Past Papers Are the Best Revision Tool, and Most Students Waste Them
Ask any maths teacher what the highest-value revision activity is and you will get the same answer: past papers, marked properly. The reasoning is solid. A past paper is the only revision resource written by the same people, to the same specification, in the same style as the thing you will actually sit. Testing yourself on it exploits the best-evidenced finding in learning research, the testing effect: retrieving something from memory strengthens it far more than re-reading it ever does.
And yet most students extract a fraction of the value. The classic failure pattern: sit a paper cold, too early, under full timed conditions, score badly, skim the mark scheme, feel terrible, avoid papers for three weeks. Every step of that is fixable. Papers are not a single activity; they are a sequence of different activities, from open-book first passes to full dress rehearsals, and each stage has its own job.
This post covers that sequence: how to run a first pass, how to build up to timed conditions, how to mark honestly with the real mark scheme, why method marks change how you self-mark, and how to space papers across the run-in to exam day. Supply is not the problem: your board publishes years of past papers, and Exam Ladder holds 276 past papers with mark schemes across AQA, Edexcel, OCR and Eduqas. What follows is about using them.
- Past papers are written by the same people, to the same specification, as your real exam
- The testing effect: retrieving under exam-style conditions beats re-reading by a distance
- Papers are a sequence of activities, not one: first passes and dress rehearsals do different jobs
- The common failure is sitting papers cold and timed too early, then avoiding them after a bad score
The First Pass: Untimed and Open Book
Your first pass at a paper should be untimed and open book, and that is not going soft. It is matching the tool to the stage you are at. Early in revision, a timed paper mostly measures what you already knew, at the cost of ninety miserable minutes. An untimed, open-book pass turns the same paper into a teaching document.
Work through every question, and when you stall, do not skip. Look the method up, in your notes, a revision guide or a lesson, then complete the question yourself with the method in front of you. Mark the question with a symbol meaning 'needed help', because that flag is data. You are building two things at once: a map of exactly which topics need work, question by question, and the experience of finishing a paper, which quietly kills the fear that a full paper is beyond you.
Two rules keep the first pass honest. Looking up a method is allowed; copying a worked solution to the exact question is not, because transcription teaches nothing. And every question you needed help on goes into your mistake log, covered below, so the flag turns into an action. One or two papers used this way at the start of revision will do more for your confidence and your planning than five sat cold under the clock. They also feed directly into choosing what to revise next, because they show you which gaps are real.
- First passes are untimed and open book: the paper works as a teaching document
- When you stall, look up the method and finish the question yourself, then flag it
- Looking up a method is fine; copying a worked solution to that exact question is not
- Flagged questions become your topic priority list and your mistake log entries
Later Passes: Building Up to Real Conditions
As topics firm up, the papers change job, from diagnosis to rehearsal, and the conditions tighten in stages. Stage two is closed book but untimed: no notes, but no clock either, so you find out what you can genuinely retrieve without the pressure of time distorting the picture. Stage three is the full dress rehearsal: closed book, timed, no phone in the room, and ideally the right kind of paper for what you are rehearsing.
That last point matters more than students realise. Every board splits calculator from non-calculator papers, and the non-calculator skills, fraction arithmetic, written multiplication and division, exact values, need their own rehearsal. AQA and Edexcel sit three 90-minute papers of 80 marks each, with the non-calculator paper first. OCR papers are 100 marks each, and OCR puts its non-calculator paper in the middle of the three. Eduqas sits two 135-minute components, which makes stamina itself something to rehearse. Timing benchmarks differ too: a rough guide is a shade over a minute per mark on an AQA or Edexcel paper, and you should know the equivalent feel for your own board rather than discovering it in the exam hall.
Under timed conditions, practise the exam skills as well as the maths: skip a stuck question and return to it, show working even when rushed, and leave the last five minutes for checking. Those habits, along with what to do in the opening minutes, are exactly what the night before your exam should be protecting, not building from scratch.
