Start and stop a hidden timer to hit target durations from 3 to 13 seconds. Counting in your head is allowed - this measures how accurately you keep time.
What this test measures
Each trial names a target duration. You tap to start a timer you cannot see and
tap again when you believe that much time has passed. Count it out however works
for you.
Your score is the mean absolute error in milliseconds across six targets
between 3 and 13 seconds. The test also records your signed bias: negative
means you tend to stop short, positive means you overshoot.
How to get a clean score
- Pick a counting method and stick with it for all six trials.
- Keep a steady pace - rushing your count near the end is the most common source
of error.
- Expect longer targets to be harder. The scoring accounts for it.
How to improve your score
The reliable gains come from consistency, not from a better sense of time. Fix
on one counting method - a sub-vocal “one-thousand-one” cadence works for most
people - and hold the tempo steady rather than accelerating as you near the
target. If you know your bias from past runs, deliberately shade the other way:
someone who always stops 400 ms early can just hold on a beat longer. Do the
test rested; fatigue and distraction both widen the error.
What the number tells you
Keeping time over several seconds combines a sense of duration with a steady
internal rhythm and the attention to maintain it. Your consistency and your bias
are usually more stable, and more revealing, than any single trial. Stimulants
tend to speed the internal pace so people stop early; fatigue and divided
attention widen the error.
FAQ
- Am I allowed to count?
- Yes. Counting in your head, tapping a beat, using "one-thousand-one" - all fine. This is a test of how accurately you can track an interval by whatever method you use, not a test of a pure internal clock.
- What counts as a good result?
- With counting allowed, most people get their mean absolute error under about 600 ms. Under 300 ms is sharp. The test also reports your bias - whether you habitually stop early or overshoot.
- Why do the targets vary in length?
- Error tends to grow with the interval, so a 13-second target will usually have a larger absolute miss than a 3-second one. Averaging across a spread of targets gives a fairer overall number.
- What does my bias mean?
- Bias is the average direction of your misses. A consistent early bias often means your internal pace is running fast - common with caffeine or high arousal. A late bias means the opposite. Bias is usually more stable across sessions than raw error, and easier to correct for once you know it.
Reference sample: SYNTHETIC. Lognormal, mean 500 ms / sd 360 ms for mean absolute error across interval-production trials (3-13 s targets), counting ALLOWED. Median player ~405 ms. Anchored to mental-counting accuracy of roughly 4-6% of the interval plus response latency, with a heavier upper tail for low-effort / distracted players (who do not count carefully).