The questions you got wrong quickly are the ones to study first

After a practice exam, most people review the questions they found hard. That is the wrong list.

The most valuable questions on your paper are the ones you answered quickly, did not flag, and got wrong. You were not uncertain. You did not slow down. You read it, knew the answer, chose it, and moved on, and it was wrong.

Why these matter more than the hard ones

A question you struggled with has already done its job. You know that topic is shaky, you flagged it, and it is on your list. Reviewing it is useful but it is not news.

A confident error is different in two ways.

You cannot find it by reviewing. Reviewing means going back over the material and asking whether you know it. On this question you will conclude that you do, because you already did, confidently, and were wrong. The belief that produced the wrong answer is the same belief that will assess it. Without a marked-up paper telling you, this error is invisible to you, and it stays invisible right up until it costs you marks on the real thing.

It is systematic, not random. You did not guess and get unlucky. You applied something you believe, and the thing you believe is wrong. Which means it is not one question. It is every question that touches that idea, on a paper of 160.

They are also the cheapest marks to recover

This is the part candidates find counterintuitive.

A topic you have never understood needs learning: hours of reading, then practice, then more practice. A confident error usually needs one correction. You do not need new material. You need to replace one wrong idea with the right one, and the rest of your knowledge around it is already in place.

Ten minutes on a confident error is often worth more than an afternoon of rereading a subject you find hard, because it converts a systematic loss into a systematic gain.

How to find them

You need three things recorded per question: your answer, whether it was right, and roughly how long you took. If you have all three, sort for fast and wrong, and read that list first.

Doing this on paper is possible but tedious, which is why almost nobody does it. Our reports surface it automatically, because it was one of the first things we wanted from our own tutoring and could never get at scale.

Once you have the list, for each one:

1. Work out what you believed. Not "I got it wrong", but the actual proposition in your head that produced that answer.
2. Find where that belief came from. Often it is a real rule applied one step too far, or a rule from a related area that does not govern here.
3. Write the correction down in one sentence, in your own words.
4. Find two more questions that touch the same idea and check the correction holds.

Step one is the one people skip and the one that matters. "I got it wrong, the answer is C" is not a correction, it is a fact about one question. "I thought the deadline ran from service rather than from filing" is a correction, and it fixes every question in that family.

The pattern worth watching for

If several confident errors turn out to share a source, you have found something significant. Usually it is a single misread rule doing damage across a whole subject, and the fix is one afternoon rather than one week.

That is the best outcome available from a practice exam, and it is not visible in a score.

The short version

• Review what you got wrong quickly and confidently, before what you found hard.
• Confident errors are invisible to ordinary review, because the belief that caused them is the one grading them.
• They are systematic, so they cost you more than one question.
• They are cheap to fix: one correction, not a subject.
• For each one, name the belief that produced it, not just the right answer.
• Watch for several that share a source. That is the highest-value finding on the paper.