Statistics vs mechanics: mastering the applied paper you hate

Almost every student I teach is clearly weaker at one of statistics and mechanics, and very few for the reason they think. The two halves of applied maths fail in completely different places: mechanics collapses before any maths happens, at the setup, and statistics collapses after the maths is finished, in the sentence you write underneath. So "I'm bad at applied" is not a diagnosis, and the two fixes have almost nothing in common. Applied is also the smallest and most repetitive part of the A-Level, which makes it the cheapest place on the course to win marks back.

Where the applied marks actually sit

A third of A-Level Maths is applied, split roughly evenly between the two. On Edexcel it all lands in one place: Papers 1 and 2 are pure, and Paper 3 is two hours and 100 marks divided into a 50-mark statistics section and a 50-mark mechanics section. AQA spreads it out — Paper 2 is pure plus mechanics, Paper 3 is pure plus statistics — and OCR's A specification does the same in the other order. Whichever board you are on, about a sixth of your total is statistics and about a sixth is mechanics.

That shape matters. On Edexcel a mechanics weakness is contained: it costs one bad hour of one morning and leaves two pure papers untouched. On AQA and OCR it bleeds into a paper you were otherwise going to do well in. I have written elsewhere about what genuinely differs between the exam boards; this is the difference that should change how you revise.

Mechanics fails before the maths starts

Students who tell me they cannot do mechanics can almost always do the mechanics. Hand them the equation and they solve it without trouble. What they cannot do is arrive at the equation — decide which body to look at, draw the forces on it, choose a direction, and commit to it.

The constant acceleration formulae are printed in the formulae booklet; I have just checked AQA's, and all five are there. What is not there is W = mg, or anything about friction, or which way a normal reaction points. The booklet gives you the algebra and expects you to supply the physics, so mechanics marks are essentially never lost to a forgotten formula.

The fix is a ritual, performed identically on every question, however easy it looks:

• Draw the body on its own, with every force as an arrow leaving it. Nothing else goes on that diagram — no velocities, no distances, no working.
• Mark the positive direction once, in the margin, and never change it. Most sign errors are a decision that was never actually made.
• Write down what the modelling words give you. "Light and inextensible" means the tension is the same all along the string; "smooth" means there is no friction term; "particle" means nothing rotates. Students read those words as scene-setting. They are instructions.
• Only then write F = ma — once per body, per direction.

Every student resists this, because the first three steps earn no marks on their own. They earn every mark that comes after them. A fortnight of doing it properly usually ends the "bad at mechanics" self-description.

Statistics fails after the maths is finished

Statistics is the mirror image, and it catches a different sort of student. The specification openly expects you to use your calculator for the probabilities — Pearson's own content guidance for the Edexcel course says students should use a calculator to find cumulative binomial probabilities and probabilities connected with the normal distribution. The computation is not where the marks are. It is also where the machine genuinely matters, which is why I am fussy about which calculator a student buys.

The marks are in everything around the number: choosing a model and saying why it is reasonable, stating hypotheses about a parameter rather than about the thing that happened, and writing a conclusion in the language of the question rather than of statistics. The loss I mark most often is a perfectly correct probability followed by "reject H0" and nothing else — or by a claim far stronger than the evidence, like "this proves the coin is biased". Two sentences fix most of it: one statistical, one in context, the second containing a noun lifted straight from the question.

So in statistics, more than anywhere else on the course, "I got the right answer" and "I got the marks" come apart. Mark your own work against the mark scheme's wording rather than your final number — the discipline I set out in the right way and the three wrong ways to use past papers.

Why students polarise, and why it is not a brain type

Mechanics rewards students who want a single right answer and a picture they can believe in; statistics rewards students who tolerate some vagueness and will write prose in a maths exam. Those are different temperaments, and most sixteen-year-olds have more of one than the other.

The bigger driver is accident. Students taking Physics meet forces and constant acceleration a year earlier, arrive fluent, and conclude they are a mechanics person. Students who do not take Physics meet a force diagram for the first time in a maths lesson and conclude the opposite. Applied also tends to be taught at the ragged end of the scheme of work. None of that is aptitude — in ten years I have not met a student who could integrate by parts and genuinely could not resolve forces.

Applied is the cheapest place on the course to buy marks

The case for doing something this term rather than in April is that the applied content is small and finite. Mechanics at A-Level is quantities and units, kinematics, forces and Newton's laws, and moments — that is the entire list. Statistics is sampling, presenting and interpreting data, probability, distributions and hypothesis testing. The questions repeat in structure year after year, because there are only so many ways to hang a mass off a string or test a proportion.

So the highest-return revision I set is deliberately lopsided. For about four weeks the student does applied sections only — not whole papers, just the applied halves, working back through their own board's past paper archive. Fifteen or twenty in a month is realistic, and by then the repetition is obvious in a way it never is at one per paper.

Mark each one strictly and log the reason rather than the topic. "Forgot to resolve vertically" is a topic. "Started writing equations before I had drawn the diagram" is a reason, and it will be the same reason every time — the error-log discipline behind the most common A-Level Maths mistakes, paying back faster here than anywhere.

When writing it off is rational, and when the arithmetic says otherwise

I will be honest about the exception. If you are eight weeks out, targeting a C, and mechanics is genuinely at zero, hours shoring up pure will return more than hours starting mechanics from nothing. That is triage and it is defensible — but only on Edexcel's shape, where the damage is sealed inside one section, and only as a decision taken deliberately rather than by drift.

Near the top of the scale the arithmetic closes that door. This summer an A* on Edexcel took 254 marks out of 300 and an A took 210. Write off the 50-mark mechanics section and your ceiling is 250: the A* is gone at any level of brilliance elsewhere, and the A means taking 210 of the remaining 250 with no margin for a bad morning. For anyone aiming high, what separates an A* from an A is decided right here, in the part of the course nobody finds interesting.