Yes. I found an available transcript of the video and summarized it simply. The video is “How To Become Red Coder? (codeforces.com)” by Errichto Algorithms, and Errichto’s own Codeforces post links this video as part of his competitive programming intro resources. (Codeforces)
Main idea of the video
To become a red coder on Codeforces, there is no magic book, secret routine, or perfect roadmap. The main thing is:
Solve many problems, at the right difficulty, learn from editorials, and make sure you actually understand the solutions.
Point-by-point simple summary
1. Red coder is a very high level
Errichto says becoming red on Codeforces is like becoming a grandmaster in chess. It is not something most people reach casually. It requires serious practice. (LinkedIn)
2. The most important tip: solve problems
His biggest advice is very direct: practice by solving problems.
Do not spend too much time asking, “Which book should I read first?” or “Which algorithm should I learn first?” Instead, start solving problems. You will learn algorithms naturally when you meet them in problems. (LinkedIn)
3. Choose problems that have editorials
He recommends solving problems that have editorials/problem analyses.
Why? Because when you cannot solve a problem, the editorial becomes your teacher. It tells you the idea, the algorithm, and what you were missing. (LinkedIn)
Example: suppose a problem needs DFS and you do not know DFS. Then you should search for a DFS tutorial, learn the basics, maybe solve 2–3 DFS problems, and then return to solving random problems. (LinkedIn)
4. Do not learn every algorithm before starting
A common beginner mistake is thinking:
“First I will learn all algorithms, then I will solve problems.”
Errichto disagrees. His view is: solve problems first, and when you encounter a new concept, learn that concept then. This keeps learning practical instead of theoretical. (LinkedIn)
5. Pick problems slightly above your level
He says solving problems that are too easy is not useful because you do not learn much.
Solving problems that are too hard is also not useful because you may just waste time and get frustrated.
The best problems are slightly outside your comfort zone: hard enough to teach you something, but not so hard that you have no chance. (LinkedIn)
6. Random problems are usually better than only topic-wise practice
He says it is not always good to solve 20 problems in a row from the same topic.
Why? Because if you already know “this is a graph problem” or “this is DP,” that itself is a hint. But in a real contest, you do not get the topic tag. You have to recognize the idea yourself.
So topic practice is okay sometimes, but random practice is important because it trains real contest thinking. (LinkedIn)
7. Red coders are not all the same
He says it is not true that every red coder:
has a computer science background,
read the same book,
follows the same routine,
has some magical habit before contests.
The common thing is that they solved a lot of problems. (LinkedIn)
8. You probably do not need 5000 problems
He says you do not necessarily need to solve 5000 problems to become red. For many people, a few hundred good problems may be enough, but it depends on the person.
If someone is already strong in math, puzzles, or logical thinking, they may improve faster. But still, practice is the core. (LinkedIn)
9. Do not stay stuck forever
If you are stuck for around 20–30 minutes, especially as a beginner or intermediate coder, he recommends looking at the editorial.
But he does not mean you should instantly copy the solution. Read a little, stop when you find a new hint or idea, then try again yourself. (LinkedIn)
10. Use editorials properly
There are three levels of using an editorial:
First, read only a small hint and try again.
Second, if still stuck, read more of the explanation.
Third, if needed, read the full solution — but then implement it yourself.
Implementation is important because sometimes you think you understood the idea, but when coding you discover missing details. (LinkedIn)
11. Understanding means you can solve similar problems later
Errichto gives a simple test: after solving a problem with help, if you later see a similar problem and can solve it after some thinking, then you probably understood it.
But if you see a similar problem and still have no idea, then you did not really understand the original solution deeply enough. (LinkedIn)
12. Upsolving is important
Errichto’s related practice notes also say you should upsolve problems you missed in contests. But do not try to upsolve every impossible problem. Usually, pick one or two problems that were just above your level. (GitHub)
The video’s advice in one simple routine
Here is the practical version:
Take a Codeforces problem slightly above your level.
Try seriously for 20–30 minutes.
If stuck, read a small part of the editorial.
Try again.
If needed, read the full editorial.
Code the solution yourself.
Check other people’s code if your implementation was messy.
Later, make sure you can solve a similar problem.
Repeat this many times.
Final takeaway
The message of the video is:
To become red, do not search for a perfect roadmap. Solve many good problems, learn from editorials, practice slightly above your level, and make sure you truly understand each problem you solve.
No comments:
Post a Comment