Gyroscope Benchmarks / Mind
Time perception test: how accurate is your internal clock?
This test measures the direction, size, and consistency of your errors when estimating short intervals without a visible timer.
Why it matters
Your sense of how long a few seconds takes runs under almost everything you do without a clock in view. Holding a stretch, leaving a pause before you answer, pacing a hard set, deciding whether the pan has been on long enough. None of that uses a timer. It uses an internal estimate, and that estimate has a direction and a spread. Some people run long, some run short, and most have never checked which.
That internal clock is not fixed. It is one of the first things a bad night moves. Attention drifts, the silent count stumbles, and the same seven seconds comes out as five in one round and nine in the next. Stress and illness shift the felt pace of time too. The hard part is that you rarely notice it happening from the inside. In one lab study, two weeks of 6-hour nights produced cognitive deficits equal to two full nights without sleep, while the people involved rated themselves only mildly impaired.
Van Dongen et al., Sleep, 2003
That is why one number, measured the same way every month, says more than how you feel. Time Keeper gives you a bias, a spread and an error, scored against peers and against your own baseline. A month where your coefficient of variation doubles is a fact you can read, not a mood you have to guess at. Over a year, the line shows whether your timing holds steady, tightens with practice, or wobbles in the weeks you sleep badly.
What this test measures
Time Keeper gives you a target such as 3 or 7 seconds. With no visible countdown, you tap when you think the interval has passed. A run takes about 70 seconds and repeats the challenge across several targets.
Two measurements matter. Bias shows whether your estimates systematically run long or short. Variability shows whether repeated estimates cluster tightly even when the average is imperfect. Mean absolute percentage error summarizes accuracy across different target lengths, while coefficient of variation divides the standard deviation by the mean to describe relative consistency.
Coefficient of variation is a standard marker in interval-timing research. Lewis and Miall tested human estimates from 68 milliseconds to 16.7 minutes and used CV to examine whether relative precision stays constant as intervals change.
Lewis and Miall, Philosophical Transactions of the Royal Society B, 2009
Context matters too. In a reproduction experiment, estimates became more variable for longer intervals and were systematically pulled toward the mean of the interval distribution participants had experienced.
Do the test right now
- 1Open Time Keeper in Gyroscope and silence other timers or audible clocks.
- 2Read the target interval, such as 3 or 7 seconds, then start the trial.
- 3Count silently with no visible timer and tap once when you think the target duration has passed.
- 4Complete every round. Keep mean absolute percentage error, the direction of bias, and coefficient of variation rather than one unusually close estimate.
Reference bands
Interval-timing protocols vary, so Gyroscope uses practical bands for the app's own short targets. These are not clinical cutoffs.
| Mean absolute error | Practical band |
|---|---|
| Under 10% | Tight timing |
| 10–20% | Typical |
| Over 20% | Noisy |
Jazayeri and Shadlen, Nature Neuroscience, 2010. Bias and variability context for Gyroscope's practical bands.
For consistency, a coefficient of variation under 0.15–0.20 is a practical target. Read it alongside bias, because tightly repeated estimates can all be early or all be late.
| Coefficient of variation | Practical read |
|---|---|
| Under 0.15 | Consistent |
| 0.15–0.20 | Transition |
| Over 0.20 | Variable |
Lewis and Miall, Philosophical Transactions of the Royal Society B, 2009. CV method context for Gyroscope's practical bands.
What moves your score
Counting strategy and attention can shift both bias and consistency. A short-term sleep-deprivation meta-analysis found its largest effect in lapses of simple attention, Hedges's g -0.776 across 70 articles and 147 tests.
Lim and Dinges, Psychological Bulletin, 2010
- Fatigue can make silent counting less stable.
- Practice can tune a counting rhythm to the task.
- Stress or illness can change the felt pace of an interval.
- Recent target lengths can pull the next estimate toward their average.
How to improve
Two things move this number: how steady your silent count is, and how well you know your own bias. The first comes from repeating one cadence; the second comes from reading your results instead of guessing.
- Pick one cadence and keep it. Count at a fixed, unhurried pace in every trial. The pace matters less than never changing it; a steady count is what tightens the spread.
- Use the same count for every target. A 3-second target and a 7-second target share one rhythm; only the number you stop on changes. Speeding up for short targets and stretching for long ones is where much of the error comes from.
- Learn your direction, then adjust by one beat. If your results run long, you tap late; if short, you tap early. Shave or add a single beat, hold it for a full session, and only then decide whether to adjust again.
- Correct between sessions, not within a run. A tap that felt early is not a reason to drag out the next one. Treat every trial as a clean repeat and read the results afterward.
- Test at the same point in your day. Your felt pace of time shifts across the day and with how rested you are, so a different hour is a different test. Pick a slot when you are not rushing.
- Sit still with the phone out of reach. In one experiment, a phone on the desk, face down and silent, measurably reduced working memory and attention. Movement, glances and a buzzing pocket all split the attention the count depends on.
Ward et al., Journal of the Association for Consumer Research, 2017
Retest with the same protocol, rested, once a month: same targets, same cadence, same slot, other clocks silenced. Change one thing at a time between sessions, so that when the line moves you know why.
Track it
Log it in the app.
Download Gyroscope and run Time Keeper in about a minute. See how you compare to your peers, to the studies, and to your own baseline over time.
Download the appGet Gyroscope One