IMO Problem Trends: A Ten-Year Analysis of Problem Types (2017–2026)

IMO Problem Trends: A Ten-Year Analysis of Problem Types (2017–2026)

Editorial overview

The International Mathematical Olympiad is often described through four familiar labels: algebra, combinatorics, geometry and number theory. That description is useful, but it can also give families the wrong idea. The IMO does not rotate through four predictable boxes, and a student cannot prepare effectively by memorising a fixed number of tricks from each chapter.

Looking across the ten completed editions from 2017 to 2026, the broad distribution remains remarkably balanced. Every year brings geometry, number theory, algebraic thinking and combinatorial reasoning in some form, but the most interesting change is how often those ideas are blended. A problem may look like a game but require an invariant; a geometry problem may be controlled by a transformation or an inequality; a sequence problem may ultimately depend on divisibility.

Our reading of the decade is therefore less about asking which topic is “most common” and more about asking what kind of mathematical maturity the IMO increasingly rewards. The answer is problem selection, structural recognition, proof control and the ability to move between representations. Students who prepare only by collecting named techniques are likely to find the paper unpredictable. Students who practise explaining why an idea works are preparing for the real competition.

This article treats each problem by its primary mathematical identity for the purpose of comparison. That is an editorial classification, not an official IMO statistic: many problems reasonably belong to two categories. The official IMO describes the competition as drawing from algebra, combinatorics, geometry and number theory, with six proof problems over two days. Read the official IMO overview.

What the IMO paper is designed to do

An IMO paper has six problems, each worth seven points, with three problems on each of two 4.5-hour days. The intended progression is often described as 1, 4, 2, 5, 3, 6: the first problem on each day is more accessible, while the final problem on each day is expected to separate the strongest contestants. This ordering is a useful guide, but it should not be treated as a promise that every student will find P1 easy or P6 impossible.

The format explains why topic distribution is only half the story. A student may recognise the area of a problem and still make no progress because the central difficulty is not content recall. IMO problems ask contestants to find a useful viewpoint, make a conjecture, test its limits and then write a complete proof. Partial progress can earn marks, but a convincing final solution must make the logical chain visible.

The four main families

Geometry

Geometry remains one of the most recognisable IMO families. The recent problems are not limited to a standard list of angle-chasing configurations. They may involve points inside triangles, iterative cuts, transformations, extremal configurations, or a carefully chosen invariant. The diagram is often an invitation to investigate rather than a picture that already contains the solution.

For preparation, this means that students need both a toolkit and the judgement to decide when not to use it. Similar triangles, cyclic quadrilaterals, inversion, ratios, directed angles and barycentric or coordinate ideas can all be valuable, but a long catalogue of lemmas is not a substitute for asking what the configuration is trying to preserve.

Number theory

Number theory is especially visible in problems involving divisibility, prime factors, greatest common divisors, least common multiples, integer sequences and Diophantine constraints. It is also a frequent source of hybrid problems: a sequence can become a number-theoretic object, while a game can be controlled by a modular or divisibility invariant.

The decade suggests that students should prepare number theory as a language of structure, not as a collection of congruence tricks. Useful habits include checking small cases, identifying what remains unchanged, looking for a minimal or maximal counterexample and separating necessary conditions from sufficient ones.

Algebra

At IMO level, algebra is broader than manipulating equations. It includes inequalities, functional equations, polynomials, sequences, substitutions and the study of how a condition behaves under repeated application. A functional equation may require a clever choice of inputs; an inequality may be solved by understanding equality cases before choosing a standard inequality.

The most transferable preparation is to learn how to interrogate a statement. What happens at zero, one or equal variables? What does symmetry allow? Can a complicated expression be normalised? Which part of the conclusion looks rigid enough to determine the form of a solution? These questions often matter more than knowing the name of a particular theorem.

Combinatorics and games

Combinatorics is the category most likely to be underestimated by students who think it means only counting. Recent IMO-style combinatorics includes colourings, arrangements, graph-like structures, extremal arguments, recursive processes and games. The hard part is often finding the right object to count or the right state to track.

Game problems are particularly useful for understanding the modern paper. They reward strategy, but the solution is rarely a description of what a clever player “would probably do”. A complete proof normally identifies a winning invariant, a strategy-stealing idea, a monotone quantity or a classification of positions.

A decade-level distribution: balanced topics, changing combinations

If each problem is assigned one primary label, the 2017–2026 set looks close to an even four-way distribution rather than a decade dominated by one subject. Small changes in the counts depend on how hybrid problems are classified, so the more reliable conclusion is qualitative:

This is why a simple league table of topics can be misleading. The official archive provides the problem sets, while resources such as MIT’s MathNet Explorer offer searchable topic metadata. Those tools are helpful for finding practice material, but metadata cannot fully capture the creative route a solution takes.

