Marks-weighted analysis of 405 questions across 3,809 marks from 17 IB Math AA HL exam sittings — with topic weightage, P1/P2/P3 splits, official grade boundaries, and Photon Academy's predictions for May 2026.
New · Live 6 Predicted Paper Sets for May 2026 — sets 1–2 free Open NowPercentage of total exam marks allocated to each topic across all 17 sittings, marks distributed among tags to avoid double-counting.
Some topics are overwhelmingly tested on one paper. This chart shows the marks split.
Of 56 questions tagged "Functions", only 23 are purely about functions (inverse, composite, transformations). The rest overlap with calculus, trig, or algebra — those marks belong to the other topic.
How topic weightage has shifted from 2021 to 2025 (% of marks per year).
Average marks per question — higher means the topic appears more often as a big Section B question.
The frequency data above tells you which topics matter. The interactive Topic Map shows you how they cluster together in past papers — so you can revise pairs that keep showing up in the same question, not topics in isolation.
Five years of data, six findings worth memorising before May 2026.
The single largest topic by marks. Almost always a standalone 15–20 mark Section B question. Reliable, predictable, high-weight — see our Vectors cheatsheet.
Functions marks share has grown 53% from 2021 to 2025. IB is testing functional reasoning (rational, composite, inverse) more and more.
But only 1.8% of P2. Complex numbers is almost exclusively a Paper 1 topic — and it dominates Q12 (the hardest question).
But only 1.9% of P1. Normal/binomial distribution and regression are almost entirely confined to the calculator paper.
Basic trigonometry marks share halved between 2021 and 2025. Less emphasis on trig equations and identities as standalone topics.
3 questions on compound interest and depreciation for the first time, cross-linked with Maclaurin series — a new IB approach to applied math.
HL only. 2 questions, 55 marks, 1 hour. The most unique and unpredictable paper.
| Year | Q1 Topic | Q1 | Q2 Topic | Q2 | Key Techniques |
|---|---|---|---|---|---|
| 2020 Spec | Inscribed/circumscribed polygons → approximate π | 30 | Chebyshev polynomials fn(x) = cos(n arccos x) | 25 | Trig identities, Maclaurin, limits, recurrence |
| 2021 May | fn(x) = xn(a−x)n — explore, differentiate, generalise | 31 | Roots of unity — product of distances | 24 | GDC exploration, complex numbers, induction proof |
| 2022 May | Polygonal numbers (triangular, square, pentagonal) | 27 | Cubics — tangent, complex roots | 28 | Number theory, induction, Argand diagram, inflection |
| 2024 May | Product rule special case: (fg)' = f'g' | 24 | Randomly generated quadratics — P(x-intercepts) | 31 | Differential equations, discriminant, counting, normal dist |
| 2025 May | Probability generating functions G(t) | 23 | Curvature k(x) = |f''|/(1+(f')²)^{3/2} | 32 | PGFs, derivatives, L'Hôpital, circles, implicit diff |
| 2025 Nov TZ1 | Normals to curves y = k²/x | 26 | Wolf population — Logistic vs Gompertz | 29 | Differential equations, Euler's method, model comparison |
| 2025 Nov TZ3 | Composed trig: cos((n+1)arccos x) | 26 | Maclaurin for arctan → approximate π | 29 | Induction, IBP, alternating series, GDC verify |
Q1 is almost always a pure mathematical investigation (functions, geometry, number theory). Q2 tends toward applied modelling (population, probability, physics-like contexts).
Every Paper 3 follows the same arc: use your calculator to explore specific cases, spot the pattern, then prove it algebraically. "Suggest" and "Hence show that" are signature command terms.
Q2 marks increased from 24–25 (2020–21) to 29–32 (2024–25). The applied/modelling question is becoming the dominant question on Paper 3.
How well candidates actually performed on each topic, session by session, as reported by IB's senior examiners. Green means well done; red means a topic where marks were routinely lost.
| Topic | M23 TZ1 | M23 TZ2 | N23 | M24 TZ1 | M24 TZ2 | N24 | M25 TZ1 | M25 TZ2 | M25 TZ3 | N25 TZ3 |
|---|---|---|---|---|---|---|---|---|---|---|
| Complex Numbers | Poor | Poor | Poor | Poor | Mixed | Mixed | Poor | Poor | Mixed | Mixed |
| Vectors | Poor | Poor | Poor | Poor | Good | Poor | Poor | Good | Good | Poor |
| Proof | Poor | Mixed | Poor | Poor | Poor | Poor | Poor | Poor | Poor | Poor |
| Differentiation | Poor | Mixed | Mixed | Good | Good | Good | Poor | Mixed | Mixed | — |
| Probability | Poor | Good | Mixed | Mixed | — | Mixed | Mixed | Mixed | Poor | Poor |
| Integration | Poor | Poor | Mixed | — | Poor | Poor | Mixed | Mixed | Poor | — |
| Differential Equations | Poor | Mixed | Mixed | Poor | Poor | — | Poor | — | Poor | Poor |
| Functions | Mixed | Mixed | Mixed | Good | — | Poor | Mixed | Poor | — | Poor |
| Sequences & Series | Good | Mixed | Good | Poor | Good | Mixed | Mixed | — | — | Good |
| Trigonometry | Mixed | — | — | Mixed | Poor | Poor | Good | Mixed | Poor | — |
| Statistics | Good | Poor | — | Poor | Poor | — | — | Poor | Mixed | Mixed |
| Polynomials | — | — | Mixed | Mixed | Mixed | Poor | — | Mixed | Mixed | Good |
| Exponents & Logarithms | Poor | Poor | Good | Good | — | — | — | — | Good | Mixed |
| Counting Principles | — | Poor | Poor | Poor | Poor | — | — | Poor | Poor | — |
| Binomial Theorem | Good | — | — | — | — | Good | Good | Poor | Poor | — |
Synthesised from 10 official IB examiner subject reports (M23–N25), across both time zones and the Paper 3 sittings.
