Past papers are a diagnostic, not a dress rehearsal
Doing thirty past papers and feeling more tired is not the same as getting better. The students who improve treat each paper as a list of things to fix, not a score to collect.
In the last month before exams, every student does past papers. The diligent ones do a lot of them — a stack of completed papers, each with a mark scrawled at the top. And many of them plateau anyway, because they're using the most useful revision tool in the world as a scoreboard instead of a microscope.
A past paper's value isn't the mark. It's the list of specific, named weaknesses it hands you for free.
The mark is the least interesting thing on the page
When a student tells me they "got 64%," I haven't learned anything I can act on. The questions that matter are: which marks did you lose, and why that kind of mistake? A 64 made of careless arithmetic slips is a completely different problem from a 64 made of one whole topic you can't do — and they need opposite fixes.
So the moment of real learning isn't doing the paper. It's the half hour afterwards, marking it honestly against the scheme and sorting every lost mark into a cause.
The routine I insist on
- Do the paper under real conditions — timed, no notes, in one sitting. A paper done in three relaxed chunks tells you nothing about exam day.
- Mark it yourself, strictly, against the official mark scheme. Learning the mark scheme's language is half of why past papers work.
- Log every lost mark by cause in three columns: careless, method gap, didn't know it. After three papers, the pattern is undeniable.
- Spend your next study session on the log, not another paper. The log tells you exactly where the marks are hiding. Chase those, then test again.
Ten papers you've analysed beat thirty you've merely completed. The paper finds the holes; filling them is a separate job, and it's the one that moves the grade.
This is the same idea as the blank-page test, scaled up to a whole exam: stop doing the comfortable thing that feels like revision, and start doing the uncomfortable thing that produces a to-do list. Then do the list.