The Pomodoro study method: a plan for a real revision week
You sat down at 19:00 with three chapters and a highlighter. It is now 22:10.
The pages are beautifully yellow. Whole paragraphs of yellow. Some of the yellow has other yellow on top of it.
Asked what chapter two said, you can produce roughly one sentence, and you are not certain it came from chapter two.
Reading feels like studying while it is happening. That feeling is the least reliable signal in the business, and almost every bad revision week is built on trusting it.
The rule: if you never closed the book, you did not study — you read
A study session has to contain a moment where the material is not in front of you and you have to produce it anyway. Three minutes is enough. Zero minutes is not a session, whatever the clock says.
The timer’s job is to make sure that moment arrives, because when you are behind it is the first thing you drop.
Why three straight hours does not stick
Two findings here, and both are about as solid as education research gets.
Retrieval beats rereading. Roediger and Karpicke (2006, Psychological Science) had students either reread a passage or test themselves on it. Five minutes later, the rereaders won. A week later, the self-testers won by a wide margin — and the rereaders had confidently predicted the opposite. Lab passages and college students, not a whole syllabus, but the pattern has replicated widely: rereading flatters you now and charges you later.
Spacing beats cramming. The same total minutes across separate days beats one block. Cepeda and colleagues (2006, Psychological Bulletin) pooled hundreds of experiments and found the advantage grows as the gap widens, up to a point that scales with how long you need to hold the material.
Dunlosky and colleagues (2013, Psychological Science in the Public Interest) then ranked ten common techniques by the evidence behind them. Practice testing and distributed practice came out high utility. Highlighting and rereading came out low — two of the most popular techniques among students, and two of the least useful.
So a three-hour block is not too long for your brain. It is bad because it is three hours of one topic, in one sitting, with the book open throughout. Change those three things and three hours is fine.
One session, set properly
- One sentence, naming an output. Not “revise chapter 4” but “write the four steps of glycolysis from memory.” If the sentence needs the word and, you have written two sessions.
- Set up before the clock starts. Notes out, tabs closed, phone in another room. Do not spend six minutes of a 25-minute session finding a pen.
- Set the timer. 25 if starting is hard, 50 if you are already warm. When in doubt take the shorter one — the point of twenty-five minutes is that it is cheap to agree to.
- Scrap of paper beside you. Every stray thought goes on it. Text Mai back. Print the form. Three seconds to write, and the urge to act on it collapses.
- Last three minutes: close everything. Blank page, no notes, no peeking. Write what you can produce.
- Then break. Off the screen, out of the chair.
Step five turns reading into studying, and step five is the one everybody drops when they are behind. Drop step one instead.
How many sessions can I do in a day?
There is no research number for this, and anyone quoting one is guessing at you with confidence.
As a planning heuristic: plan four, allow eight, and judge the day by whether you could repeat it tomorrow. The last two sessions of a long day are reliably worse than the first two, and a schedule you can hold for nine days beats one heroic Sunday followed by nothing — which is the whole argument behind keeping a streak small enough to survive.
Change the shape by material
The 25/5 rhythm is a default, not a finding. Different material wants different blocks.
| Material | Block | Break | What changes |
|---|---|---|---|
| Problem sets — maths, physics, chemistry | 50 min | 10 min | Mix problem types instead of doing twenty of the same. Interleaving feels worse while you do it and tests better later (Rohrer and Taylor, 2007) |
| Memorisation — vocabulary, anatomy, dates, cases | 20 min | 5 min | Short and frequent. Three 20-minute sessions across three days beats one hour tonight, and it is not close |
| Dense reading | 25 min | 5 min | One section per session, then close the book and write three sentences from memory |
| Essay drafting | 50 min | 10 min | Stop mid-paragraph. Restarting is cheap when a half-finished sentence is waiting |
| Past papers | Full paper | After | No timer method here. Sit the whole thing at exam length — the stamina is part of what you are training |
That last row matters more than the other four. A past paper broken into pomodoros measures your knowledge and says nothing about whether you can hold your head together for three hours, which is what actually goes wrong in the hall.
