A revision schedule you will actually keep

Why the same hour a day before the exam is worth less

Memory fades fast, and the rate is not constant. Whatever you learn settles noticeably within a day, then more slowly, then flattens out. Revisiting a topic just before it fades costs almost nothing to learn and tops it back up. Revisiting it after it has gone costs much more.

That asymmetry is the whole argument for spacing, and it is why the advice sounds so strange: the harder a session feels at the time, the more efficient it is, provided you find the topic again slightly after the point where it started slipping.

A schedule that survives contact with a real week

Three passes over the same material, spread out, beats six passes crammed together. The exact intervals matter less than the shape, so this one is deliberately modest:

• Pass one, the day you finish the material. Get it in once while it is fresh, so you know what you are dealing with.
• Pass two, two to three days later. This is the session that does the work. Expect it to be harder than pass one and that is the point.
• Pass three, a week later. By now you should only be looking at the things you got wrong.
• After that, only revisit what is still wrong. A topic you have answered correctly three times across three separate weeks does not need a fourth.

If you only have one more session before an exam, do not split it evenly. Cover less material and cover it twice, with a break in between, rather than touching everything once.

When the schedule slips

It will slip, and the usual response is to give up on the schedule entirely and return to cramming. A better response is to shorten the material rather than compress the time.

Pick the reviewers you have touched least recently, take them in one sitting, and mark everything you got wrong for the next session. A reviewer you have already taken twice and still fail on is a signal, not a scheduling problem: either you do not understand it yet, or the reviewer is testing something the notes never covered. Both are worth a look before you queue it again.

Using attempt history to pick today's session

Hachi's History page records every attempt with the reviewer, the mode, the score, and the date. Read it as a to-do list rather than a scoreboard.

The useful signals, in order:

• Scored badly and long ago. The highest value session available, because the decay has had time to work.
• Scored badly recently. Read the explanations rather than re-quizzing. You do not need more attempts, you need the material you missed.
• Scored well long ago. Due for a top-up. One pass now is much cheaper than relearning it the night before.
• Never attempted. The only genuine blind spot in the list.

Streaks count days, not topics, so treat them as a nudge rather than a target. A day spent reviewing one weak section honestly is worth more than a day spent opening five reviewers and closing them.