Pure Mathematics 2
Error Pattern Tracker
Provenance: written 2026-09-29, against research/veridian/WMA12-verified-facts.md (read in full, §5–§8 in close detail for source scope and per-topic citations), content/veridian/technique-hub/wma12/forge-exam-technique-guide.md (Part 4, "Common-error digest," read in full), content/veridian/technique-hub/wma12/signal-frequency-and-boundaries.md (Part 1, topic-frequency table, read in full), and the trap-taxonomy block of each of the 11 real WMA12 lesson files in content/veridian/curriculum/wma12/ (proof.ts, factor-and-remainder-theorems.ts, the-circle.ts, sequences-and-arithmetic-series.ts, geometric-series.ts, binomial-expansion.ts, logarithms-and-exponentials.ts, trigonometric-equations.ts, stationary-points-and-curve-sketching.ts, definite-integrals-and-areas.ts, the-trapezium-rule.ts — each read directly this session). content/veridian/question-bank/wma12.ts was also checked directly; it holds two real past questions and no error-pattern content that overlaps with this document. Every quoted phrase below is trimmed to under 15 words and attributed to the series and question it came from; every category is labelled with its actual source (an examiner-report finding, a mark-scheme requirement, or — for the handful flagged explicitly — no citation at all, carried over honestly from the lesson file's own admission).
WMA12 — Error Pattern Tracker
What this document is, and isn't
This is a fillable self-tracking template, not a teaching document. It has two jobs: give a student a category to log a lost mark against every time they mark a timed practice paper (Part 2), and give them blank tables to actually do that logging in, session after session (Parts 3–4). It reuses the exact category names — the trap-taxonomy name slugs — already built into the 11 real WMA12 lesson files, so a student's own log lines up with the course's own taxonomy rather than inventing a second, incompatible one.
It is deliberately not:
- The trap-taxonomy itself. Every category below is a one-line pointer with a trimmed quote, not the full worked exemplar, wrong-path, and MCQ that already exists in its home lesson file — go there to actually learn the fix.
- FORGE's own locator (
forge-exam-technique-guide.md, Part 4), which indexes the same 66-plus taxonomy items by lesson for revision routing. This document indexes them differently — as loggable categories with citation and confirmation status — and adds the fillable tracking tables FORGE does not have. - SIGNAL's frequency table (
signal-frequency-and-boundaries.md), which counts how often a topic has been confirmed tested. Part 5 below borrows those counts once, to weight where a tracker will fill up fastest — it does not re-derive them. - A source of any invented statistic. No likelihood, percentage, or grade boundary appears anywhere below that isn't already stated in one of the source documents listed in the provenance note. Where a category has no real citation behind it, that is stated plainly, not smoothed over.
Part 1 — How to use this tracker
- After every timed practice paper — a real past paper, marked against its real mark scheme — go through every lost mark, not just the ones that feel like "silly mistakes."
- For each lost mark, find the closest matching category tag in Part 2, grouped by topic below. If nothing genuinely fits, write it in your own words at the bottom of that topic's table rather than forcing a bad match — a mismatch logged as a false category is worse than an ungrouped one.
- Tally it in Part 3, under that paper's column.
- Write the specific detail in Part 4's log. The tally alone tells you that a category recurs; the log entry is what tells you why, which is what actually changes what you do differently next time.
- Before your next timed paper, re-read Part 3. Whichever row carries the most tallies is your next targeted drill — using the technique-hub lesson file named against that category in Part 2.
- Re-read Part 6 (flagged limitations) occasionally. A handful of categories below carry no real citation; treat those as plausible-but-unconfirmed, not as equally weighted against the ones with a named series and quote behind them.
Part 2 — Confirmed category reference, by topic
Each row: the category tag exactly as it appears in its lesson file's trap-taxonomy, what it looks like in practice, and where it's confirmed from. Type marks whether the source shows it actually costing a mark ("Error"), costing time/risk without necessarily costing the mark on its own ("Inefficient choice"), or carrying no citation at all ("Not confirmed" — included only because the lesson file itself includes it, flagged the same way here).