The ten mistakes IB examiners flag most often across the reports — ordered by how frequently they recur. Fixing these is the fastest route to the marks most candidates leave on the table.
The single most-cited error. Candidates substitute or verify the given answer instead of building a valid forward argument — examiners award no marks for circular reasoning, and "verify" is not "show that".
Rounding intermediate values (instead of storing them on the GDC) corrupts final answers, and results are repeatedly given to 2 s.f. — or 6+ s.f. — rather than the required 3 s.f. or exact form.
Long analytic work is used where the calculator's solver or graph was intended (wasting time and inviting slips), yet the GDC is sometimes used when an exact value was required. Calculator-based working is also often poorly set out.
Base case values are not stated, the assumption of truth for n = k is vaguely or wrongly worded ("assume n = k", "let n = k+1"), and the concluding statement is missing — each costs a method mark.
Working is scattered across the page, cramped or illegible, and sketches have unruled, non-perpendicular or unlabelled axes with wrong asymptotic behaviour — leading to avoidable lost marks.
The GDC is repeatedly left in degree mode when radians were required (or vice versa) for trigonometry, producing incorrect answers throughout a question.
The constant is forgotten, or the initial condition is never used to find it — particularly in kinematics and differential equations — and it is often mistreated when a modulus is involved.
Candidates state the imaginary part without the i, fail to equate real and imaginary parts when solving equations, and draw lax Argand diagrams with roots at unequal moduli or angles.
Dividing an equation by cosθ without checking it is non-zero discards valid roots, while extra solutions outside the given domain (and missing obtuse solutions) are frequently offered.
Lines are written with "l =", an arrow, or a stray "l.." precursor instead of the required "r =", and equations of planes are given without the required integer coefficients.
Official IB grade boundaries as percentages. Source: published IB Grade Boundary documents.
| Session | G1 | G2 | G3 | G4 | G5 | G6 | G7 |
|---|---|---|---|---|---|---|---|
| N25 TZ3 | 0 | 12 | 24 | 35 | 49 | 63 | 75 |
| N25 TZ1 | 0 | 13 | 24 | 34 | 49 | 64 | 78 |
| M25 TZ3 | 0 | 13 | 24 | 35 | 49 | 64 | 78 |
| M25 TZ2 | 0 | 13 | 24 | 31 | 44 | 59 | 72 |
| M25 TZ1 | 0 | 12 | 24 | 36 | 46 | 60 | 72 |
| N24 TZ0 | 0 | 10 | 19 | 30 | 44 | 59 | 73 |
| M24 TZ2 | 0 | 10 | 19 | 30 | 44 | 59 | 74 |
| M24 TZ1 | 0 | 14 | 24 | 35 | 48 | 64 | 76 |
| N23 TZ2 | 0 | 13 | 24 | 34 | 49 | 62 | 74 |
| M23 TZ2 | 0 | 10 | 21 | 31 | 43 | 57 | 70 |
| M23 TZ1 | 0 | 13 | 24 | 34 | 49 | 64 | 75 |
| N22 TZ0 | 0 | 10 | 16 | 27 | 44 | 59 | 73 |
| M22 TZ2 | 0 | 10 | 18 | 28 | 39 | 52 | 65 |
| M22 TZ1 | 0 | 7 | 15 | 24 | 37 | 51 | 65 |
| M21 TZ2 | 0 | 4 | 12 | 24 | 38 | 53 | 69 |
| M21 TZ1 | 0 | 4 | 13 | 24 | 36 | 49 | 63 |
Source: Official IB Grade Boundary documents. Some 2023–2024 values estimated where data is incomplete — verify with your IB coordinator if needed.
HL Grade 7 boundary ranged from 63% (May 2021 TZ1, COVID) to 78% (Nov 2025 TZ1 & May 2025 TZ3). Average across all sessions: 72.5%.
HL Grade 5 requires roughly 45% of marks. Achievable by mastering Section A and getting entry marks on Section B.
Post-COVID boundaries stabilised at 72–78%. With the guided Nov 2024+ style, expect ~73% for May 2026 — the 5-year HL average.
Six full predicted paper sets written by Photon Academy in IB exam style. Each set has Paper 1 (no calc), Paper 2 (calc) and Paper 3 (extended investigation) with full markschemes. Sets 1 and 2 are free. Sets 3–6 unlock with Math Resources lifetime access.
Open the Predicted Papers viewer to read sets 1–2 in your browser, side-by-side with markschemes. Sets 3–6 unlock automatically once you have Lifetime Library access for AA HL.
Hand-built by an IBO-certified tutor. Print or learn online — no signup needed.