Four pieces worth 80% of the course — measure a queue, change who goes first, simulate the change, then redesign something real and defend it. The other 20% is the weekly quizzes.

Assessment#

Four pieces, worth 80% of the course between them, and each one can only be attempted with the tools the weeks before it handed over. That is deliberate: the assessment follows the same arc as the syllabus rather than sitting beside it.

You measure a real queue in week 3, having just learned how (10%). You change who goes first in week 6, having spent two weeks on what that costs (15%). You simulate the change in week 9 (15%). Then you spend the back half of semester on one design brief worth 40% of the course on its own.

The remaining 20% is not a fifth brief. It is a short quiz at the start of every lecture, running from week 1 — how that works, and how labs fit in, is on the arc page, next to the timeline it sits inside of.

Each brief says what good looks like, names the failure modes that account for most weak submissions, and explains how every marking criterion is read. They are longer than assessment briefs usually are, on the theory that a student who knows exactly what is being asked writes a better answer than one who has to guess, and that nothing in here is worth hiding.

Three things are worth knowing before you read them. Extra capacity is not an answer — the week 6 task forbids it, because it is the response that lets you avoid thinking about priority at all. Naming who is worse off carries real marks, because the mathematics makes it very easy to forget that anybody is. And the final brief is marked hardest on the alternative you rejected, argued at its strongest rather than its most convenient.

Late work, extensions and the rules on generative AI are on the policies page rather than repeated in each brief.