6.1 Proof by exhaustion / counter-example — proof.ts
| Category tag | What it looks like | Confirmed by | Type |
|---|---|---|---|
algebraic-proof-attempted-when-exhaustion-required | Reaching for algebra/logic instead of the bounded case-check the question restricts you to | "unnecessarily long-winded" (Oct 2022); "more demanding than... a numerical one" (Oct 2023) | Error |
numerical-examples-mistaken-for-general-proof | Checking a handful of values on an infinite/unbounded domain and treating that as proof | worst-scoring question in its series — "considered that several numerical examples... constituted a proof" (Oct 2021) | Error |
multiple-of-k-assertion-without-quotient-shown | Asserting divisibility without showing the actual whole-number quotient | "insufficient without further clarification" (Oct 2023) | Error |
product-misread-as-sum | Checking the sum of the values when the question named their product (or vice versa) | "not widely understood... considering the sum" (Oct 2022) | Error |
case-defining-relation-applied-backwards | Using the stated relation in reverse, generating the wrong case list from line one | "using c = b − 2 instead of c = b + 2" (Oct 2022) | Error |
column-values-transposed-between-derived-variables | Individually-correct values written under the wrong variable's column | "a mix up between columns b and c" (Oct 2023) | Error |
case-list-not-exactly-complete | Extra rows beyond the bounded set, or rows missing from it | "the addition of extra rows" (Oct 2023) | Error |
conclusion-omitted | Correct, complete case list with no closing statement that the claim therefore holds | "omission of a conclusion" (Oct 2023) | Error |
counter-example-search-continued-past-the-first-valid-one | Finding and justifying more than one counter-example when one suffices | "many found far more than necessary" (June 2019) | Time cost, not a mark |
6.2 Factor theorem / remainder theorem / algebraic division — factor-and-remainder-theorems.ts
| Category tag | What it looks like | Confirmed by | Type |
|---|---|---|---|
root-sign-taken-straight-from-the-bracket | Reading the root off a factor (2x+1) as x = ½ instead of solving the bracket for zero | "setting f(½) = 0... and/or f(−2) = −25" — wrong signs, both directions (Oct 2023) | Error |
remainder-equation-set-to-zero | Substituting the wrong value, or setting the remainder expression equal to zero by reflex | "using f(−2) or f(3) rather than f(2)" (Jan 2024) | Error |
long-division-chosen-when-substitution-was-asked-for | Valid but measurably slower; a real risk of abandoning it mid-way | "many giving up part way through the division" (Jan 2024) | Inefficient choice (can become an Error) |
factorisation-stopped-one-step-short | Leaving a further-factorisable quadratic factor unfactorised | "leaving the answer to (c) as (2x+1)(2x²+x−15)" (Oct 2023) | Error |
hence-factorise-reached-without-using-the-given-factor | Reaching the correct final factorisation by a route that ignores the "hence" instruction | "score M0 and therefore A0 A0" for the correct answer, wrong route (Oct 2023) | Error |
non-integer-factor-form-given | Writing a factor as (x − 7/2) instead of the required integer-coefficient (2x − 7) | "This is incorrect as integer values were required" (June 2025 Regional 01R) | Error |
answer-left-as-a-list-not-a-product | Listing the factors or roots instead of writing their product | mark scheme requires it "written out as a product of factors (not as a list)" (Jan 2020) | Error |
show-that-steps-omitted | Reaching the printed target answer without showing the intermediate line that gets there | "not showing the = 0 or intermediate lines" (Oct 2023) | Error |
6.3 The circle — the-circle.ts
| Category tag | What it looks like | Confirmed by | Type |
|---|---|---|---|
centre-sign-not-flipped | Reading a centre coordinate straight off a completed-square bracket instead of its negative | confirmed sign-flip error on a real centre/radius question (Jan 2023, Q6) | Error |
general-form-read-as-centre-radius-form | Reading centre/radius off the multiplied-out equation without completing the square first | source notes most candidates do complete the square correctly here — included for the mechanism, not as a common failure | Not the observed error (context only) |
solved-parameter-not-substituted-into-earlier-general-centre | Solving for an unknown constant but not feeding it back into an earlier general centre expression | confirmed as the single biggest mark-loss point on the question it was checked against (Jan 2023, Q6) | Error |
radius-given-as-decimal-not-exact-surd | Giving r = 4.47 instead of the required exact r = 2√5 | "a common error was giving the answer r = √20" as a decimal (Jan 2023, Q6) | Error |
implicit-differentiation-chosen-over-the-property | Differentiating the circle's equation implicitly instead of using radius-perpendicularity | "very few attempted... these often made little progress" (Jan 2021, Q9) | Inefficient choice, often becomes an Error |
origin-and-circle-centre-both-labelled-o | Treating the origin, O, as the circle's centre when both appear in the same question | "the misunderstanding that O was the centre... not the origin" (Jan 2021, Q9) | Error (mark scheme was generous this once) |
substitution-direction-not-checked-for-fractions | Substituting in the direction that leaves fractions everywhere, when the other direction doesn't | "(−2y−10) for x was more successful than (−0.5x−5) for y" (Jan 2023, Q6) | Inefficient choice |
general-area-formula-chosen-over-axis-aligned-base-height | Using ½ab sinC (or a shoelace formula) for a triangle whose base sits on an axis | "many candidates not using the simple formula for the area" (Jan 2021, Q9) | Inefficient choice — both credited, no mark actually lost |
Two related patterns flagged by FORGE (Part 4) but not yet built into the-circle.ts's own taxonomy — log against them anyway if you make either:
- Double-reciprocal normal-gradient error: after correctly finding a normal's gradient, taking its negative reciprocal again and using that. "Some went on to find the negative reciprocal and use this as their normal gradient" (Jan 2023, Q6).
