What is covered, week by week, and which topics each lecture draws on.
One lecture runs in each of the twelve teaching weeks, delivered by the course convenor. Each lecture page below covers that week’s topic in more detail, and links to the tutorial session and slides (where a deck exists) it feeds into.
The digits are a claim you did not make
Why written precision overstates what is known, and why the first question in any estimate is what the decision needs
A number is a chain, not a guess
Why breaking a quantity into parts beats guessing it whole, and why defining the target is part of the estimate rather than a preliminary to it
The assumption you didn’t notice you made
Why an estimate’s fragility concentrates in a few inputs, why people audit the wrong ones, and why hindsight makes this look easier than it was
A range says more than a number
What a point estimate hides, why a national agency publishes its uncertainty on purpose, and how to choose a width honestly
Measured well, about something else
Why most evidence is borrowed, the three ways a proxy fails, and why a proxy relationship can decay without anybody noticing
Disagreement is information
Why splitting the difference destroys evidence, the small set of axes real disagreements sit on, and why nobody warns you at the moment you need it
- Teaching break5–18 April 2027 — no lecture or tutorial.
Revision, and its imitation
How pressure actually arrives, the test that separates an honest change from a concession, and why the defences have to be built early
What your estimates have in common
The single mechanism that makes combining estimates work, what happens when it is absent, and why agreement is not evidence
Accurate on average is not the same question
Why expected accuracy stops being the point once an estimate gets acted on, and the two questions the rest of the course has kept merged
The one thing forecasting adds
Why a claim about the future needs no new machinery beyond what you already have, and the one structural difference that changes everything
What prose cannot force you to say
Why a written justification can hide exactly the weakness a live question exposes in seconds, and what the Court is actually testing for
One miss tells you nothing. Fifteen might.
Why no single estimate can reveal whether you are a good estimator, and what an accumulated, precommitted record can show instead