karthikeya_cp's blog

By karthikeya_cp, history, 5 months ago, In English

Most contestants think rating is a function of the number of problems solved.

It is not.

A contestant who solves 300 carefully selected problems often improves faster than someone who solves 1500 random ones.

The difference is not volume.

The difference is feedback.

A beginner solves a problem and moves on.

A strong contestant solves a problem and asks:

• Why did I miss the observation? • What alternative solutions exist? • What pattern was hidden? • How can I recognize it faster next time?

The solution is not the asset.

The observation is.

Every rating jump is usually the result of seeing a pattern you previously missed.

1200 → Learn implementation.

1400 → Learn standard algorithms.

1600 → Learn pattern recognition.

1900 → Learn problem decomposition.

2100+ → Learn observation generation.

At higher ratings, coding speed matters less than identifying the correct model.

The code is often twenty lines.

The idea is everything.

Most contestants practice coding.

Few practice thinking.

That is where ratings are gained.

Full text and comments »

  • Vote: I like it
  • -23
  • Vote: I do not like it