- Diameter mistaken for radius: finding the distance between two given points that are a diameter's endpoints, then using that distance itself as the radius. "Treated it as though it was the radius" (June 2025, Q2).
6.4a Arithmetic sequences and series — sequences-and-arithmetic-series.ts
| Category tag | What it looks like | Confirmed by | Type |
|---|---|---|---|
sum-proof-last-term-off-by-one | Writing the standard sum-formula proof's opening line with the wrong last term | "often having the last term incorrect" (Jan 2020, Q8(i)) | Error |
sum-proof-reversal-skipped | Never writing the reversed copy of the sum the pairing argument needs | "failed to reverse the terms... or at all" (Jan 2020, Q8(i)) | Error |
sum-proof-insufficient-terms-shown | Too few terms shown at each end to make the next step legible | "did not display a sufficient number of terms" (Jan 2020, Q8(i)) | Error |
induction-attempted-for-standard-proof | Reaching for proof by induction on a question that wants the standard pairing proof | "little progress was made in most of these" (Jan 2020, Q8(i)) | Inefficient choice / Error if abandoned |
sum-proof-assumes-the-result | Starting the proof from a formula that is itself the thing being proved | "started with... something that they were trying to prove" (Jan 2020, Q8(i)) | Error |
sum-proof-pairing-fails-for-odd-n | Pairing terms within a single copy of the sum without handling an odd number of terms | candidates "did not give due consideration" to this case (Jan 2020, Q8(i)) | Error |
arithmetic-and-geometric-formulas-confused | Reaching for the geometric formula mid-question on an arithmetic series, or vice versa | recurring across at least two series (June 2019, Oct 2020) | Error (shared category, see 6.4b) |
6.4b Geometric series, incl. sum to infinity — geometric-series.ts
| Category tag | What it looks like | Confirmed by | Type |
|---|---|---|---|
arithmetic-and-geometric-formulas-confused | See 6.4a — the identical category, shared across both lessons | June 2019, Oct 2020 | Error |
overlap-term-double-counted | A term shared between an arithmetic and a geometric part of one question, counted in both | "double counted this term" (Jan 2022) | Error |
geometric-series-restarted-from-the-original-first-term | Starting the geometric part from the sequence's very first term instead of the hand-off term | "assuming they had to start... from the first term... again" (Jan 2022) | Error |
sum-to-infinity-first-term-misidentified | Using the sequence's original first term as a, not the first term of the sum actually asked for | "often used a = 6 instead of a = 1.5" (Jan 2024, Q5(i)) | Error |
calculator-summation-shown-without-formula | A correct numeric sum-to-infinity value with the formula never written down | "even if [the value] was found, no marks were scored" (Jan 2024, Q5(i)) | Error |
standard-proofs-not-known | Not knowing the standard sum-formula proof for either series type | examiners "advised us to make sure they know" both proofs (Jan 2020) | Error |
geometric-proof-subtraction-direction-reversed | Subtracting S_n − rS_n the wrong way round in the geometric sum-formula proof | "subtracted the wrong way round" (Oct 2020, Q8(a)) | Error |
convergence-condition-read-as-r-less-than-one | Reading |r| < 1 as r < 1, silently admitting every negative ratio below −1 | no citation — mechanism-derived only, per the lesson file's own note | Not confirmed |
6.5 Binomial expansion — binomial-expansion.ts
| Category tag | What it looks like | Confirmed by | Type |
|---|---|---|---|
term-index-mismatched-to-power | Using the wrong power or wrong binomial coefficient for the required term | "using the wrong power... or the wrong binomial coefficient" (Jan 2024, Q2) | Error |
general-series-form-chosen-over-combinatorial-form | Reaching for the general (1+x)ⁿ series form when (a+b)ⁿ is simpler and sufficient | confirmed to make an already-hard two-unknown question harder still (June 2025 Regional 01R, Q7) | Inefficient choice, can become an Error |
bx-term-not-raised-as-a-whole | Treating (4x)⁷ as 4x⁷, raising only the variable and not the coefficient | "using invisible brackets... resulting in multiplication by 4 instead of 4⁷" (Jan 2024, Q2) | Error |
sign-lost-on-odd-powers-of-a-negative-b | Writing (−2)ʳ as 2ʳ, dropping the sign on an odd power of a negative term | no citation — mechanism-derived only | Not confirmed |
