](https://drive.google.com/file/d/1CFumWbNLTUEywkVDNF-LKT3SzYd1r15L/view), though I think the old English version did not explain something clearly. --- [problem:102114A] This problem requires to calculate $s$-$t$ min cut between any two vertices on a weighted cactus graph having $n$ vertices, denoted by $\mathrm{flow}(s, t)$. You need to report $\sum_{s < t}{(s \oplus t \oplus \mathrm{flow}(s, t))}$. $n \leq 10^5$, $\sum{n} \leq 10^6$, weights are $\leq 10^9$. Try to find some features of this graph. <spoiler summary="solution"> By contradiction, we can prove that for an undirected graph, each ed...

Full text and comments »

Is It The Easy Way Lose Weight Fast By natural means How to loose weight fast --- instance training for busy men and women One of many reasons good reasons individuals do not work out is because they not have the time. Not every person can dedicate an hour each day to training. Exactly who don't understand is quick interval training workout workouts will get you greater final results than very long boring cardio can. Because of this article I will show you a fantastic training it can be done will trim your workout time by 50 percent yet still bring you utmost outcomes. Todd is really a Licensed Strength & Training Professional and produces for Guys Health and fitness, Mens Exercise, Highest Fitness, Muscle tissue and Health Hers, and Oxygen newspapers. Choose a diet which don't allow you to unfavorable because of the food items trying like card board. The most important challenge why individuals don't adhere to their exercises is: Do not add too much within the first two weeks. No other will make you extra adverse than the usual t...

Full text and comments »

000 — CSEC CPD KICKOFF Contest — (DIV 1) — Editorial [here](https://codeforces.me/contestInvitation/20d27b98a275636e2ceac807ccc79bfaefd490c2) is the contest link if u want to practice! ### [A. Garland](https://codeforces.me/gym/558367/problem/A) <spoiler summary = "Understand the Problem"> The problem asks us to turn on all four colored light bulbs in a garland, with the restriction that a bulb can only be switched if its color differs from the one switched previously. The goal is to find the minimum number of operations needed to turn all bulbs on, or determine if it’s impossible. </spoiler> <spoiler summary="Tutorial "> If u observe there are only Five possible configurations of the bulbs: 1. **AAAA**: All bulbs are of the same color. This makes it impossible to turn on all bulbs, as you cannot alternate between different colors. **Answer**: `-1`. 2. **AAAB**: Three bulbs are of the same color, and one bulb is different. In this case, it's impossible to turn all bulbs on in just 4 moves because at least on...

Full text and comments »

I built a better CPH — and a better Codeforces while I was at it (FREE) **TL;DR -** CPH and Competitive Companion already parse samples and submit for you. CPOS does that too &mdash; but that's the boring part. The point is everything *around* it: a **browser companion** that turns the Codeforces page itself into an IDE (write and run your solution right there, plus rating-change predictions, deep profile analytics, head-to-head compare, a focus mode, and a cleaner UI), a **VS Code extension** that &mdash; unlike CPH &mdash; also renders the problem statement next to your code, and a **terminal app** that actually tells you what to solve next and how to reach your next rating. Free, open-source, MIT, and 100% local &mdash; no accounts, no servers, no tracking. - **Source (MIT):** [github.com/Soham109/cpos](https://github.com/Soham109/cpos) - **Browser companion (the good part):** [Chrome / Edge / Brave](https://chromewebstore.google.com/detail/gjnbapmjonegeeamdeahcoojgokeogmm) (Firefox source install in the repo) - **VS Code extension:** [Marketplace]...

