Time allocation and the psychology of a live contest
Learn
Rating is lost to the clock and to tilt at least as often as to not knowing an algorithm. Two failure modes do most of the damage: sinking forty minutes into one problem out of sunk-cost stubbornness, and unraveling after a wrong submission. Both are manageable, and managing them is a skill you practice like any other.
The allocation rule of thumb: give each problem a soft budget and respect it. If you have read a problem and have no idea after about ten minutes of genuine thought, switch — your subconscious keeps chewing on it while you bank points elsewhere, and problems often crack on the second look. Having a buggy idea is different from having no idea: a concrete-but-broken approach is worth pushing a little longer. And never, ever leave an easy problem unread because you got stuck on a hard one.
A Wrong Answer is information, not a judgment. The disciplined response is a fixed checklist, in order: re-read the statement (a large fraction of WAs are simple misreads), check the obvious edge cases (n = 1, integer overflow, empty input, all-equal), and only then stress test. What you must not do is resubmit the same logic hoping the judge changes its mind — that just burns penalty time and feeds the tilt.
Forty minutes into a contest you have spent all of it on problem B with no working idea, and problems C and D are still unread. What is the best move?
Tilt is the silent rating-killer: one wrong submission rattles you, the next decision is worse, errors compound. The counter is mechanical — after a WA, take thirty seconds, breathe, and if you are stuck, switch problems to reset your head. Treat your own composure as part of the toolkit, because over a season it moves your rating more than any single algorithm will.