Pure Mathematics 1

Error Pattern Tracker

Provenance: compiled 2026-09-29. Every error code below is drawn from research/veridian/WMA11-verified-facts.md §4 (verified per-topic examiner-report material) and cross-checked against this same folder's forge-exam-technique-guide.md §4 (common-error digest) and signal-frequency-and-boundaries.md (frequency counts) so the two documents stay consistent rather than drifting apart. No new primary-source extraction was done for this document — it is a repackaging of already-verified findings into a fillable tracking format, not a new research pass. Where the facts bank states an exact multi-series count ("confirmed across two independent series," etc.) that wording is reproduced here; nowhere does this document state a likelihood percentage or a grade boundary, because none exists in the verified sources. Not affiliated with or endorsed by Pearson Edexcel.

#WMA11 — Error Pattern Tracker

#Pearson Edexcel IAL Mathematics, Unit 1: Pure Mathematics P1 — a self-tracking template

#0. What this is, and how it's different from the rest of this folder

forge-exam-technique-guide.md explains why each error costs marks, with the verbatim mark-scheme wording behind it. codex-formula-and-method-bank.md gives the methods. signal-frequency-and-boundaries.md gives topic frequency. None of them give you a place to write down your own mistakes and watch them repeat — that's what this document is. It has two parts:

§1 — a short, coded catalogue of every specific error pattern this course's research has confirmed against a real WMA11 mark scheme or examiner report. Each code is deliberately terse (a scannable label, not a re-explanation) because the full explanation already exists elsewhere in this folder and is cross-referenced, not repeated.
§3 — a blank log template you fill in after every timed practice set, matching your own dropped marks against these codes (or a new one you name yourself, per §4), so that after five or six practice papers you have a real, personal frequency table instead of a vague feeling that "I keep messing something up."

If a mistake you make doesn't fit any code below, that is expected — this catalogue only contains patterns a real examiner report or mark scheme has confirmed, across the nine WMA11 series this course's research has actually read (five fully, four partially — see WMA11-verified-facts.md §7 for the exact breakdown). It is not a claim that these are the only ways to lose marks on this paper.

#1. The error-code catalogue

Codes are grouped by spec strand (1–5, matching the spec's own numbering) plus a cross-cutting group (X). The See column names the WMA11 lesson file with the full worked treatment — go there for the marked solution, not here.

#Strand 1 — Algebra and functions

Discriminant with a parameter (spec 1.4) — See: discriminant-with-a-parameter.ts

Code	Pattern	Confirmed in
EP1a	Discriminant written as b²+4ac instead of b²−4ac	Jan 2023 Q4
EP1b	A coefficient not squared before substituting (e.g. "6" left unsquared), giving a wrong bound	Jan 2023 Q4
EP1c	A stated non-zero/positivity constraint on the parameter dropped after solving the inequality (losing a lower bound like k > 0)	Jan 2023 Q4
EP1d	Boundary written non-strict (≤) where "no real roots" needs a strict inequality (<)	Jan 2023 Q4
EP1e	The quadratic itself never formed/recognised before trying to apply the discriminant	Jan 2019 Q9

Hidden quadratic via substitution (spec 1.1/1.5/1.6) — See: hidden-quadratic-substitution.ts

Code	Pattern	Confirmed in
EP1f	Exponent laws applied incorrectly when converting the substituted terms (mapping the wrong power onto the substitution variable)	Jan 2023 Q5
EP1g	Solves for the substituted variable, then stops — never converts back to the original variable	Confirmed across two independent series: Jan 2023 Q5(b), Jan 2025 Q4(b)
EP1h	Jumps to a calculator-only answer on a step that requires a shown method	Jan 2025 Q4(b)
EP1i	General laws-of-indices fluency slip when rearranging an index expression	Jan 2019 Q2

Quadratic/reciprocal inequalities, graphically (spec 1.7–1.9) — See: quadratic-inequalities-graphically.ts; EP1l has no dedicated lesson yet

Code	Pattern	Confirmed in
EP1j	Final answer given in terms of the region label itself (e.g. "R < …") instead of x/y — this is a method-mark failure (M0), not just a lost accuracy mark	Jan 2019 Q4; Oct 2021 Q3(ii)
EP1k	Wrong-variable boundary (e.g. stating a bound on y where the question needs one on x)	Jan 2019 Q4
EP1l	A reciprocal inequality solved by a one-sided rearrangement instead of the multiply-by-x² route, silently dropping a branch	Oct 2021 Q3(i)
EP1m	Inconsistent or wrong inequality signs when combining a line condition with a curve condition	Oct 2021 Q3(ii)