Breaks that make the next session start
The break has one job: make the next session easy to begin. Judge every break by that and the rules write themselves.
Five minutes means stand up, water, look at something more than two metres away, walk into a different room. A feed rests your legs and nothing else, and makes the next start harder — the opposite of the job.
Where does active recall fit?
At the end of the session, not in the break. Three minutes, closed book, blank page, still on the clock. Putting it in the break looks efficient and quietly kills the system: the break stops being a break, and within a week you stop taking breaks at all.
The one recall move that does belong in the break is verbal. On the walk to the kitchen, say out loud what the last session was about, in ordinary words, as if to someone who does not do your subject. Thirty seconds. If you cannot, you have just written the goal sentence for the next session.
After four rounds take 20 to 30 minutes and eat something. That long break is where the day either continues or quietly ends, and deciding which in advance saves an hour of guilt.
A seven-day plan you can copy
Three topics, one exam, seven days. Adjust the topic count, keep the shape.
| Day | Morning · 2 × 50 | Afternoon · 3 × 25 | Evening · 1 × 25 |
|---|---|---|---|
| Mon | Topic A, first pass, closed-book summary each session | A, mixed problems | Blank-page recall: A |
| Tue | Topic B, first pass | B, mixed problems | Recall: A + B |
| Wed | Topic C, first pass | C, mixed problems | Recall: B + C |
| Thu | Past paper, full length, timed | Mark it honestly, list every gap | Recall: A |
| Fri | Thursday’s gaps only | Mixed problems, A + B + C together | Recall: C |
| Sat | Second past paper, timed | Mark, list gaps | Off |
| Sun | Gaps only, 3 × 25 | Off | Off |
Four rules make this work rather than merely look tidy.
Every topic gets touched on at least three separate days. That is the spacing, and the only non-negotiable line in the table. If you must cut something, cut a morning, never a day.
Thursday’s paper is early on purpose. A past paper is a diagnostic, and a diagnostic on Saturday night is a horoscope — it names what you got wrong at the exact moment you can no longer fix it.
Evening recall is closed-book and short. Ten to twenty minutes, no notes on the desk, then check and mark your own blanks. This block quietly does most of the week’s work.
Sunday is mostly off. A rested Monday beats a fourth pass at topic C, and if you disagree you have never sat an exam on four hours of sleep.
When Pomodoro is the wrong tool for studying
Timed papers, as above. Sit them whole.
Lab sessions, group work, anything on someone else’s clock. You do not control the interval, so stop pretending you do.
Problems with a long load-in. If a proof takes twenty minutes to get into your head, a bell at minute 25 arrives just as you became useful. Move to 50 or 90.
When the system becomes the subject. Colour-coding a tracker, rebuilding the schedule, counting sessions to two decimal places — procrastination with excellent handwriting. If you have spent more than ten minutes this week on the plan, close it and open a book.
And if 25 minutes has never fitted you, that is not a failure to follow the method. Some people work best at 10, some at 45. The number was always meant to be adjusted, and the case for the technique rests on the bounded promise rather than any particular figure.
One more, for anyone who studies alone and cannot start at all: that is often not a discipline problem but an absence-of-other-people problem, and studying alongside someone fixes more of it than another timer will.
Tonight
Write one sentence naming what you will be able to do at 19:25. Put the phone in the hall. Set 25 minutes.
At 19:22, close everything and write what you remember on a blank page. Those last three minutes are the session. The other twenty-two were preparation for them.
Fomodo measures the daily goal in minutes rather than sessions, so a day of three long blocks and a day of nine short ones can both come out as a good day, which is closer to how revision actually goes. The limitation is the obvious one and worth saying anyway: it will not test you. No question bank, no flashcards, and no way for the app to know whether you filled the blank page or stared at it.
What it does is make the record hard to argue with. The heatmap darkens as the minutes go in, which is a slightly childish thing to want from software, and people want it anyway — reliably including the ones who announced they did not care.
Three chapters is not a plan. One sentence, twenty-five minutes and a blank page at the end is.
