Most IB Math students don’t fail to revise—they fail to revise in the right order. Re-reading notes, drilling isolated topics with no bridge to exam conditions, and jumping straight to full papers can all feel like progress. The shared problem is that none of them trains the complete skill set the exam demands, and reaching for papers too early produces a score you genuinely cannot interpret. Did you drop marks because you didn’t know the method, couldn’t recognize which method applied, or knew both and made a slip at execution? That kind of opacity isn’t a motivation problem. It’s a sequencing problem.
The failure isn’t any single activity—it’s conflating what each stage is built to do. Running full papers before technique-identification is solid generates feedback with nowhere to go: you know you dropped marks, but not which gap caused them. Staying inside topic-labeled drills too long builds recall and execution in a condition the exam never replicates. A three-stage sequence—applicable across every IB Math pathway, whether AA or AI, SL or HL—keeps those cognitive demands separate so each stage trains what the next one depends on.
- Step 0—Lock your course constraints (once, for ten minutes): note your pathway (AA/AI, SL/HL) and, for each paper you will sit, calculator allowed (Y/N), question format (shorter vs. longer multi-part), and time limit.
- Stage 1—Close topic gaps (most days): work one micro-topic at a time using topic-labeled sets. Tag each miss as (A) method not recalled or (B) execution error. Gate: across two separate sittings, you choose the correct method without hesitation; remaining errors are execution detail.
- Stage 2—Mixed-topic identification (several times per week): build unlabeled sets mixing across repaired topics. Gate: you select the correct technique on first read most of the time; errors are slips, not repeated wrong-method choices.
- Stage 3—Course-matched simulations (one to two times per week, once gated): replicate timing, calculator rules, and formula booklet access. Route misses: wrong technique → Stage 2 mixed sets; wrong execution → Stage 1 drill. If you’re repeatedly routing large portions back to Stage 1, reduce full papers and rebuild.
Table of Contents
Stage 1 — Isolating and Closing Topic Gaps
Stage 1 is deliberate isolation. Every question carries its topic label in advance, so the method is announced before you read the problem—a scaffold that lets you focus entirely on mastering the technique rather than identifying it. Work one micro-topic at a time, tag each miss as a recall failure (method unknown) or an execution failure (method known, applied incorrectly), and advance when you can select the correct method without hesitation across two separate sittings.
Deciding where to start is itself part of the work. For each weak area, weigh three signals: weakness severity (how often you get stuck or choose the wrong steps), exam exposure drawn from observed past-paper patterns, and prerequisite leverage (whether fixing this topic unlocks others). Start with topics that are both high-weakness and high-exposure—in AA HL, Calculus accounts for roughly 29% of marks across 1,730 questions from 57 past papers (2021–2026), a practical signal worth taking seriously, though these are historical distributions rather than IBO guarantees. If a topic carries clearly higher prerequisite leverage, address it first even when its exposure signal is slightly lower.
Stage 2 — Mixed-Topic Practice and Technique Identification
Stage 1’s scaffold—knowing the topic before you start—is also its limit. A real IB Math paper never announces which method applies; recognizing the problem type and reaching for the correct technique family is the first cognitive task on every question. Topic-labeled practice never trains that recognition because the label does the work for you. Stage 2 removes it. Earlier classroom research with 140 grade 7 students confirmed the mechanism clearly. Over nine weeks of lessons followed by an unannounced test two weeks later, interleaved mixed problem sets produced mean scores of roughly 72% compared with 38% for blocked sets (effect size around d = 1.05)—because students had to match each problem type to its solution strategy without being told which strategy to use. Blocked practice, by contrast, is comfortable precisely because it pre-solves the hardest step.
The strongest current evidence for treating Stage 2 as a distinct phase comes from a quasi-experimental study in a large undergraduate calculus course (N = 585), published in npj Science of Learning in 2026. Interleaved, mixed-topic homework improved final exam performance for all students and disproportionately benefited those with lower initial performance—the population for whom technique gaps are most costly at exam time. Mixed problem sets don’t merely approximate exam conditions more closely; they train the discrimination skill that exam performance actually measures.