- Tighten conditions in stages: open book, then closed book untimed, then fully timed
- Rehearse calculator and non-calculator papers separately: they are different skill sets
- Know your board's format: AQA/Edexcel 3 × 90 min at 80 marks; OCR 100-mark papers with the non-calculator paper in the middle; Eduqas two 135-min components
- Timed passes are for exam skills too: skipping and returning, showing working, checking time
Marking Honestly with the Real Mark Scheme
Marking is where most of a past paper's value lives, and where most of it is thrown away. The single rule: mark with the real mark scheme, published by the board for that exact paper, question by question, with the generosity of a stranger. Not the answers in the back of a revision guide, and not your own sense of 'that's basically right'.
Mark schemes have their own vocabulary, and learning it changes how you self-mark. M marks are method marks, awarded for a correct step even if the final answer is wrong. A marks are accuracy marks, usually dependent on the method that precedes them. B marks are independent marks for a correct result or statement on its own. When the scheme says 'oe' it means 'or equivalent', so earns what earns; 'ft' means 'follow through', where later marks can still be earned after an earlier slip; 'cao' means 'correct answer only', where nothing but the exact answer scores.
The discipline cuts both ways. Do not give yourself a mark you would have to argue for: if the scheme demands a specific step in the working and your page does not show it, the mark is gone, however sure you are that you 'did it in your head'. But do not dock marks the scheme awards either. Under-marking your own method marks makes papers feel more brutal than the real exam will be, and it hides the encouraging truth that you are often closer than your raw answer count suggests.
- Always mark with the board's real mark scheme for that exact paper
- Learn the vocabulary: M for method, A for accuracy, B for independent, plus oe, ft and cao
- Never award a mark you would have to argue for: unshown working does not score
- Never dock marks the scheme awards: under-marking hides real progress
Why Method Marks Change Everything About Self-Marking
Here is the mindset shift that separates students who improve through papers from students who just sit them. GCSE Maths is not marked answer-by-answer; it is marked mark-by-mark, and on longer questions most of the marks are for the journey, not the destination. A five-mark problem might carry three method marks and two accuracy marks. A student who sets up the right equation, rearranges it correctly and slips on the final arithmetic can score four out of five. A student who writes down a bare wrong answer scores zero for identical understanding.
If you self-mark answer-by-answer, right or wrong, you destroy this information. Two students can both 'get question 14 wrong' where one is a single sign error from full marks and the other could not start. Answer-only marking scores them identically, so it cannot tell you what to do next. Mark-by-mark marking tells the first student to slow down on arithmetic and the second to relearn the topic, which are completely different prescriptions.
Method marks should also change how you write, and how you total. In the exam, working is not something you show to keep teachers happy: it is how you get paid. Practising with mark-scheme marking trains you to write scaffolding an examiner can award. And when you total a paper, count marks earned, not questions right, then track that number against grade boundaries. Boundaries move between series and differ by tier and board, so treat any single year's boundaries as a rough gauge, but marks earned is the number the real exam cares about.
- Long questions carry more method marks than answer marks: the journey outscores the destination
- Right-or-wrong self-marking destroys the information that tells you what to fix
- A sign error and a blank page are different problems needing different fixes
- Total marks earned, not questions right, and read the total against grade boundaries as a rough gauge
The Mistake Log: Ten Minutes That Outvalue the Whole Paper
Every marked paper should end with ten minutes updating a mistake log. One page, one line per lost mark: the paper, the question, what went wrong in your own words, and the category. Three categories cover nearly everything. Careless: you knew it and slipped, a sign error, a misread question, a dropped unit. Method gap: you started but stalled or picked the wrong approach. Blank: you could not begin.
The log works because mistakes repeat. You do not make hundreds of different errors; you make your handful of favourites over and over, and until they are written down in one place you cannot see the pattern. Three papers in, the log starts saying things like 'loses marks on reverse percentages every single time' or 'forgets to give answers to the required accuracy'. Those are fixable, specific and worth several marks each, and no amount of general revision would have surfaced them as clearly. Our tour of common GCSE Maths mistakes is effectively a pre-written starter log of the errors examiners see most.