Graph transformations, described in words (spec 1.12) — See: graph-transformations-in-words.ts

Code	Pattern	Confirmed in
EP1n	Transformation described with "shift" or "move" instead of the required word "translation"	Jan 2023 Q7(b)
EP1o	Stretch direction confused (e.g. reading cos(2x) as a y-axis stretch rather than an x-axis stretch of scale factor ½)	Oct 2021 Q4(a)(i)
EP1p	Sign error on a reflected value/asymptote	Jan 2019 Q8(a)
EP1q	The "use the transformed graph to find k" part left blank more often than any other part of the question	Jan 2019 Q8(c)

Simultaneous equations, line meets curve (spec 1.6) — See: simultaneous-equations-by-substitution.ts

Code	Pattern	Confirmed in
EP1r	A known value substituted the long way (re-deriving a combined equation) instead of directly, inviting extra algebra errors	Jan 2023 Q7(c)
EP1s	A shifted-denominator fraction mis-split as if the shift could be pulled out additively (6/(x−2) treated as 6/x − 3)	Jan 2023 Q7(d)
EP1t	Finds x, forgets to substitute back to find y	Jan 2023 Q7(d)
EP1u	Reaches for the discriminant when the question already gives (or asks for) a specific value	Jan 2023 Q7(d)
EP1v	A stated modelling constraint (e.g. x > y) dropped, leaving two unfiltered solution pairs at the end	June 2023 Q2

Reconstructing a quadratic from its turning point (spec 1.3/1.5) — See: quadratic-functions.ts

Code	Pattern	Confirmed in
EP1w	The leading multiplier (k / a) ignored or forgotten when rebuilding f(x) from a turning point plus one other point	June 2022 Q5(a) — examiner report: "1 out of 4 was a very common score"

Cubic factorisation / graph sketching (spec 1.10/1.11) — See: cubic-graph-sketching-from-factored-form.ts

Code	Pattern	Confirmed in
EP1x	A cusp drawn at a turning point instead of a smooth curve	Oct 2021 Q6(a)
EP1y	y-intercept left unlabelled on an otherwise-correct sketch	Oct 2021 Q6(a)

#Strand 2 — Coordinate geometry — See: `perpendicularity-proofs-and-coordinate-geometry.ts`

Code	Pattern	Confirmed in
EP2a	A correct calculation given with no closing statement connecting it to what was actually asked to be proved	Jan 2023 Q2(a)
EP2b	Difference-of-squares confused with square-of-a-difference in a distance calculation	Jan 2023 Q2(a)
EP2c	The least reliable method chosen (simultaneous equations for an unknown vertex) over a faster, more reliable one (vectors or the diagonal midpoint)	Jan 2023 Q2(b)
EP2d	A Pythagorean triple applied from the wrong reference point (the origin instead of the relevant midpoint)	Oct 2020 Q6(c)

#Strand 3 — Trigonometry

Sector area / arc length / composite perimeter (spec 3.2) — See: sector-area-and-arc-length.ts

Code	Pattern	Confirmed in
EP3a	½ omitted from the sector-area formula	Jan 2023 Q6
EP3b	A chord/base line mistaken for a diameter of the circle	Jan 2023 Q6
EP3c	Composite-perimeter slip: a length not doubled, a straight/base length missed, or a radius omitted	Jan 2023 Q6
EP3d	The minor angle used where the major/reflex angle was needed (the 2π subtraction skipped)	Oct 2021 Q7 — the facts bank notes this echoes a related finding elsewhere in the Jan 2023 series (§4.3)
EP3e	The whole composite shape wrongly treated as a single sector	Oct 2020 Q3(b)(i)

Sine/cosine rule, ambiguous case (spec 3.1) — See: sine-and-cosine-rules-ambiguous-case.ts

Code	Pattern	Confirmed in
EP3f	The obtuse solution missed in an ambiguous-case sine-rule question	Jan 2019 Q7(a)
EP3g	Angle measure mixed mid-question (working in radians where the question's given angle is in degrees)	Jan 2019 Q7(a)

Trig graphs, period and domain (spec 3.3) — See: trig-graphs-period-and-domain.ts

