You generally want a balance between decisiveness and verification. Rechecking every step can destroy your pacing, but blindly trusting every first answer can turn small mistakes into avoidable losses.

Build checks into the process where the risk is highest: calculations, interpretation of a key data point, or a decision that depends on several conditions. For simpler choices, a quick reasonableness check is usually enough.

During practice, note how often a second check actually changes a wrong answer versus changes a correct one. That will tell you whether your current problem is carelessness or overthinking. If useful, this guide on McKinsey Solve guide gives more detail.

I was a chronic double checker. extra checking honestly created more mistakes sometimes.

rn I only recheck calcs / decisions that actually change the outcome.

practice helped me learn when first instinct was solid vs when I was just guessing.

the timer is part of the problem. being 'more accurate' doesn't help if you never finish the important stuff.

I literally track changed-right-to-wrong vs changed-wrong-to-right now. humbling but useful.