What changed in 2026?

The 2026 Shanghai IMO paper is a useful example of the decade’s mixed character. The official 2026 problem page presents:

• a number-theoretic process involving gcd and lcm on a blackboard;
• a geometry problem built around points inside triangles and a circumcentre;
• a game about marking and claiming pieces of a stick;
• an iterative triangle-cutting game;
• a functional equation on positive real numbers; and
• an integer sequence defined through a minimality condition and divisibility.

Using a primary label, this is approximately two number-theory problems, two geometry or geometric-game problems, one combinatorial game and one algebra problem. Using a hybrid label, the picture becomes even more interesting: the paper includes number theory plus process, geometry plus iteration, combinatorics plus strategy, and algebra plus functional structure.

The important point is not that 2026 suddenly changed the IMO syllabus. It did not. Rather, it makes the existing direction easy to see. The paper continues to reward students who can ask “what is invariant here?”, “what is the smallest useful example?” and “what would a complete proof need to establish?”

The real distribution is in the methods

Topic labels become more useful when translated into methods. Across the decade, students repeatedly need to practise:

1. Invariant thinking: identify a quantity, relation or property that survives a move or transformation.
2. Extremal thinking: choose a largest, smallest or earliest counterexample and exploit its forced structure.
3. Construction: build an object with exactly the properties the problem asks for rather than hoping a pattern continues.
4. Local-to-global reasoning: understand small cases, then prove why the observed mechanism works for all cases.
5. Equality and boundary analysis: find where an inequality or geometric condition becomes rigid.
6. Proof communication: turn an excellent idea into a solution that another mathematician can verify line by line.

This list is more useful for a student’s training plan than “complete geometry by March”. It also explains why a student can be strong in one contest and still struggle at IMO level: speed, factual knowledge and proof maturity are related, but they are not the same skill.

What this means for students in the UK and Ireland

The IMO is reached through national selection systems, so a student in the UK or Ireland normally builds towards it through several stages rather than entering the international contest directly. Our Olympiad Pathways Explained guide gives the broader picture, while the UKMT Competition Pathway Explained and Ireland Maths Competition Pathway Explained focus on the two local ecosystems.

For UK students, the progression from challenge papers towards written Olympiad work is explained in our guides to the UKMT Junior Mathematical Olympiad and the British Mathematical Olympiad Round 1. For Irish students, the Irish Mathematical Olympiad pathway article is the best starting point before looking at local competition details.

The decade analysis supports a practical recommendation: students should not wait until they feel “finished” with one topic before trying another. Rotate the training. Work on a geometry problem, then a number-theory problem, then a combinatorial game. Keep a written record of ideas that nearly worked. The goal is not to recognise the category instantly; it is to become comfortable when the category is unclear.

How parents should interpret the analysis

A topic distribution is not a prediction of whether a particular student will medal. It is a planning tool. If a child is already confident with algebra but avoids geometry, the answer is not to abandon algebra or to force a six-month geometry boot camp. A better plan is to identify the smallest regular practice that builds tolerance for unfamiliar diagrams and proof writing.

Parents should also be cautious about treating an IMO score as a simple ranking of mathematical potential. The contest measures a very specific combination of creativity, endurance, proof technique and performance under severe time constraints. It is a remarkable experience for students who enjoy that combination, but it is not the only valid route into mathematics, computer science, physics or engineering.

Our guide What Is Olympiad-Level Maths? explains this distinction in more detail. The short version is that Olympiad mathematics is not just “harder school maths”; it is a different kind of mathematical communication.

Key Takeaways

The 2017–2026 IMO problem set remains broadly balanced across geometry, number theory, algebra and combinatorics. The more important trend is the growth of hybrid problems: games with invariants, geometry with iteration, sequences with divisibility, and algebra with functional structure. The 2026 Shanghai paper illustrates this clearly. Students preparing for IMO should build all four foundations, practise moving between topics, and give equal attention to proof writing, structural thinking and problem selection. For UK and Ireland students, the local pathway guides on CompeteMap provide the practical route from school competitions to national Olympiad selection.

Final thoughts

The best way to read ten years of IMO problems is not to ask which year was “the hardest” or which topic is “most important”. Ask what the problems repeatedly demand from a solver. The answer is a willingness to explore, the discipline to discard a weak idea, and the precision to explain a strong one. That is the common thread running through all four areas, and it is the most durable lesson in the decade’s papers.