The return flight: why reviewing today beats relearning

The most expensive hours of a semester are the ones spent learning something for the second time that you already knew once. Nothing new is produced in them. They restore a position you had reached and then lost, they cost as much as the first attempt, and often more, because the second time comes with the small humiliation of reading your own handwriting and not understanding it.

A short return flight on the same day prevents most of that. It takes minutes rather than hours, and it is the cheapest transaction available in studying.

What forgetting actually does

Forgetting is not a character flaw and it is not random. Something learned this morning fades quickly at first and then much more slowly. The steep part is over within a day or two. After that the loss continues, but gently, which is why a lecture from three weeks ago feels about as vague as one from four weeks ago, while the difference between yesterday and the day before is enormous.

The second half of the mechanism is the part worth knowing. Every renewed encounter with the material slows the decay that follows it. The memory does not become permanent, but the next drop is flatter than the last one. Work on how memory behaves over time points consistently in one direction: material met again after a delay lasts longer than material met again immediately.

The practical consequence is slightly counterintuitive. The right moment to revisit something is not when it feels gone. It is earlier, while it is still faintly warm, when the visit is quick and the answer still arrives with a bit of effort.

The five minute return flight

At the end of a day in which you learned something, take five minutes with nothing in front of you. No script, no notes, no slides, no laptop. Ask what happened today. Which topic, which three or four claims, how the argument went, what the exercise was actually testing.

The absence of material is the whole point. With the notes open, the exercise turns into reading, and reading is comfortable and almost free of information. Without them you either produce the content or you find a gap, and the gap is what you came for. A gap found on Tuesday evening costs a minute of looking something up. The same gap found in the exam costs the question.

It does not have to be written down. Saying it out loud on the way to the kitchen works. If it is written, keep it to four or five lines on the back of a page, because the moment it becomes a document it becomes a task, and tasks get postponed.

The second return flight

One pass on the same day is not enough by itself, but it changes what every later pass costs. Two or three days afterwards, do the same thing again: five minutes, still without materials.

The second attempt feels worse, and that is the correct signal. Some of it will not come. You will remember that there were three conditions and produce two of them. Most people take this as evidence that the first review was wasted, which is exactly backwards. Retrieving something that has started to fade is what makes it stay. Retrieving something you read four minutes ago does almost nothing.

Only after that second attempt is it worth opening the notes. At that point you are not rereading, you are checking one specific answer, and the correction lands on a question you have already asked yourself. That is a completely different experience from reading a chapter you have read twice before.

Why the gaps should grow

After the first few passes the intervals should get longer, not shorter. A week, then two, then whenever the subject shows up in a tutorial or a problem set anyway. The reason is simple. The visit is only worth making while the answer is uncertain. If it arrives instantly and without effort, nothing was learned and the time went into confirming something that was never at risk.

This also explains a familiar illusion. A subject reviewed at short intervals feels efficient and leaves very little behind. Everything comes fast, nothing is difficult, and the pleasant sense of mastery is a product of the short gap rather than of the knowledge.

A rhythm that holds up across a semester: the same day, two or three days later, a week later, then a couple of weeks later. None of this needs to be precise. Being three days late does no damage. It is a direction, not an obligation to be defended.

Fitting it into a day that already exists

The usual reason same day review does not happen is that it gets treated as another item on a list which is already too long. At half past ten in the evening a five minute task competes with sleep, and it loses, even at five minutes.

The way around this is to attach it to something that happens anyway, without a decision. After dinner, while the plates are still on the table. On the bus back from the library, where a script cannot sensibly be opened. On the walk from the lecture hall to the station. The trigger is an event you already have, not a time you have to remember and not a reminder you will swipe away.

Two things break this more reliably than laziness does. The first is doing it with the material in view, which quietly turns it back into reading. The second is scale. Someone who decides to review three subjects every evening has built a thirty minute obligation and will abandon it inside a week. One subject, the one from today, is enough.

What review cannot repair

Repetition cannot repair something that was never understood. If a derivation was opaque the first time through, recalling it after three days recalls the confusion, and after two weeks it recalls a slightly better established confusion. Spacing preserves whatever is there and has no opinion about whether it is correct. When the same review keeps producing the same blank, that is not a signal to review more often. It is a signal to go back to the source, to the tutorial, to somebody who can explain the step that never landed.

The second limit is administrative. Card systems are the natural tool here and they have a well known failure mode: the collection becomes the project. Making cards is pleasant, sorting them is pleasant, adjusting the settings is pleasant, and a deck of nine hundred cards with a daily queue of two hundred is not studying. It is maintenance with a progress bar. If more time goes into producing and organising the material than into answering it, the tool has taken over. A single page with ten questions in your own handwriting will beat an elaborate system you have quietly begun to avoid.

None of this is difficult, which is the strange part. Five minutes on the day, five more three days later, and the expensive month of relearning never has to be booked.