What fatigue actually takes away
Tiredness does not lower your intelligence evenly, like a dimmer switch. It attacks specific functions first, and the order is consistent enough to plan around. Sustained attention goes first — you can still read, you just stop noticing when you have stopped reading. Working memory goes next, which is why a three-step derivation becomes impossible while a one-step one is fine. Inhibition goes third, which is why your phone wins at 1am and loses at 10am.
What survives longest is recognition and well-practised retrieval. You can still answer a flashcard. You can still tell whether an answer is right. You can still recognise the shape of a problem you have seen before. Those are the operations to build a tired session out of.
The tired-hour hierarchy
| Task | Survives fatigue? | Do it tired? |
|---|---|---|
| Flashcard review | Yes — retrieval of near-known items is robust | Yes. This is the best possible tired task. |
| Redoing problems you've already solved | Mostly — the method is already in memory | Yes, and it's underrated. |
| Marking your own past answers against a mark scheme | Mostly — recognition, not generation | Yes. |
| Tidying, filing and tagging notes | Yes — near-zero cognitive load | Yes, but be honest that it isn't studying. |
| Reading a new chapter | No — encoding needs attention you don't have | No. You'll re-read it tomorrow anyway. |
| Multi-step problem sets you haven't seen | No — working memory is the bottleneck | No. You'll fail them and learn helplessness instead. |
| Essay planning and argument structure | No — needs held-in-mind comparison | No. |
The pattern: anything where the material is already in your head is fair game. Anything where it has to get in is not. That single rule sorts almost every task you own.
A twenty-minute tired session that actually works
- 1
Set a hard stop before you start
Twenty or thirty minutes, on a timer, out loud. Tired sessions have no natural end — you drift until you feel guilty enough to stop, which teaches you that studying is an unpleasant thing that goes on forever. A stated stop is the difference between a short win and a long defeat.
- 2
Do due reviews only — nothing new
Open your review queue and clear what's due. No new cards, no new chapters. If you use a real scheduling algorithm, the queue already knows what's about to fall out of memory, so the tired hour spends itself on exactly the highest-value items.
- 3
When a card fails, don't investigate — flag it
The tired-brain trap is failing one card, opening the textbook, and losing forty minutes to a rabbit hole you won't remember. Mark it, move on, deal with it in the morning when you can actually encode the fix.
- 4
Finish with one deliberately easy thing
Re-solve a problem you already know how to solve. You want the session to end with a success, because the emotional memory of last night's session is most of what decides whether you sit down tomorrow.
- 5
Write tomorrow's first task on paper before bed
One line. The single hardest thing you're avoiding. Tired-you is a bad executor but a fine planner, and it removes the twenty-minute morning dither about where to start.
The all-nighter question, answered honestly
If the exam is tomorrow morning and you genuinely have not covered the material, staying up will get you more marks than sleeping will. That is true and pretending otherwise is dishonest. But it is a loan, not income: you are borrowing from the consolidation that sleep performs on everything you studied this week, and from the exam-hall attention you will need to read the questions properly. The trade is only worth it when the alternative is arriving with nothing at all.
If you are more than one day out, the trade is never worth it. Sleep is when the day's learning is actually filed — see sleep and memory for the mechanism, and all-nighters for what they cost you the day after.
The trap: mistaking motion for work
Tired studying is where highlighting, re-reading and "organising my notes" thrive, because all three feel like effort and none of them demand any. You will finish two hours of it convinced you worked, and remember nothing — the same illusion of fluency described in why highlighting doesn't work.
- Re-reading a chapter you already read — feels productive, produces recognition you'll mistake for knowledge in the exam hall.
- Watching a lecture recording at 2x with no output — passive video is the lowest-yield thing you can do at any energy level, and worst when tired.
- Making brand-new notes from scratch — this is encoding work, and encoding is exactly what fatigue takes.
- Colour-coding anything — the most seductive fake work there is.
When tired means something else
Tired every single evening for weeks is not a study-technique problem and no scheduling trick will fix it. Chronic exhaustion, dread before opening a book, and the sense that nothing you do makes a dent are the recognisable shape of student burnout, and the intervention there is rest and load reduction, not a better flashcard routine.
The honest test: after a genuinely free weekend, does the tiredness lift? If yes, you are tired. If no, treat it as the separate problem it is.