The pomodoro technique is a trap for maths students. Try this instead.

Watch a student work a hard, multi-step problem and you can almost see the structure being built: what's given, what's wanted, which tools might bridge them, which dead end was just ruled out. That structure lives in working memory. It took twenty minutes to assemble. It is not saved anywhere. Now a timer goes off, a five-minute break begins, a phone appears — and the structure is gone. The next "work session" starts with ten minutes of rebuilding what the break demolished. The student concludes they're bad at maths. They're not. Their study system is optimised against the subject.

What the timer actually interrupts

Two findings from cognitive psychology explain the damage:

• Loaded context is expensive. Hard problem-solving holds a web of intermediate state in working memory — relationships, partial results, ruled-out paths. Unlike reading or flashcards, this state doesn't survive interruption; it must be rebuilt from scratch, and rebuilding costs far more than the break "recharged".
• Breaks leave attention residue. Research on task-switching (Sophie Leroy's work is the classic) shows attention doesn't transfer cleanly: part of your mind stays with whatever the break contained. A phone-shaped break maximises the residue, because the feed is engineered to be exactly what attention clings to.

Pomodoro was invented for work that is shallow-restartable — admin, drafting, studying flashcards — where progress is linear and any minute's work stands alone. Mathematics at exam-deciding depth is load-then-think work. The 25-minute slice systematically truncates thinking right at the point where the load is finally in place.

When pomodoro IS the right tool

This is not an anti-break article. Pomodoro remains excellent for maths work that is genuinely shallow-restartable: fluency drills (times tables, index laws, differentiation standards), flashcard retrieval (formulae, definitions, theorem statements), error-log admin, and marking your own paper. The skill isn't picking a side — it's matching the protocol to the work. That's the whole table:

The depth-first protocol

1. Enter through retrieval (5 minutes). Start every session with five questions from memory — yesterday's topic, last week's, last month's. This warms up the exact machinery deep work needs and doubles as spaced practice. No phone allowed past this gate.
2. One thread, 50–90 minutes. One topic or one paper section. The block ends at a milestone, not a timer: a problem solved, a method securely repeated, a paper section completed. If you must stop mid-problem, that's fine occasionally — see step 4.
3. Break like an athlete, not a scroller (10 minutes). The break's job is recovery without residue: move, water, stare out of a window. The phone is not a break; it's a second cognitive task. (This single substitution is the largest improvement most students ever make to their studying.)
4. Exit through a bridge note (3 minutes). Before closing the books, write two lines: where I am, what I try next. Hemingway famously stopped mid-flow on purpose — the note makes tomorrow's load-in cost minutes instead of half a session.

Run the two-week experiment

Don't take the article's word for it — this is testable on yourself. For two weeks, run problem-solving sessions as depth-first blocks (retrieval entry → 50–90 min thread → movement break → bridge note) and keep pomodoro for drills. Track one number: long questions completed per week. Then compare against your previous fortnight. Every student I've run this with has kept the protocol — not because I argued for it, but because their own number did.