10 common mistakes in maths olympiad papers, and how to fix each one
By The Sophoz Curriculum Team
The short answer. Most marks lost in a class 4 to 8 olympiad paper are not lost to mathematics the child has not learned. They are lost to ten repeatable process failures: reading the question wrong, dropping units, applying a remembered formula to a case it does not cover, refusing to draw, and six more below. Each one has a specific fix and a drill that takes ten minutes.
This is written for the parent or tutor sitting beside the child with a marked paper, trying to work out what actually went wrong. Every mistake below includes what it looks like on the page, why it happens, the fix, and a drill.
One framing point before the list. In SOF's published cutoff data for IMO 2025-26, the Level 2 qualifying mark for class 7 was 38 out of 60. That is not a paper where a child needs to be brilliant. It is a paper where a child needs to stop leaking marks.
1. Answering a different question from the one asked
What it looks like. The working is correct, the arithmetic is correct, and the answer is wrong, because the question asked for the perimeter and the child found the area, or asked how many were left and the child found how many were taken.
Why it happens. Olympiad questions bury the actual demand in the last four words after two sentences of setup. A child who has been trained on school questions, where the demand comes first and is usually obvious, reads for the scenario and stops reading before the instruction.
The fix. Underline the question's final demand before starting any working. Not the numbers, the demand: the verb and the noun. "How many more." "The smallest such number." "In minutes."
The drill. Take ten olympiad questions. The child does not solve any of them. For each, they write down only what the question is asking for, in under six words. Mark that. This takes eight minutes and it is the highest-yield drill in this article, because it isolates a skill that solving practice never isolates.
2. Dropping or mixing units
What it looks like. An answer of 150 where the answer was 150 centimetres and the options were in metres. Or a speed question where minutes and hours were used in the same expression.
Why it happens. In school questions the units are usually consistent throughout, so conversion is not a step the child has learned to look for. Olympiad questions deliberately mix them, because mixing units is a cheap way to make a routine problem non-routine.
The fix. Write the unit next to every number in the working, not just the final answer. It slows the working down and it makes an inconsistency visible on the page rather than invisible in the head.
The drill. Give twelve short questions where the answer requires a conversion the question does not prompt. Length in cm asked in m, time in minutes asked in hours, money in paise asked in rupees. No solving beyond the conversion. The child is learning to notice, not to calculate.
3. Applying a remembered formula to a case it does not cover
What it looks like. Using the area formula for a rectangle on a shape that is not a rectangle. Using the average formula on a set where the groups are different sizes. Using a perimeter rule on a compound shape where two sides are internal.
Why it happens. This is the most instructive mistake on the list, because it is a symptom of the child having learned the formula as a procedure rather than as a statement about a situation. A child who knows that the area of a rectangle is length times breadth, but does not know why, has no way to tell whether this shape qualifies.
The fix. Before using any formula, the child says out loud what the formula requires. "Length times breadth works when all four corners are right angles and opposite sides are equal. Does this shape do that?" The check takes four seconds and it is the difference between a formula and a habit.
The drill. Give six shapes or situations, three of which the formula applies to and three of which it does not. The child's only job is to sort them into applies and does not apply, and to write one line saying why. Do not let them calculate anything. The skill is recognition.
4. Not drawing the diagram
What it looks like. A geometry or arrangement question attempted entirely in words and numbers, usually with a wrong answer and very little working.
Why it happens. Drawing feels like a delay when the clock is running, and many children have absorbed the idea that drawing is what you do when you cannot do it properly. It is also not usually rewarded in school marking, so it has never been practised as a step.
The fix. Treat drawing as compulsory for every question involving shapes, positions, orders, journeys or arrangements. Not an optional aid. A step.
The drill. Take eight questions and forbid answers entirely. The child's only output is a labelled diagram for each one. Mark the diagrams, not the answers. Draw a diagram is one of Polya's original heuristics from How to Solve It, and it is the one most often skipped precisely because it produces no marks of its own.
5. Abandoning a hard question instead of trying a smaller case
What it looks like. A blank space, or a guess, on a question that asks about a hundred terms, a ten-digit number, or a large arrangement.
Why it happens. The child reads "the 100th term" and correctly concludes that writing out a hundred terms is not viable, and then stops. Nobody has taught them that the move is to write out the first five.