pascal-triangle-row-miscounted | Starting the coefficient count at "row 1" instead of "row 0" | no citation — mechanism-derived only | Not confirmed |
required-term-not-isolated-from-a-list | Finding the right value but leaving it unlabelled inside a longer expansion or list | "failed to isolate the required term from a list" (Jan 2024, Q2) | Error |
6.6 Exponentials and logarithms — logarithms-and-exponentials.ts
| Category tag | What it looks like | Confirmed by | Type |
|---|---|---|---|
log-subtraction-rewritten-as-division | Rewriting log(x) − log(y) as log(x)/log(y), then cancelling the log symbols like a fraction | "crossed out log₃ in the numerator and denominator" (Jan 2023, Q4(i)) | Error |
log-of-a-sum-expanded-as-sum-of-logs | Writing log(5x+7) as log(5x) + log(7) | "expand the brackets within the log function" (Jan 2023, Q4(i)) | Error |
log-squared-confused-with-log-of-the-square | Rewriting (log y)² as 2 log y | "a very common mistake" on one of the paper's hardest parts (Jan 2023, Q4(ii)) | Error |
exponent-rearrangement-left-unfinished | Setting up the correct log equation but not finishing the algebra to isolate x | mark-by-mark drop from ~90% to ~70% (Jan 2021, Q3(i)) | Error |
single-log-form-left-uncombined | A genuinely correct expression such as x = log₇3 − 2, submitted without combining to the required single-log form | "did not reach the required form, failing to deal with the −2" (Jan 2021, Q3(i)) | Error |
logs-taken-before-the-equation-is-rearranged | Taking logs immediately instead of isolating the exponential term first | "not always advisable to take logs first" (Jan 2021, Q3(i)) | Inefficient choice |
6.7 Trigonometric equations — trigonometric-equations.ts
| Category tag | What it looks like | Confirmed by | Type |
|---|---|---|---|
divides-out-a-common-trig-factor | Cancelling sinθ from sinθ(2+3cosθ)=0 instead of setting both factors to zero | "only produced solutions from cosθ = −⅔" (Oct 2023, Q3) | Error |
wrong-identity-substituted-for-cos-squared | Replacing cos²θ with 1 − sinθ instead of 1 − sin²θ | "cos2θ was occasionally replaced with 1 − sinθ" (Oct 2022, Q5) | Error |
coefficient-dropped-when-distributing-after-substitution | Expanding 3(1 − sin²θ) to 3 − sin²θ, multiplying only one term | verbatim example given in the report (Oct 2022, Q5) | Error |
incomplete-multiply-through-after-tan-substitution | Clearing a cosθ denominator from only some terms of the equation, not all | same distribute-across-every-term hazard, at the fraction-clearing step | Error |
compound-angle-not-isolated-before-inverse-function | Taking arctan before isolating the compound-angle term it applies to | "some did not recognise that they should isolate" it first (June 2025, Q8(i)) | Error |
coefficient-distributed-across-the-whole-compound-angle | Multiplying a coefficient into the angle inside a trig function instead of its output | "distributing the coefficient... resulting in tan(6x + 3π/5)" (June 2025, Q8(i)) | Error |
compound-angle-bracket-range-scaled-by-coefficient | Generating too few (or wrong) bracket values when the range is scaled by a coefficient | mark scheme's own credited multi-value method (June 2025, Q8(i)) | Error if a value is missed |
decimal-given-instead-of-exact-kπ-radian-form | Giving a decimal angle, or working in degrees, when an exact kπ radian answer is required | "used decimals... worked in degrees" (June 2025, Q8(i)) | Error |
squared-compound-angle-sign-or-period-copy-dropped | Taking only the positive square root, or missing a solution once the range is doubled | "many ignored the negative root" (Oct 2021, Q10(i)) | Error |
premature-rounding-before-final-angle | Rounding an exact intermediate value (e.g. cosθ) before finding the final angle | "rounding errors when finding θ" (June 2025, Q8(ii)) | Error |
6.8 Differentiation — stationary points and curve sketching — stationary-points-and-curve-sketching.ts
| Category tag | What it looks like | Confirmed by | Type |
|---|---|---|---|
second-derivative-found-but-never-substituted | Finding the correct second-derivative expression, never substituting the x-value into it | "did not then substitute the values for x" (Jan 2021, Q2) | Error |
y-coordinates-found-when-only-x-was-asked | Extra, unrequested work finding y-coordinates when a part asked only for x | "unnecessary work by finding the y coordinates" (Jan 2021, Q2) | Time cost, not a mark |
