Every problem on this ladder is graded by an automatic judge. You never talk to it directly. It does three mechanical things, the same way every time:
- It feeds your program some text on standard input (stdin).
- Your program runs.
- It compares what your program printed to standard output (stdout) against a stored correct answer, character for character.
If your output matches, the test is Accepted (AC). If it does not, you get Wrong Answer (WA) — even if you were "close". The judge is not reading your intentions; it is comparing two blocks of text.
That is the whole game. A problem is not "write a clever algorithm" — it is "read the input the problem describes, compute the answer, and print it in exactly the format the problem asks for." Get comfortable with that loop and the rest is just widening what compute can do.
One practical note: while you are learning, the ladder can run in a mock mode that marks a test AC without truly executing your code, just so the interface works with no servers. Any result produced that way is badged as a mock so you always know whether a real judge ran your program.