The fix. Teach the explicit rule: when the number in the question is too big to handle, do the same question with a small number, then look at what happened. Solve a smaller case is a named heuristic, and it converts a large class of olympiad questions from impossible to routine.
The drill. Give five questions that all involve a large quantity. For each, the child works the case with n equal to 1, 2, 3 and 4, writes the four results in a row, and only then looks for the pattern. Do not ask for the final answer on the first pass.
6. Guessing the pattern from two terms
What it looks like. A sequence question where the first two gaps are both 3, so the child answers as though the rule is add 3, and the actual sequence was 1, 4, 7, 12, 19.
Why it happens. Two data points support infinitely many rules, and school sequence questions almost always use the simplest one, so the habit of checking never forms.
The fix. A rule is not a rule until it has been tested against a term that was not used to make it. Find the rule from the first three terms, then verify it on the fourth. If it fails, the rule was wrong, which is information rather than a setback.
The drill. Give ten sequences where four of them break the obvious pattern after the third term. The child must state the rule and then write "checked against term 4" with the verification. Mark the check, not the rule.
7. Losing the Achievers Section to the clock
What it looks like. The last five questions attempted in ninety seconds, or not attempted at all, with the rest of the paper complete and accurate.
Why it happens. The Achievers Section sits at the end of every SOF paper, and children work front to back. By the time they arrive, the time and the attention are both gone.
Why it is expensive. The Achievers Section is five questions carrying 25 % of total marks at every class level, worth two marks each for classes 1 to 4 and three each for classes 5 and above. Two children each get five questions wrong. The one who dropped five ordinary questions scores 55 out of 60. The one who dropped the Achievers questions scores 45. Seventeen percentage points apart for the same number of errors.
And there is a second reason, which almost nobody knows. SOF's published cutoff data states that where students tie on total marks, section-wise marks decide who advances, with weightage given in the order M4, M2, M3, M1. M4 is the Achievers Section. It is the first tie-breaker.
The fix. In practice papers, do the Achievers Section second, after a quick pass on the easiest twenty questions. Not last.
The drill. Once a week, a five-question Achievers-only paper, timed at fifteen minutes, done first thing rather than at the end of a session. The point is to practise meeting those questions with a fresh brain, because that is the condition you want on the day.
8. Doing all the practice one topic at a time
What it looks like. A child who scores well on the fractions worksheet, well on the ratio worksheet, and poorly on the paper.
Why it happens. Workbooks are organised by chapter, so practice is blocked by construction. Blocked practice removes the hardest step in an olympiad question, which is working out what kind of question this is. When every question on the page is a ratio question, the child never has to decide that it is one. The exam restores that step and they meet it for the first time on the day.
The evidence. In a randomised controlled trial with 787 seventh-graders across 54 classes over four months, students whose practice was interleaved scored 61 % on a test a month later against 38 % for students whose practice was blocked, an effect size of d = 0.83 (Rohrer, Dedrick, Hartwig and Cheung, 2020, Journal of Educational Psychology). In an earlier study, blocked practice produced 89 % accuracy during the practice session and 20 % a week later, while mixed practice produced 60 % during and 63 % after.
That last pair of numbers is the thing to hold onto. The method that looks better while the child is doing it is the one that fails on the paper.
The fix. Once a method is secure, stop grouping. Twelve mixed questions is worth more than thirty blocked ones.
The drill. Ten minutes a week, with a workbook and a pen. Pick four questions from chapter 3, four from chapter 7 and four from chapter 11, and write the page numbers on an index card. That card is a mixed practice set, it costs nothing, and on the evidence above it is worth more than any product you could buy.
9. Reading the worked solution instead of reconstructing it
What it looks like. A child who has been through fifty worked solutions and still cannot start question one unaided.
Why it happens. Reading a solution and following it produces a strong feeling of understanding. That feeling is recognition, and recognition is not what the exam tests.
The evidence. Students who read a passage then tested themselves recalled 56 % a week later; students who read it four times recalled 42 %. And the rereaders predicted they would do better (Roediger and Karpicke, 2006). In the largest review of study techniques, only practice testing and distributed practice earned a high utility rating, while rereading and highlighting both landed in the low tier (Dunlosky et al., 2013).
The fix. Read the solution once. Close the book. Write it out from scratch on blank paper. The gap between those two activities is the whole difference between recognising a method and being able to produce one.