second-derivative-set-to-zero-and-solved | Solving d²y/dx² = 0 as if it were a fresh stationary-point equation | "a common error was to try to solve for the second derivative" (Jan 2021, Q2) | Error |
y-values-eyeballed-instead-of-gradient-tested | Judging a stationary point's nature from the curve's shape/y-values instead of gradient sign | "not given credit as it did not use further calculus" (Jan 2021, Q2) | Error |
critical-value-inequality-wrong-region-chosen | Choosing the outside range of two critical values when the inside range was correct (or vice versa) | wrong answer "x < −5 or x > ⅔" recorded verbatim (June 2025, WMA12/01A, Q8) | Error |
second-derivative-test-inconclusive-mishandled | Treating a zero second derivative as automatically meaning "no information" or "must be inflection" | no citation — mechanism-derived only, per the lesson file's own note | Not confirmed |
no-turning-points-asserted-without-a-discriminant-shown | Stating "no solutions" for a non-factorising quadratic without forming the discriminant | "did not know what to do next... without showing how they knew this" (Oct 2019, Q10(b)) | Error |
stationary-point-x-coordinate-rounded-instead-of-exact | Rounding a genuinely fractional stationary-point x-value instead of leaving it exact | general marking guidance's exact-answer rule, applied to this pattern | Error |
6.9a Definite integrals and area — definite-integrals-and-areas.ts
| Category tag | What it looks like | Confirmed by | Type |
|---|---|---|---|
answer-produced-with-no-method-shown | A calculator-only decimal for a definite integral where an exact answer was required | "full method must be shown for full credit" (Jan 2020, Q4) | Error |
curve-and-line-subtracted-in-the-wrong-order | Subtracting line − curve instead of upper − lower, flipping the sign of the result | "not uncommon to see the difference the wrong way round" (Jan 2020, Q4) | Error (self-correcting only if you check the sign) |
line-boundary-ignored-entirely | Integrating only the curve, never subtracting the bounding line at all | "completely ignoring the requirement to subtract the area under the line" (Jan 2023, Q9(c)) | Error, confirmed on two separate series |
sign-change-inside-the-interval-not-split | Integrating straight across a root instead of splitting the interval either side of it | no citation — mechanism-derived only, per the lesson file's own note | Not confirmed |
slowest-valid-method-chosen-under-time-pressure | Picking a mathematically valid but slower route (e.g. five terms instead of one) | "a more time-consuming approach... occasional sign errors" (June 2025, Q3) | Inefficient choice |
y-values-used-as-x-limits | Using two y-values off the diagram as the limits of an integral with respect to x | "use the y values of 7 and 17 as their limits" (Jan 2020, Q4) | Error |
6.9b The trapezium rule — the-trapezium-rule.ts
| Category tag | What it looks like | Confirmed by | Type |
|---|---|---|---|
strip-width-from-point-count-not-gap-count | Dividing the interval by the number of ordinates instead of the number of strips | "dividing 1 by the 5 points rather than the 4 spaces" (Oct 2020) | Error |
endpoints-and-interior-ordinates-swapped | Doubling the wrong ordinates, or doubling every value including the endpoints | "the commonest error was to double the three given values" (Oct 2020) | Error |
bracket-closed-before-scaling | Closing the sum's bracket after only the first two terms, adding the rest outside it | "gaining no credit" for the resulting mis-scaled sum (Jan 2023) | Error |
conclusion-restated-instead-of-reasoned | Restating "it is an underestimate" without saying why | "the most difficult mark for candidates to achieve" on its paper (Oct 2023) | Error |
trapezium-rule-assumed-tested-in-isolation | Expecting a stand-alone numeric calculation, not a rule embedded in another context | tested "via logarithms" (Jan 2025) and in a real-world context (Oct 2023) | Not itself a marks-lost error — a preparation gap |
Part 3 — Frequency tally sheet
Fill in a tally mark under the paper you were practising, every time that category cost you a mark (or, for "Time cost" / "Inefficient choice" rows, every time it slowed you down even without costing the mark). Add more paper columns as needed — six are provided as a starting grid.