Code	Pattern	Confirmed in
EP3h	Period confused with domain/interval — an interval given where a single period value was asked for	Jan 2023 Q9(a) — fewer than 5% scored full marks; more than half scored none
EP3i	Sketch imprecision traced to a missing scale/axis label	Jan 2019 Q5

#Strand 4 — Differentiation

Negative/fractional indices, exact gradients (spec 4.2/4.3) — See: differentiating-negative-and-fractional-indices.ts

Code	Pattern	Confirmed in
EP4a	"+c" appended after differentiating — the procedure confused with integration	Jan 2023 Q1
EP4b	A decimal given where an exact value was required	Jan 2023 Q1(b)
EP4c	Sign/index error differentiating a negative-power term (e.g. 2/x mapped to 2x or 2, rather than −2x⁻²)	Jan 2025 Q5(a)

Hidden higher-degree equations from differentiation (spec 4.1–4.3) — See: hidden-higher-degree-equations-from-differentiation.ts

Code	Pattern	Confirmed in
EP4d	A quartic-in-x² not recognised as a solvable three-term-quadratic substitution	Oct 2021 Q2
EP4e	A claimed factorisation that doesn't actually match the unsimplified line it came from (a skipped "divide through" step)	Oct 2021 Q2
EP4f	x² and x blurred mid-solution while solving the substituted quadratic	Oct 2021 Q2
EP4g	Forgets to square-root back from x² to x	Jan 2025 Q5(c)
EP4h	Both signs of x given despite a stated domain restriction (x > 0) that rules one out	Jan 2025 Q5(c)
EP4i	A zero gradient treated as a dead end rather than a valid horizontal tangent	Oct 2021 Q6(c)
EP4j	Stops at computing f(value); never writes the actual tangent/normal equation	Oct 2021 Q6(c)

#Strand 5 — Integration

Splitting a fraction before integrating (spec 5.1/5.2) — See: splitting-fractions-before-integrating.ts

Code	Pattern	Confirmed in
EP5a	No route found to split the fraction into separate terms before integrating	Jan 2023 Q3
EP5b	All terms integrated before dividing through	Jan 2023 Q3
EP5c	Constant of integration omitted	Confirmed in 4 series checked in this pass: Jan 2023 Q3, Oct 2021 Q1, Oct 2020 Q9, Jan 2025 Q6(b)
EP5d	Final answer left with a bare exponent and no base attached	Jan 2023 Q3 (2026 re-audit)

Integrate twice, find two constants (spec 4.1/5.1/5.2 combined) — See: integrate-twice-two-constants.ts

Code	Pattern	Confirmed in
EP5e	Wrong point/values substituted when finding a constant (transposed coordinates)	Oct 2021 Q10
EP5f	A coordinate point used where a gradient value was needed, or vice versa, across a two-stage integration	Jan 2023 Q11(b)
EP5g	Both constants deferred to the very end instead of being fixed immediately after each integration step	Jan 2023 Q11(b)

#Cross-cutting (X) — not tied to one strand

Code	Pattern	Confirmed in
EPXa	A solution reached straight from a calculator on a part carrying the "solutions relying on calculator technology are not acceptable" warning, with no shown method	Jan 2025 Q4(b); the warning phrase itself is confirmed on 4 sub-parts in Jan 2023 and 2 in Jan 2025 (`WMA11-verified-facts.md` §2)
EPXb	A rounded decimal given where the question's wording, or surd-based working, clearly calls for an exact answer	General Pure-Mathematics marking principle (`WMA11-verified-facts.md` §3), plus Jan 2023 Q1(b) directly
EPXc	No evidence of a full method shown at all, on a part that requires it	Jan 2025 general examiner comments, Q7(b)

#2. If you only track a handful of codes, track these

The facts bank only states an exact multi-series count for a small number of patterns — everything else above is a real, confirmed finding but from a single series, which is not the same claim. These are the ones with a stated count:

EP5c (forgotten constant of integration) — named explicitly across 4 of the series checked (Jan 2023, Oct 2021, Oct 2020, Jan 2025). The single most repeatedly-documented "silly" mark loss in this whole catalogue.
EP1g (stops at the substituted variable) — the facts bank states this is "confirmed across two independent series" (Jan 2023, Jan 2025) sharing the exact same failure mode.
EPXa (calculator-only answer on a flagged part) — the warning phrase itself is confirmed on 4 sub-parts in one series and 2 in another; how often candidates actually ignore it and lose marks is not separately counted, so log it yourself in §3 rather than trusting a number here.

