What do you prefer or you believe to be effective more?
1.Spending more and more time on a single problem (above your comfort level) and try until finding a complete solution and then write notes about the observations you got and the subtask you solved
2.If you don’t get it, after 20-30 minutes,go to editorial and solve. And prefer not to write the observations you got, simply move to the next one.
Which one?
Read problem then read editorial.
lol
Read editorial only :D
And then guess the problem from it.
.
It's funny that Benq seriously said that during practice he reads the editorial first, and then reads the problem. You can find this in his github "/General Resources"
Similarly in contests, he submits the solution first and then reads the problem.