For IB Math revision, Stage 2 means building problem sets where the topic is not stated, mixing questions across areas already repaired in Stage 1. The cognitive work is identification first: reading the problem and selecting the correct method before any calculation begins, not executing a pre-announced technique. The gate criterion is behavioral—when you’re selecting the correct technique on first read most of the time, with remaining errors concentrated in working steps rather than method choice, the identification skill is in place. What it hasn’t tested is the sustained pressure of a full sitting: multi-part questions that compound across a timed paper, and the specific constraint bundle—calculator rules, formula booklet, timing—of each paper you’ll actually face. That’s precisely what Stage 3 is built to expose.
Stage 3 — Full-Length Simulation Matched to Your Course
Stage 3 is only productive once Stages 1 and 2 are genuinely complete. Running full papers before topic gaps and technique-identification skill are solid generates data you have no framework to act on: you see a mark but can’t determine whether the loss came from a method you don’t know, a method you couldn’t identify, or a step you executed incorrectly. Completing papers without a diagnostic routine is revision theater—the effort is real, the learning isn’t.
Course-specific matching is non-negotiable once you reach this stage. In AA HL, students sit three papers: Paper 1 is no-calculator and accounts for roughly 40% of marks in past-paper distributions; Paper 2 is calculator-allowed at roughly 40%; Paper 3 is a longer calculator paper with a different question architecture at roughly 20%. That three-paper structure needs its own simulation strategy—simulating the wrong paper gives systematically misleading readiness data. Students ready for this stage should locate IB Math practice exams matched precisely to their course and level, since paper count, calculator permissions, and question architecture differ enough between pathways that a mismatched paper can’t serve as a reliable test. These paper-level figures come from the AA HL past-paper archive referenced earlier and represent historical patterns, not guaranteed future weightings.
The obstacle is that simply rereading solutions on a marked paper doesn’t reveal whether each lost mark came from wrong method selection, weak execution, misreading a command term, or time pressure—and without that distinction, you can’t determine whether to return to Stage 1 or Stage 2.
- Tag each lost-mark item by primary cause—Technique choice (wrong method or didn’t know what to do), Execution (right method but wrong algebra, notation, graphing, or rounding), Interpretation (misread the command term or didn’t answer what was asked), or Time/strategy (ran out of time or got stuck too long).
- Write a one-line fix—the single smallest action to take before the next simulation.
- Route the fix—Technique choice: back to Stage 2 mixed sets for the same cluster, topic unlabeled. Execution: back to Stage 1 drill on the same micro-skill until the steps are automatic. Interpretation or time/strategy: add one explicit constraint to the next simulation—for example, re-read command terms before writing the final line, or apply a two-minute move-on rule.
- Weekly review—count your cause tags. A majority of technique-choice tags means more Stage 2 work next; a majority of execution tags means more Stage 1.
Tagging each lost mark by cause and routing it back under matched paper constraints is what turns practice papers into genuine diagnostic feedback rather than corrected scripts. Held consistently, this loop keeps Stage 3 in its proper role: a diagnostic phase that continuously feeds repair in Stage 1 and Stage 2, not an accumulation of completed papers you’ve reviewed and filed.
Making the Three-Stage IB Math Sequence Your Default Revision Plan
The sequence pays off at its gates, not just in its stages. Moving on too early means entering the next phase with gaps still open; staying in Stage 1 too long creates false security from topic-labeled questions that real exams never replicate. Each gate is a check against your own optimism.
The underlying logic is not tied to any particular syllabus year. As IB course content evolves, the sequence stays applicable because it mirrors the cognitive demands of the exam at each level of preparation. A student who respects the gates enters the exam room knowing what each stage trained—and when marks are dropped, they can trace each one back to a cause they already know how to address. A student who treats the sequence as a checklist of tasks to complete arrives with a folder of reviewed papers and a score that still defies explanation.