Frequency and Boundaries

Provenance: written 2026-09-28. Part 1 (topic frequency) reuses the two real, complete, question-by-question mark tables already built in forge-exam-technique-guide.md §1 (January 2025, WME01_01_2501_MS/ER, and June 2025, WME01_01_2506_MS) — no new primary-source mining was done for those two series; this document only re-tags their questions by spec sub-topic and counts. The cross-series trap citations in Part 1's second table are pulled directly from research/veridian/WME01-verified-facts.md §3/§4/§6 and the FORGE guide, already verified in prior passes. Part 2 (grade boundaries) required genuinely new lookups: this pass fetched six official Pearson grade-boundary PDFs directly from qualifications.pearson.com/content/dam/pdf/Support/Grade-boundaries/International-A-level/ (Jan 2023, Jun 2023, Jan 2024, Jun 2024, Jan 2025, Jun 2025 series) and read the WME01 row of each with pdftotext -layout, since no grade-boundary data existed anywhere in this course's WME01 research before this document.

#SIGNAL · WME01 Topic Frequency and Grade-Boundary Reference

#Pearson Edexcel IAL Mechanics 1 (WME01/01)

This is deliberately the smallest document in this paper's technique hub. forge-exam-technique-guide.md teaches the paper's shape and marking; codex-formula-and-method-bank.md is the formula masterlist. This document does one narrow job: how often each spec topic actually shows up, and what a real grade has actually needed. Nothing here is a worked method — for that, use FORGE and CODEX.

---

#1. Topic frequency and mark weight

#Basis: the two fully mark-mapped series (Jan 2025, Jun 2025)

FORGE §1 tags every question of both series against the spec numbering. Re-sorted by topic, every spec sub-topic except modelling (1.1) was examined in both series — a real, verified finding, not an estimate. Marks shown are each question's full mark total, not a spec-point split: most WME01 questions blend two or three spec areas inside one scenario, and the real mark schemes never sub-divide a question's marks by spec point, so a question's total is listed against every topic it draws on and should not be summed down the table.

Read this as: in two real, consecutive, fully-checked series, only modelling (1.1) failed to anchor its own question — every other spec topic did, every time. That is two data points, not seventeen; treat "every topic shows up almost every series" as a strong, narrowly-sourced pattern, not a guaranteed one.

#Cross-series signal beyond the two full mark-maps

The wider 17-series archive (Oct 2018–Jun 2025, per the facts bank) was never tagged topic-by-topic — that full weighting table was explicitly out of scope for the original research pass (facts bank §8) and stays out of scope here. What the archive does give, already verified in prior passes, is named-series confirmation of specific recurring traps — a narrower claim than "this topic is tested in N of 17 series":

• Momentum sign-convention trap (reporting a signed velocity where an unsigned speed was asked for): confirmed in Jan 2020, Oct 2022, and Jun 2024 — three named series — plus Jan 2025 independently reproduces it on a kinematics question. Four named series out of the 17 checked.
• Moments — choosing the pivot that eliminates an unknown reaction: confirmed in Jan 2023, Jun 2024, and Oct 2023 — three named series.
• Kinematics two-stage motion treated as one suvat equation: confirmed in Jan 2023 and Jun 2024 with the identical error description — two named series (Oct 2019 and Jan 2020 show related but distinct multi-stage/multi-particle errors, not counted here as the same trap).
• Connected-particles pulley-then-free-motion as the paper's final, heaviest question: confirmed in Jan 2025 and Jun 2025 only — two named series, explicitly flagged in FORGE as short of the 3+ series this course otherwise wants before calling something a standing convention.
• Bearing-to-component sin/cos swap, and wrong angle substituted into the cosine rule on a triangle-of-forces question: one named series each (Jun 2024) — not yet cross-series-confirmed.

None of the above is a percentage. Each number is a count of series this course has actually read and can name.

---

#2. Grade boundaries

Real, published Pearson data exists, and none of it was in this course's WME01 research before this pass. Fetched directly from qualifications.pearson.com/content/dam/pdf/Support/Grade-boundaries/International-A-level/ and read with pdftotext -layout this pass (2026-09-28) — six series, chosen to bracket the two fully mark-mapped series above plus show recent trend, not the complete 17-series history:

All figures are raw marks out of 75 — the number that actually matters when self-marking a real paper. UMS boundaries are not shown because they are fixed by definition (80/70/60/50/40 for a, b, c, d, e on every unit, every series) and carry no real information of their own.

Two real findings:

• M1 is always graded on Pearson's "International AS unit" table — lowercase a–e, no a*, in all six series checked, regardless of whether the unit is cashed into IAS Maths, IAL Maths, or IAL Further Maths. A unit-level A* does not exist for M1; A* only applies to a whole qualification's cash-in calculation from A2-unit UMS averages, not covered here.
• Boundaries moved by a lot once, then settled. Grade a jumped from 52 (Jan 2023) to 67 (Jun 2023) — a 15-mark swing in one series — then held at 64–65 from Jan 2024 through Jun 2025, including both series with a full mark table above. Don't read that four-series stability as a rule the next series must follow: Pearson sets each series' boundaries after the exam, from that exam's own difficulty, and only six of the archive's 17 series were checked here.

---

#3. How to use this

• Use Part 1 to sanity-check revision coverage, not to justify skipping anything: in the only two series checked question-by-question, every spec topic except modelling anchored real marks. Don't deprioritise a topic on the theory that it's "less likely" — the real record doesn't support that theory for this paper.
• Use Part 2 to translate a self-marked mock into roughly what grade that score needed in a real recent series — a rough calibration, not a guarantee. The next real series can set a different boundary, as Jan→Jun 2023 shows it already has.
• This document replaces neither FORGE (timing, mark-coding, self-mark checklist) nor CODEX (formula masterlist, worked-method skeletons). Use those for technique; use this one only for "how often" and "what did a grade need."

---

#Genuine gaps flagged, not papered over

• Part 1's "seen in N of 2" basis is real but narrow — not a claim about all 17 archive series, since no full per-question spec-tag was ever built across that wider archive (facts bank §8 flags this out of scope, and it stays out of scope here).
• Part 2's boundary table covers 6 of the archive's 17 series, not the complete published history back to Oct 2018 — earlier and later series were not checked and shouldn't be assumed to resemble the six shown, given how much Jan 2023 differs from its own neighbour.
• Marks in the Part 1 table are whole-question totals, not a spec-point-exact split — no mark scheme reviewed for this course sub-divides a question's marks that finely, so this document doesn't invent a split either.