| Category tag | P1 | P2 | P3 | P4 | P5 | P6 | Total |
|---|---|---|---|---|---|---|---|
| 6.1 Proof | |||||||
| algebraic-proof-attempted-when-exhaustion-required | |||||||
| numerical-examples-mistaken-for-general-proof | |||||||
| multiple-of-k-assertion-without-quotient-shown | |||||||
| product-misread-as-sum | |||||||
| case-defining-relation-applied-backwards | |||||||
| column-values-transposed-between-derived-variables | |||||||
| case-list-not-exactly-complete | |||||||
| conclusion-omitted | |||||||
| counter-example-search-continued-past-the-first-valid-one | |||||||
| 6.2 Factor/remainder theorems | |||||||
| root-sign-taken-straight-from-the-bracket | |||||||
| remainder-equation-set-to-zero | |||||||
| long-division-chosen-when-substitution-was-asked-for | |||||||
| factorisation-stopped-one-step-short | |||||||
| hence-factorise-reached-without-using-the-given-factor | |||||||
| non-integer-factor-form-given | |||||||
| answer-left-as-a-list-not-a-product | |||||||
| show-that-steps-omitted | |||||||
| 6.3 The circle | |||||||
| centre-sign-not-flipped | |||||||
| solved-parameter-not-substituted-into-earlier-general-centre | |||||||
| radius-given-as-decimal-not-exact-surd | |||||||
| implicit-differentiation-chosen-over-the-property | |||||||
| origin-and-circle-centre-both-labelled-o | |||||||
| substitution-direction-not-checked-for-fractions | |||||||
| general-area-formula-chosen-over-axis-aligned-base-height | |||||||
| double-reciprocal-normal-gradient (FORGE-flagged) | |||||||
| diameter-mistaken-for-radius (FORGE-flagged) | |||||||
| 6.4a Arithmetic series | |||||||
| sum-proof-last-term-off-by-one | |||||||
| sum-proof-reversal-skipped | |||||||
| sum-proof-insufficient-terms-shown | |||||||
| induction-attempted-for-standard-proof | |||||||
| sum-proof-assumes-the-result | |||||||
| sum-proof-pairing-fails-for-odd-n | |||||||
| arithmetic-and-geometric-formulas-confused | |||||||
| 6.4b Geometric series | |||||||
| overlap-term-double-counted | |||||||
| geometric-series-restarted-from-the-original-first-term | |||||||
| sum-to-infinity-first-term-misidentified | |||||||
| calculator-summation-shown-without-formula | |||||||
| standard-proofs-not-known | |||||||
| geometric-proof-subtraction-direction-reversed | |||||||
| convergence-condition-read-as-r-less-than-one (unconfirmed) | |||||||
| 6.5 Binomial expansion | |||||||
| term-index-mismatched-to-power | |||||||
| general-series-form-chosen-over-combinatorial-form | |||||||
| bx-term-not-raised-as-a-whole | |||||||
| sign-lost-on-odd-powers-of-a-negative-b (unconfirmed) | |||||||
| pascal-triangle-row-miscounted (unconfirmed) | |||||||
| required-term-not-isolated-from-a-list | |||||||
| 6.6 Logs and exponentials | |||||||
| log-subtraction-rewritten-as-division | |||||||
| log-of-a-sum-expanded-as-sum-of-logs | |||||||
| log-squared-confused-with-log-of-the-square | |||||||
| exponent-rearrangement-left-unfinished | |||||||
| single-log-form-left-uncombined | |||||||
| logs-taken-before-the-equation-is-rearranged | |||||||
| 6.7 Trigonometric equations | |||||||
| divides-out-a-common-trig-factor | |||||||
| wrong-identity-substituted-for-cos-squared | |||||||
| coefficient-dropped-when-distributing-after-substitution | |||||||
| incomplete-multiply-through-after-tan-substitution | |||||||
| compound-angle-not-isolated-before-inverse-function | |||||||
| coefficient-distributed-across-the-whole-compound-angle | |||||||
| compound-angle-bracket-range-scaled-by-coefficient | |||||||
| decimal-given-instead-of-exact-kπ-radian-form | |||||||
| squared-compound-angle-sign-or-period-copy-dropped | |||||||
