Видео

My comment: Here is an idea to experiment on some future duels. Duel can be between two pairs of 2400+ rated users where for each round you choose 2 (up to 2400 codeforces difficulty rating) problems that are not solved by any of 4 duelists and then each pair split 2 problems so that each duelist solves only 1 problem without trying to solve the other one at all and then the main point is that each duelist can not write the solution for the problem player A solved but player A must collaborate with the other teammate (player B) in a way where the other teammate is using whiteboard and explaining all about the problem that this other teammate (player B) solved and giving instructions on what to write in order to solve the problem solved by the problem B. The same thing for problem solved by player A has to give instruction and explain all bits of the solution to player B. You could also have 3 streams separate where you have one main stream for your commentary and splitting between the other 2 streams where these 2 other streams each is for each pair of duelists with their commentary and screen share. Like the main points is in this solution explaining to the other teammate which could be interesting to see how top rated people come up with solutions. Maybe there doesn't even have to be any round but basically a set of problems that are not solved by any competitor like 5 problems or such and then each duelist marks the problem he is doing so the other one does not try to solve it and they have some time limit like 1h or 2h and are just trying to solve as many as possible. Also problems should be picked so they are not implementation heavy because that would be really hard I assume to code for the other person getting instruction, so maybe keep problems around 2000 cf rating difficulty or such so it keeps the stream quick paced and have something going at all time for majority of viewers to watch and learn something from the stream. This is just a rough idea with main focus on this solution explaining to the other teammate, I guess other people can give more suggestions and figure what could be done better and so. Anyway this ended up bit of a longer write up. Thanks for the streams. ================================================= Revised and more concise version of my comment by chatgpt: Here is an idea to experiment with for future duels. A duel could involve two pairs of 2400+ rated users. For each round, you could select two problems (with a maximum difficulty of 2400 Codeforces rating) that none of the four duelists have solved before. Each pair would then split the two problems so that each duelist solves only one problem without attempting the other. The twist is that no duelist is allowed to write the solution to the problem they solved. Instead, they must collaborate with their teammate. For example, player A explains the solution to the problem they solved to their teammate, player B, who codes the solution based on player A's explanation. The same applies in reverse-player B must explain and guide player A in coding the solution for the problem they solved. This format could be enhanced by having three separate streams: A main stream for your commentary, which alternates between the pairs. Two additional streams, one for each pair of duelists, with commentary and screen sharing. The main focus is on the solution explanation and collaboration between teammates, which could be fascinating to watch as it showcases how top-rated players approach problem-solving and communicate their ideas. You could also eliminate the concept of "rounds" and instead have a set of unsolved problems (e.g., 5 problems). Each duelist selects a problem to solve, ensuring no overlap, and the teams have a set time limit (e.g., 1 or 2 hours) to solve as many as possible. To keep the stream engaging, the problems should not be implementation-heavy, as this would make it difficult for one person to code based on another's instructions. Instead, problems around a 2000 Codeforces rating would likely be ideal, as they strike a balance between difficulty and fast pacing, keeping the stream interesting for viewers. This is just a rough idea focusing on the solution explanation aspect. I’m sure others can provide additional suggestions to improve this concept further. Thanks for the streams! ================================================= TLDR also by chatgpt lol chatgpt is the best: Proposed a duel format with two pairs of 2400+ rated users. Each duelist solves one problem (max 2400 rating) from a shared set of unsolved problems, then explains their solution to their teammate, who codes it based on the explanation. Focus is on collaboration and showcasing top players' problem-solving and communication skills. Suggested streams for commentary and screen sharing, a time limit (e.g., 1-2 hours), and selecting problems (~2000 rating) that aren't implementation-heavy to keep it engaging for viewers. Open to further suggestions to refine the idea.

First time seeing dominator

denominator to bhout neend mai lag raha hai😪

Cool ❤

Nice duel ...

keep them coming!