The drill. Three worked solutions a session. For each: read, close, reconstruct, then compare. Mark the reconstruction, not the reading. This takes roughly three times as long per solution and covers a third as many, which is the correct trade.
10. Treating a wrong answer as a verdict rather than as information
What it looks like. The child gets a question wrong, the mood drops, the session ends badly, and the same error reappears three weeks later because nobody looked at what caused it.
Why it matters more than it sounds. Maths anxiety does not make a child worse at mathematics in any permanent sense. It consumes the working memory the child needs to do mathematics. Under a light memory load, high-anxious and low-anxious participants barely differ; under a heavy load, the high-anxious group makes significantly more errors (Ashcraft and Kirk, 2001).
And here is the finding that should change how a tutor handles a marked paper. In a study of 154 first- and second-graders, maths anxiety predicted lower achievement only among the children with higher working memory. Among lower-working-memory children the relationship was absent (Ramirez, Gunderson, Levine and Beilock, 2013). The children most damaged by pressure are the capable ones, which is exactly the population sitting olympiad papers.
The fix. Every wrong answer gets sorted into one of four buckets before anything else happens: misread the question, arithmetic slip, wrong method, or genuinely did not know. Those four need four different responses, and only the last one requires teaching. Most marked papers, when sorted this way, turn out to contain very little of the fourth kind.
The drill. After every practice paper, the child does the sorting themselves, out loud, before seeing the score. A tally of "four misreads, two slips, one wrong method, one unknown" is a useful document. A score of 41 out of 60 is not.
The one-week repair plan
If a marked paper has just come back and something needs to change this week, do these in order.
| Day | Drill | Time |
|---|---|---|
| Monday | Sort the marked paper into the four error buckets | 15 min |
| Tuesday | Ten questions, write the demand only, no solving (mistake 1) | 10 min |
| Wednesday | Eight questions, diagrams only, no answers (mistake 4) | 15 min |
| Thursday | Five-question Achievers set, done first, timed (mistake 7) | 15 min |
| Friday | Mixed set of twelve from three different chapters (mistake 8) | 20 min |
| Saturday | Three worked solutions: read, close, reconstruct (mistake 9) | 25 min |
Total under two hours. None of it requires buying anything.
Sophoz builds olympiad practice around the same shape: worked examples the child reconstructs rather than reads, mixed practice rather than chapter-blocked drills, and a schedule that works backwards from the exam date so earlier topics come round again. The method is not proprietary, though. It predates every platform selling it, this one included, and a tutor with a workbook and an index card can run all of it.
Frequently asked questions
What is the most common mistake in maths olympiad papers? Answering a different question from the one asked. Olympiad questions place the actual demand in the last few words after a long setup, and children trained on school questions read for the scenario and stop before the instruction. The fix is to underline the demand before starting any working.
Why does my child score well in practice but badly in the olympiad? Most often because practice was blocked by topic. Working through one chapter at a time removes the hardest step, which is identifying what kind of question this is. A randomised trial found interleaved practice produced 61 % on a delayed test against 38 % for blocked practice.
How important is the Achievers Section in SOF papers? It carries 25 % of total marks in five questions at every class level. It is also the first tie-breaker: SOF states that where students tie on total marks, section weightage is applied in the order M4, M2, M3, M1, and M4 is the Achievers Section.
What score does my child need to qualify for SOF Level 2? It varies by class and year. In SOF's published IMO 2025-26 data, the top-5 % cutoff was 38 out of 60 for class 7, 42 for classes 5 and 8, and 46 for classes 9 and 10. Note also that the top-5 % route is only one of three qualifying criteria.
Should my child draw diagrams in an olympiad exam? Yes, as a compulsory step rather than an optional aid, for any question involving shapes, positions, orders, journeys or arrangements. Drawing is skipped because it earns no marks of its own, which is exactly why it needs to be drilled separately from solving.
How do I fix careless mistakes in maths? Stop treating them as one category. Sort every wrong answer into misread the question, arithmetic slip, wrong method, or did not know. Those need four different responses, and sorting them takes the emotional weight out of a marked paper, which matters because anxiety consumes the working memory the child needs.
Cutoff and pattern figures verified 21 September 2026 against sofworld.org. Research citations are linked in full in what the research says about olympiad preparation.