| premature-rounding-before-final-angle | |||||||
| 6.8 Stationary points / curve sketching | |||||||
| second-derivative-found-but-never-substituted | |||||||
| second-derivative-set-to-zero-and-solved | |||||||
| y-values-eyeballed-instead-of-gradient-tested | |||||||
| critical-value-inequality-wrong-region-chosen | |||||||
| second-derivative-test-inconclusive-mishandled (unconfirmed) | |||||||
| no-turning-points-asserted-without-a-discriminant-shown | |||||||
| stationary-point-x-coordinate-rounded-instead-of-exact | |||||||
| 6.9a Definite integrals / area | |||||||
| answer-produced-with-no-method-shown | |||||||
| curve-and-line-subtracted-in-the-wrong-order | |||||||
| line-boundary-ignored-entirely | |||||||
| sign-change-inside-the-interval-not-split (unconfirmed) | |||||||
| y-values-used-as-x-limits | |||||||
| 6.9b Trapezium rule | |||||||
| strip-width-from-point-count-not-gap-count | |||||||
| endpoints-and-interior-ordinates-swapped | |||||||
| bracket-closed-before-scaling | |||||||
| conclusion-restated-instead-of-reasoned |
Part 4 — Detailed mistake log
One row per lost mark. The tally in Part 3 tells you a category recurs; this row is where you write down enough to actually fix it.
| Date | Series/paper & Qn | Category tag | Marks lost / available | What I actually wrote | What the mark scheme required | Fix for next attempt |
|---|---|---|---|---|---|---|
| (example — overwrite before your first real entry) | Jan 2024, Q1 | long-division-chosen-when-substitution-was-asked-for | 2/3 | Started polynomial long division, ran out of time, abandoned it | f(2) substituted directly into the cubic, set equal to the given remainder | Try substitution first on any "find a coefficient" opener; only divide if a later part needs the quotient |
Part 5 — Where a tracker fills up fastest (not a prediction)
This table borrows, without re-deriving, the per-topic "seen in N of 20 series checked" counts already established in signal-frequency-and-boundaries.md Part 1, alongside this document's own count of how many distinct, examiner- or mark-scheme-confirmed error categories exist per topic (Part 2 above; "unconfirmed" rows excluded from the count).
| Topic | Seen in N of 20 series (SIGNAL) | Confirmed error categories logged here |
|---|---|---|
| Proof | 6 | 9 |
| Factor/remainder theorems | ≥4 | 8 |
| The circle | ≥5 | 7 confirmed + 2 FORGE-flagged |
| Arithmetic series | 4 | 7 (shared category with geometric) |
| Geometric series | 2 | 6 confirmed + 1 unconfirmed |
| Binomial expansion | 2 series / 3 papers | 4 confirmed + 2 unconfirmed |
| Logs and exponentials | 5 | 6 |
| Trigonometric equations | 4 | 10 |
| Stationary points / curve sketching | 4 | 6 confirmed + 1 unconfirmed |
| Definite integrals / area | 3 | 5 confirmed + 1 unconfirmed |
| Trapezium rule | 6 | 5 |
How to read this, and how not to. A topic with more confirmed categories has had more of its specific ways of losing marks individually documented so far — it says nothing about how likely that topic is to appear on your paper, and every row above is a compulsory spec area regardless of its count. Trigonometric equations currently carries the most named categories (10) largely because two single series (Oct 2021, June 2025) each yielded several distinct, closely-related traps on compound-angle questions — depth of documentation on those two series, not a claim that trig is inherently more error-prone than the circle. Use this table to decide which topic's tally rows in Part 3 to check first when deciding where to drill next, not to skip revising a topic that happens to have fewer rows.