Full text and comments »

Invitation to Insomnia Qualifier 2022 Hello Codeforces! Computer Club, MNNIT Allahabad, India is glad to invite you to the annual programming competition of MNNIT, INSOMNIA, which is an ACM-ICPC style team programming contest of 2.5 hours duration held on Codeforces during its annual technical fest Avishkar. The team can consist up to 3 members. <b>Contest Details: </b> <ol> <li> Qualifiers: </li> <ul> <li> Start Time: Wednesday, November 9, 21:00 IST</li> <li> Duration: 2.5 hours</li> </ul> <li> Finals: </li> <ul> <li> Start Time: Sunday, November 13, 12:00 IST</li> <li> Duration: 2.5 hours</li> </ul> </ol> **Top 25** global teams, and **Top 25** teams from MNNIT (based on the result of Qualifiers) will qualify to the Finals. The prize distribution for global teams is mentioned below: <br>  Teams consisting of MNNIT students only will be eligible for a seperate prize pool. Register your team for the qualifiers here: https://forms.gle/BKsbKfoGzSxC394R8 <br>...

Full text and comments »

National Competitive Programming Challenge — Editorial Here is the link to the contest: [Link](https://codeforces.me/contestInvitation/6071a436743e70518e953453aadcf924bca069aa) ### [A. The Blacksmith](https://codeforces.me/gym/656191/problem/A) <spoiler summary="Credits"> - **Problem Idea**: [user:eulmelk,2025-12-12] - **Problem Setter**: [user:eulmelk,2025-12-12] </spoiler> <spoiler summary="Rate the Problem"> - **How good is this problem?** - Very Good - Good - Bad - Very Bad - **How hard is this problem?** - Very Easy - Easy - Hard - Very Hard </spoiler> #### Approach 1 <spoiler summary="Tutorial"> **Step 1**<br> $n \le 500$, so the total number of pairs is at most $\frac{n(n-1)}{2} = \frac{500 \cdot 499}{2} = 124750$. This is easily manageable within time limits. **Step 2**<br> By checking every possible pair, we never miss any valid combination. Therefore, keeping the maximum among all va...

Full text and comments »

Free Weight Loss Tips - You Have Anything You Pay Money For Even so the most really serious result's attained by individuals elements from the area of increased high blood pressure levels and cardiovascular acceleration. Coffein might activate vitality spending thereby greatly enhance tariff of weight. Having said that, coffein raises put on our cardiovascular and may induce climbing hypertension levels also. It is usually well-known that coffein boost our hunger. Additionally, there are supplements that be involved in extra fat oxidizing, like vitamin c and N supplements. But in fact polyvitaminic dietary supplements employed for this purpose might have some supplements that prevent breakdown of weight it may be PP vitamin and mineral or nicotinic plaque created by sugar. Have a look at evaluated simple ingredients in normal weight loss supplements which can be hazardous for your health and fitness or disfunctional as fat reduction treatment. Keep in mind supplements aren't strictly regulated by Food and drug administration, so its safe practi...

Full text and comments »

Learning Notes on the Hopcroft-Tarjan planarity algorithm (With C++ Implementation) **Section 1. Introduction** I have been struggling for weeks to understand the first linear-time planarity algorithm, the [Hopcroft-Tarjan algorithm (HT)](https://www.cs.princeton.edu/courses/archive/fall05/cos528/handouts/Efficient%20Planarity.pdf). Fortunately, I successfully understood it and implemented it. In fact it is very hard to find a verified HT implementation online. Some code, e.g., [LEDA](https://en.wikipedia.org/wiki/Library_of_Efficient_Data_types_and_Algorithms) is not open source. Other open-source planarity tests, e.g., [networkx (LR planarity)](https://github.com/networkx/networkx/blob/main/networkx/algorithms/planarity.py), [boost (Boyer-Myrvold)](https://www.boost.org/doc/libs/1_53_0/libs/graph/doc/boyer_myrvold.html), [Pigale (LR planarity)](https://pigale.sourceforge.net/), use other algorithms. There is [an implementation by Shawn Anderson](https://github.com/shawnwanderson/Hopcroft-Tarjan-Planarity-Testing), however it is partial. I would like to open a blo...

Full text and comments »