Study Timer

You have a chapter to get through and a phone that keeps asking for attention. A timer that splits the evening into fixed stretches of work and rest decides one thing for you: when to stop. Here is how the cycle adds up, what the evidence on fixed breaks shows, and what the timer does when you leave its tab or close the lid.

The default cycle, step by step

The timer alternates 25-minute focus sessions with 5-minute short breaks until four focus sessions are done. The fourth is followed by a 15-minute long break, and the cycle starts again. This 25-minute pattern was popularised by the Pomodoro Technique, a trademark of Francesco Cirillo; this site is not affiliated with him or endorsed by him.

The long break takes the place of the last short one. For a long break after every n focus sessions:

cycle = n × focus + (n − 1) × short + long

focus share = n × focus ÷ cycle

The default cycle takes 130 minutes, and 100 of them are focus: 76.9%.

What the research says

Two peer-reviewed studies bear on fixed breaks, and both have limits worth stating.

Fixed breaks against self-chosen ones. Biwer and colleagues (2023) followed 87 bachelor's and master's students at a Dutch university through one day of their own self-study. Some chose their own breaks; the others took either a 6-minute break after every 24-minute block or a 3-minute break after every 12-minute block. Those choosing their own breaks studied and rested in longer stretches, which went with more fatigue and distraction and less concentration and motivation than in the fixed groups. There was no difference between the groups in mental effort or in how much of the task got done. The authors conclude that systematic breaks "had mood benefits and appeared to have efficiency benefits" compared with self-chosen ones.

Brief switches during a long task. Ariga and Lleras (2011) had observers do a visual vigilance task while holding digits in memory. When the digits were recalled only at the end, performance fell steeply over time. When observers recalled them now and then during the task, the decline was averted. That was a laboratory task, not revision, and the switch was to another task, not rest.

Neither study tested a 25-minute session, and neither says anything about health. Both fixed schedules in the first study give the same share: 24 minutes in every 30, or 12 in every 15, is 48 minutes of study per hour.

Choosing other lengths

Focus, short break and long break each accept 0.05 to 240 minutes, with decimals: 0.5 is 30 seconds and 0.05 is 3 seconds. The long break can come after 1 to 12 focus sessions. A value outside those limits shows an error and the last good value stays in use.

The last two rows are the 2023 study's schedules, which had no long break. To copy a schedule with no long break, set the long break to the same length as the short one.

A change made while a session is running or paused leaves that session's length alone; the new length applies from the next one. A session not yet started takes it straight away.

Planning a study day

To reach a target amount of focus, work out the sessions first:

sessions = target focus minutes ÷ focus length, rounded up

Four hours of focus is 240 minutes, or 9.6 sessions of 25 minutes, so 10 sessions. With two long breaks and seven short ones between them, the last session ends 315 minutes after the first Start: begin at 9:00 am and you finish at 2:15 pm, if each session starts as soon as its break ends.

With auto-start on, each new session is timed from the moment the last one really ended, not from when the page noticed, so a late check never pushes the day back. For a single countdown with no cycle, the countdown timer is simpler.

What happens in a background tab

The timer never counts ticks. It notes when a session started and how long was left, and every check works out the time left from the clock: four times a second while running, and again the moment you return to the tab. In testing, 3-second sessions changed over within about 0.2 seconds of schedule.

Browsers check a hidden page's timers far less often. Chrome checks them only once a minute in a page hidden for more than 5 minutes, under conditions Google documents. So the digits, the countdown in the tab title and the tally can lag by up to about a minute. The times they show are exact, and the page catches up as soon as it is checked.

The alarm does not wait for those checks. It is booked on the browser's audio clock when the session starts, so it can sound on time in a background tab. If the browser suspends audio, as a phone browser in the background may, it plays once when the page runs again.

Sleep, lids and clock changes

Every check reads two clocks: one that keeps running through sleep but moves if the system time is changed, and one the system time cannot move.

time gone = the larger of the two clocks' changes

A laptop closed for 20 minutes shows 20 minutes gone. A clock wound back an hour adds nothing.

An example with default lengths and auto-start on. You press Start at 10:00, so focus ends at 10:25, and you close the lid at 10:20.

• Open it at 10:28. The page noticed more than 2 minutes late, so it says the session ended at 10:25. The break is timed from 10:25, so 2 minutes are left.
• Open it at 10:40. The break would have ended at 10:30, so it is left stopped, with a note, rather than counted as taken.

Only one session ends per wake. One case cannot be told apart from sleep: the system clock jumping forward while the page is open. A session then ends early by the size of the jump.

The daily tally

The two figures under the timer measure different things.

Finish three default sessions and skip the fourth after 10 minutes 30 seconds. Sessions today reads 3, and Focus today holds 85 minutes 30 seconds, shown as 1 h 25 min in whole minutes rounded down. The skip still moved the cycle on, so the long break is next.

The tally measures the timer, not your attention: a focus session that ended while the laptop slept still counts. Focus minutes are saved when you pause, reset, skip or finish, and when the tab is hidden or closed. Both figures restart at midnight on your device, and the clear button empties them sooner. The tally lives in this browser only; a private window forgets it.

Sound, and why you might not hear it

Browsers do not let a page make sound until you have pressed something on it. Chrome's autoplay policy, for example, puts Web Audio created before any press into a suspended state. The timer therefore sets up its alarm on your first press of Start or Test sound.

The alarm is made in the browser, with no sound file: three soft notes played twice, falling into a break and rising back to focus, about 4.6 seconds in all by the timings in its code. The volume slider is applied on a curve:

gain = (slider ÷ 100)²

The default of 60 gives 0.36 of full output, and halfway at 50 gives a quarter.

If you hear nothing, check the mute box, the slider and your device's volume, then press Test sound. On an iPhone in silent mode, sound from web pages may be muted.

Keeping the screen on, full screen and keys

Where the browser supports the Screen Wake Lock API, an option to keep the screen on appears, switched on by default. The timer holds the lock while a session runs and lets go when you pause, reset or leave. Under the W3C specification only a visible page can hold the lock, and it is released when the page is hidden, so the timer asks again when you return. The browser can refuse or drop the lock, for example in a power-saving mode, and the page tells you when it does.

Full screen uses the browser's own full-screen mode. Where that is unavailable or refused, as on iPhone Safari, the timer fills the window instead, with the browser's toolbar still visible. Esc closes it.

Space starts or pauses, R resets and S skips. They do nothing while you type in a box or hold Ctrl, Cmd or Alt. To time something open-ended instead, use the stopwatch.

What this tool assumes

• Lengths from 0.05 to 240 minutes; defaults of 25, 5 and 15 with a long break after 4.
• The session in progress is not saved. Reloading or closing the tab loses it; settings and today's tally are kept.
• Time comes from the device's clocks, so a forward jump of the system clock ends a session early.
• If the device sleeps past the end of one session and the whole of the next, that next one is left ready rather than counted.
• Settings and the tally stay in this browser. Nothing is sent anywhere.
• The timer divides time. It makes no claim about grades, memory or health.