Part 6 — Flagged limitations
- Six categories above carry no examiner-report or mark-scheme citation at all, and are marked "Not confirmed" or "unconfirmed" throughout:
convergence-condition-read-as-r-less-than-one,sign-lost-on-odd-powers-of-a-negative-b,pascal-triangle-row-miscounted,second-derivative-test-inconclusive-mishandled,sign-change-inside-the-interval-not-split. Each is carried over honestly from its own lesson file, which states the same thing about itself — they are genuine, checkable mathematical failure modes, not fabricated ones, but no real student's error has been documented against any of them in this course's archive to date. Log against them if you actually make the mistake; don't treat a tally there as confirming a pattern the source material doesn't support. - Two categories in the circle table are not, on inspection, confirmed marks-lost errors:
general-form-read-as-centre-radius-form(the source states most candidates complete the square correctly) andgeneral-area-formula-chosen-over-axis-aligned-base-height(both routes are equally credited by the mark scheme). Both are kept in Part 2/3 because a student can still lose time on them, but neither should be read as "commonly wrong," because the sources say the opposite. - The two FORGE-flagged circle patterns and the two 3-D-optimisation/Q2-diameter patterns FORGE names in its own Part 4 (double-reciprocal normal gradient; diameter mistaken for radius) are real, examiner-confirmed findings, but — per FORGE's own note — not yet built into any lesson file's trap-taxonomy. This tracker lists the two circle ones (rows added to Parts 2–3) because they fit an existing lesson's topic cleanly; the June 2025 Q9 3-D-optimisation family (open-vs-closed prism, halving a cross-sectional area, cube-root-vs-square-root slips) is not included here because it does not yet correspond to a topic this document's structure covers — see
forge-exam-technique-guide.mdPart 4, items 4–5, directly for that material. - Coverage across series is uneven, per the facts bank's own §8: roughly two-thirds of the 17 series with an examiner report were read in full depth; the remainder were confirmed present but not exhaustively mined. A category with zero tallies in your own log may reflect a genuinely rare error, or a series this course's own source material hasn't fully mined yet — not necessarily the same thing.
- No examiner report exists for October 2025 or January 2026 (too recent, not a confirmed permanent gap) and none for June 2021 (a confirmed, COVID-disrupted absence). Nothing in this document is sourced from those three series' commentary.
- This document invents no likelihood, percentage, or grade boundary anywhere. Where a frequency is stated (Part 5), it is a direct, unmodified reuse of a count already published in
signal-frequency-and-boundaries.md, itself sourced to the facts bank's own §5 archive scope — never a new calculation performed for this document. - A blank tracker changes nothing by itself. Its only value is in being filled in, honestly, against real timed practice — and in being read back before the next paper, not just written into.
Sources used
research/veridian/WMA12-verified-facts.md— read in full; §5 (archive scope: 20 series with question paper + mark scheme, 17 with an examiner report, June 2019–January 2026) and §6.1–§6.9 (the per-topic verified quotes this document's Part 2 draws its citations from) read in close detail; §8 (flagged issues) is the direct source for several of this document's own Part 6 caveats.content/veridian/technique-hub/wma12/forge-exam-technique-guide.md— Part 4 ("Common-error digest") read in full; source of the two circle patterns flagged in Part 2/3 as not yet inthe-circle.ts's own taxonomy, and of the note (Part 6 above) about the June 2025 Q9 3-D-optimisation family this document does not attempt to cover.content/veridian/technique-hub/wma12/signal-frequency-and-boundaries.md— Part 1 read in full; direct, unmodified source of every "seen in N of 20" count reused in this document's Part 5.- The trap-taxonomy block of each of the 11 real WMA12 lesson files in
content/veridian/curriculum/wma12/— read directly this session, file by file, as the source of every category tag, its exact wording, its trimmed quote, and its confirmation status (including the handful each lesson file itself flags as uncited). content/veridian/question-bank/wma12.ts— checked directly; contains two real past questions (Oct 2019 Q10, Oct 2021 Q10(i)) and no error-pattern content that overlaps with this document.