Wow! This is so cool. Always wanted to see how Dominater solves problems

Damn, the elimination of n colour states was so good. Thanks.. Please keep uploading videos of such hand-picked good questions.

Scrasse orz orz Dominator orz orz

Cp is about to be chess....1v1

This is amazing to see🎉❤

wow this can be the future of cp🎉 looking forward to tourist vs jiangly 1v1 someday

I guess vMix can help to stream both the participants screen at the same time

Congrats 👏🎉

Congrats to both the participants.

Congo.. shreyan for day 1 and best of luck for day 2

❤🎉

Rank of dhirubhai ambani institute of information technology

Very good explaination 👏👏

When two legends met together💀💥

when will an online mirror be available for practice?

Bhai can't we solve D question using DP? If yes, can you explain and give some hints?

Can you please do a discussion stream for ICPC Kanpur 2024 regionals.

Orgasm aa jata hai CP dekh ke🥵🥵🥵

Dominater069,

Oh dominator!

You teach really well but why you write what you are saying like a whole sentence it is not required !!!! <3

Very Clear Explanation of all questions bhaiya ❤

Why cant we just calculate min sum and max sum of sub array from entire array? Why we are taking two separate part.

in the 3rd question we should check for first 30 primes but by mistake u checked for first 30 numbers while proving it is greater than 1e18

Thumbnail has wrong problems

Which drawing tool are you using?

So damn good explanations

Bro, can u give some direction about xor getting 0 in 1 + log 50

Sir i am very new to codeforces , but have decent knowledge of dsa, but one thing.i don't understand is how much question should i aim for like for div 4 should i try upsolving all questions and same for div 2 , should i upsolve the first 3.. i mean my rating is just 1100 rn.. so can you please tell.. btw i really like your approaches and way of teaching

For problem C, for finding max sub arr sum and min subarray sum for left and right of the unique number why will kadane algorithm not work? Thanks in advance <3

is G hard for u?

What is the point of using INF and -INF in "E. Best Price" inside the impPoints vector

Sir very thankful to you for your post contest discussions . Really helps !!! . Please keep doing it .

@demoralizer welcome back

my only inspiration of competitive programming ,

Which os u use

thank you so much sir

can you share the software you use for these annotations

Lord Aryan with the solutions🙇‍♂🙇‍♂🙇‍♂🙇‍♂

can any one answer me? how to prove greedy approach? please

cant we use binary search for prpblem E

please add timestmaps for each problem

hi bro i am new to cp , how to build these concepts to solve problem like i was trying to do with python but unable to solve tried few codechef contest able to solve only first question butttt any advice you can give

my approach for B is as follows : we need to find OR of of range [L,R], lets says there is only 30 bits in the number at max ; then we will see if any number in the range of L,R can set the Result as 1 if not then in result it will be 0, next question is how we see that ? for lets say i th bit we want to see this; then we will set all the bit of lesser significance from ith to 0 and set i th bit as 1 in L if the this number exist the range then that bit will be set 1 in result ; now we need to prove if this number does not exist in the range then no number with this bit set will be in the range , so what we did ? we set the bit which was 0 in L as 1 and made all the bits with lesser significance as 0 hence resulting number will be bigger than the L hence if it doesnot exist in the range it will be bigger then the R, hence you may argue there may be a number in the range with i bit set which is smaller then number , that can only happen if we set bit of higher significance which was 1 as 0 but if we do that then it less than L hence no othen number can exist in the range ..... code is : void solve(long long n, long long m) { long long number = max(0ll, n - m); long long orignalNumber = max(0ll, n - m); long long base = max(0ll, n - m); long long j = 1; for (int i = 0; i <= (long long)(log2(n + m) + 1); i++) { long long reqNumber = ((base >> i) << i) | j; if ((reqNumber <= n + m) && (reqNumber >= (n - m))) { number = (number | j); } j = j << 1; } cout << number << endl; } time complexity is ig log(max(n+m)) cause in the result at max bits can only be set till the bits of significance present in n+m