Most exam-week plans are calendars of intentions — "Monday: chapter 4, Tuesday: chapter 5" — which is rereading with dates attached. The day says what you'll look at, not what you'll produce, so the week ends with everything covered and nothing tested. Coverage is not preparation; the exam doesn't ask what you saw, it asks what you can pull out.

The plan that works assigns stems, not chapters. Day 1 is a mixed diagnostic — a scored pass across everything, closed-book, that produces the repair list. Days 2–5 repair only the misses, interleaved so nothing stays warm from yesterday's session. Day 6 is a timed mixed set — exam conditions, the real test of whether the repairs held. Day 7 is two light retrieval loops and then sleep, because the night's consolidation is worth more than a panic reread.

Why chapter calendars fail

A calendar that lists chapters trains nothing — it schedules exposure, which is the weakest form of practice. The week of "read, reread, review" produces familiarity with the material and no evidence you can use it. The diagnostic version of the same week produces a map of what's actually broken and spends the remaining days fixing exactly that.

The week works because it's spacing plus retrieval — the two effects with the strongest evidence in the learning literature. Each day is a retrieval event on a spaced schedule: the diagnostic on day 1, the first repair pass on day 2, the re-test on day 4, the timed set on day 6. The misses get attempted three or four times across the week, with days between attempts, which is exactly the schedule that makes memory durable. A chapter calendar is rereading with dates; this is retrieval with dates.

The seven days, day by day

Day 1: diagnose. One timed 40-minute mixed set across every lecture — no notes, no warming up by rereading. Score it honestly and tag every miss by topic. The output is the repair list: the ranked set of things you can't do, which the next four days will work through. The diagnostic feels bad; that's the information arriving.

Days 2–5: repair the misses, interleaved. Each day takes a share of the repair list — not one topic per day, which re-blocks the week, but a mixed set across several flagged topics. Interleaving prevents the false fluency of doing the same topic until it feels easy. Every item gets attempted closed-book before notes are allowed, and every repaired item gets a re-test date later in the week.

Day 6: timed, mixed, everything. A full set under exam conditions — time pressure, no notes, all topics shuffled. This is the week in miniature: a diagnostic of whether the repairs held after spacing. The misses here go to day 7's light pass; the hits are done. If day 6 goes well, the exam is a repeat of a performance you've already given.

Day 7: light loops and sleep. Two short retrieval passes over the day-6 misses — 20 minutes each, morning and evening — then stop. No new material, no all-nighter; the marginal gain from late cramming is smaller than the cost of sleeping badly. Consolidation happens in sleep; day 7's job is to arrive rested with the material still warm.

A 40-minute session on the clock

The weak answer treats day 1 as the start of coverage — rereading everything, which feels productive and produces nothing measurable. The strong answer treats day 1 as the start of evidence: a diagnostic that shows what's broken before the week is spent fixing the wrong things. The refusal matters as much as the plan — "warm up" rereading is how the day becomes coverage again.

Why the shape matters more than the hours

The plan fails when the shape is violated — ten hours of heroic effort on day 1 followed by collapse, or a "review day" that's secretly a reread day. The seven days work because each has a defined output: the diagnostic list, the repaired items, the timed score. Days without outputs are days that drifted back to coverage; the plan's discipline is that every day produces something you can check.

This is also why sleep is on the schedule rather than implied. The retrieval across the week is only half the mechanism — the consolidation that stabilizes it happens at night. A week of perfect practice days and 4-hour nights retains less than a week of decent practice and real sleep; the schedule includes sleep because the science does.