Nothing else in §1 has a stated cross-series count in the verified sources — treat single-series citations as "this is real and it happens," not "this is the most common thing that happens." Your own log in §3 is the only way to know which of these actually recur for you.

#3. Your practice log

Copy this table (or the row format) into your own notes after every timed practice set — a full past paper, a question-bank set, or even a single multi-part question done under time pressure. Log every mark you dropped, not just the ones you don't understand — a code you already "know" that still costs you a mark is exactly the pattern worth catching early.

#	Date	Source (series/paper or set)	Q#	Marks lost	Code(s)	What actually happened (one line)	Fix for next time
e.g.	2026-09-29	Jan 2023 QP, timed, 90 min	6	2	EP3a	Forgot the ½ writing the sector area — wrote r²θ, not ½r²θ	Write "A = ½r²θ" out before substituting, every time, no exceptions
1
2
3
4
5
6
7
8

The first row is an illustrative example of how to fill the table in — it is not a real result from any student or series, and should be deleted once you understand the format. Add rows freely; this table is meant to run for an entire revision cycle, not one sitting.

#4. When a mistake doesn't fit a code

Log it anyway. Write a short one- or two-word name for it in the Code column instead of an EP-code (e.g. "arithmetic-slip", "misread-question"), and keep using that same name every time it recurs. Two rules for this:

Don't force a real mistake into the nearest EP-code just to keep the table tidy — a mismatch there defeats the point of tracking in the first place.
If the same self-named pattern shows up three or more times across different practice sets, that is itself worth flagging to whoever is supervising your revision (a tutor, a teacher, or your own written notes) as a real gap this catalogue doesn't yet cover — it may simply be a pattern this course's research pass hasn't read the right series to confirm yet (see WMA11-verified-facts.md §7 for exactly which series remain unread).

#5. Review protocol

Do this every 3–4 practice sets, not after every single one — a single sitting is too small a sample to tell a real pattern from one bad day.

Tally how many times each code appears in your §3 log so far.
Sort by count. Your top 3 are your actual, evidenced weak points — not the ones that feel most embarrassing, the ones that keep recurring on paper.
For each of your top 3, go to the lesson file named in its See column in §1 and re-work that lesson's own marked solution and trap-taxonomy block before your next timed set.
After re-working, don't just move on — deliberately look for that same code's trigger condition in the next practice set (e.g. if EP5c keeps recurring, write "+c" as the very last thing you do on every integration question, checked off before you move to the next question).
If a code stops appearing for two consecutive practice sets after you've worked on it, it's fixed for now — keep the log entries (don't delete history) but stop actively targeting it, and let a new top-3 candidate take its place.

#6. Cross-references (not repeated here)

Full worked explanation of any code above, with the real mark-scheme M/A/B breakdown: the WMA11 lesson file named in its See column, content/veridian/curriculum/wma11/.
Why the mark scheme is built this way (mark types, dependent marks, the "quote the formula first" habit): forge-exam-technique-guide.md §2, this same folder.
How often each spec strand and sub-topic actually appears across the series this course has read: signal-frequency-and-boundaries.md, this same folder.
Real past questions to practise against, already mark-coded: content/veridian/question-bank/wma11.ts.
The full, unfiltered research behind every code above, including what could and couldn't be verified: research/veridian/WMA11-verified-facts.md, especially §4 (source material), §6 (which topics have enough evidence for a full lesson) and §7 (flagged gaps).

#7. Honesty notes

This catalogue draws on nine WMA11 series, read at uneven depth: five with a fully-read examiner report (Jan 2019, Oct 2020, Oct 2021, Jan 2023, Jan 2025), two confirmed-genuine mark schemes not mined question-by-question (Oct 2022, Jan 2024), and two read for one targeted question each (June 2022 Q5; June 2023 Q1–Q2). A code's absence from a series is not evidence the error doesn't happen there — it may simply mean that series wasn't read closely enough to catch it.
No grade boundary appears anywhere in this document, for the same reason none appears in signal-frequency-and-boundaries.md §2: the one boundary figure this course's research has seen (October 2022) was never checked against Pearson's own primary grade-boundaries document, and is explicitly marked unverified in the source file.
Every code above is traceable to a named series and question in research/veridian/WMA11-verified-facts.md. If a future update to that file corrects or retracts a finding, this tracker's matching code should be corrected in the same pass, not left stale.