The log then drives two things. Between papers, method gaps and blanks become your revision list, attacked in sensible dependency order rather than at random. And re-reading the log is the perfect light-touch activity for the final days, when sitting new papers has stopped being worth it: five minutes with your own favourite errors is the cheapest mark-protection there is.
- One line per lost mark: paper, question, what happened, and a category
- Categorise as careless, method gap, or blank: each needs a different fix
- Patterns emerge within three papers, and patterns are fixable in a way vague weakness is not
- Re-reading the log is the ideal final-days activity when new papers no longer pay
Spacing Papers Across the Run-In
When you sit papers matters almost as much as how. The tempting pattern, ignoring papers until the final fortnight and then binging one a day, fails twice over: it wastes the diagnostic power of early papers, and it leaves no time to act on what late papers reveal. A paper that exposes a gap three days before the exam has told you something you can no longer use.
A better shape for the run-in, counting back from your real exam dates on the exam calendar: start with one or two open-book first passes early in revision, months out if you have them, purely for diagnosis. Through the middle weeks, one paper a week under progressively tighter conditions, with the days between spent acting on what the marking found. That gap between papers is the point, not dead time: a paper you never respond to is just a sad quiz. In the last three or four weeks, step up to two papers a week, fully timed, alternating calculator and non-calculator, and if your board sits three papers make sure each type gets a rehearsal.
Spacing the papers out also puts the spacing effect to work: coming back to mixed material after a gap forces the retrieval that makes it stick, which is the same principle behind spaced repetition. In the final two or three days, stop sitting full new papers. The marginal information is tiny, the cost to confidence and energy is real, and re-reading your mistake log covers the same ground for a fraction of the effort. If your whole run-in is compressed, the plan in how to revise for GCSE Maths in 2 weeks builds this rhythm into a tighter window.
- Do not binge papers in the final fortnight: early papers diagnose, and diagnosis needs time to act on
- Middle weeks: one paper a week, with the days between spent fixing what marking found
- Final weeks: two timed papers a week, alternating calculator and non-calculator
- Last two or three days: no new full papers, just your mistake log and light recall
Frequently asked questions
How many past papers should I do before GCSE Maths?
Quality beats quantity, but as a guide: one or two open-book diagnostic papers early in revision, one paper a week through the middle of your revision, and two a week in the final month, alternating calculator and non-calculator. Every paper must be marked with the real mark scheme and followed by action on what it found, otherwise adding more papers adds very little.
Should GCSE Maths past papers always be done in timed conditions?
No. Early papers are more useful untimed and open book, used as teaching documents to map your gaps. Tighten conditions in stages: closed book but untimed next, then fully timed dress rehearsals in the final weeks. Timed conditions test exam skills like pacing and question-skipping, which only matter once the underlying maths is mostly in place.
How do I mark my own GCSE Maths past paper?
Use the board's real mark scheme for that exact paper and mark mark-by-mark, not answer-by-answer. Award method marks where your working shows the step the scheme asks for, and be strict the other way: no marks for steps you did in your head. Learn the notation, M for method, A for accuracy, B for independent marks, oe for 'or equivalent' and ft for 'follow through'.
What is a mistake log and do I really need one?
It is a single page with one line per lost mark: where it happened, what went wrong, and whether it was a careless slip, a method gap or a blank. It works because your errors repeat, and the log makes the pattern visible within about three papers. It also gives you the perfect final-days revision activity: re-reading your own most likely mistakes.
Why did I fail a past paper on a topic I thought I knew?
Usually one of three reasons: the paper mixed topics without labelling them, which is a different skill from topic-by-topic practice; time pressure exposed methods that were understood but not yet automatic; or you lost method marks by not writing working the mark scheme could award. All three are trainable, and they are precisely what staged past-paper practice exists to train.