The score is the first thing you see and the least informative thing in the result. It is a single number over 180 questions, heavily influenced by which particular items came up, and it collapses four quite different situations into one figure.
What to read instead
- Where in each section your accuracy fell. A decline towards the end of sections is a pacing or stamina problem, not a knowledge one, and it has a completely different fix.
- Confident errors. The questions you were sure about and got wrong. These are the highest-value output of any mock.
- Time spent on questions you got wrong. If you spent a long time and still missed it, the decision to persist was the error.
- Blanks. Given +4/−1, every blank was probably a small expected loss. Count them and ask why each one happened.
A mock's value is in the twenty minutes after it, and most people spend those twenty minutes reacting to the number instead.
One mock is a small sample
180 questions across an entire curriculum means only a handful per topic. A swing of ten marks between two mocks can easily be which questions came up rather than any change in you. Read the trend across several; treat any single score as approximate.
Full-length mocks are for rehearsal, not diagnosis
A three-and-a-half hour paper mainly tests whether your concentration and pacing hold up over that distance. It is a poor instrument for finding individual weak topics, because the sample per topic is tiny. Use full mocks to rehearse the day, and use targeted sets to diagnose. Trying to do both with the same tool does neither job well.
Sit them under real conditions or do not bother
A mock taken in three sittings, with the phone nearby and a pause for tea, tests your knowledge and nothing else. The things a full mock is uniquely good for — stamina, pacing across five sections, the discipline of a locked section — all require the conditions. Without them you have taken a long practice set and called it a mock.
A review protocol worth following
- Confident errors first. Highest value, and you will have the energy for them only at the start.
- Then blanks. Ask why each one happened — no idea, no time, or misplaced caution. Three different problems.
- Then slow-and-wrong. Questions you spent a long time on and still missed. The error was usually the decision to persist.
- Skip fast-and-right entirely. Nothing to learn. Everyone reviews these because they are pleasant.
How often
Every one or two weeks in the last couple of months is plenty, and more often than weekly is usually counter-productive: a mock plus a proper review consumes a day, and you need days to act on what it found. A mock you do not review has cost you three and a half hours for a number.
If you have not got time to review a mock properly, you have not got time to sit one.
The review is the work
Sitting a mock and not reviewing it is close to wasting the three hours. The sitting produces the data; the review is where marks come from. If you do not have time to review a mock properly, you do not have time to sit it.