Time Duration Game
Watch a dot, then say how many seconds it lasted. No counting aloud.
About Duration
Duration is the simplest question you can ask about your inner clock. A coloured dot appears on the screen, stays for a short while and vanishes. You then slide to the number of seconds you think it was visible. There is no timer on the screen during the display and nothing to count except your own sense of time.
The dot stays visible for a randomly chosen time. On the first levels that is between one and four seconds, a range in which most people do well. On later levels it can be as short as 0.4 seconds or as long as ten, which stretches the skill in both directions. The slider is logarithmic, so you can set a fraction of a second as precisely as a number of whole seconds.
How to play
- Watch the dot appear. Try not to count in your head with 'one-Mississippi' rhythm if you want to test your raw feeling; counting works well, but it measures your counting more than your intuition.
- When the dot disappears, a slider appears. Move it to the duration you think you saw. The scale runs from 0.2 to 20 seconds.
- Lock in before the 12 second answer timer ends. The true duration, measured to a hundredth of a second, is shown on the slider.
- Check the feedback to see whether you tend to guess too long or too short; the Stats page keeps that tendency for you.
A coloured dot stays on screen for a while, then disappears. Slide to how many seconds it was visible. No counting aloud and no stopwatch! The slider is logarithmic.
How scoring works
The score depends on the ratio between your answer and the real duration. A perfect answer scores 100 and the score drops linearly to 0 at a factor of about 2.2 on level 1 (twice as long or half as long still earns about 15 points) and 1.5 on level 10. Because the scale is a ratio, being 20 percent off costs the same at 1 second as at 8 seconds.
The real duration is measured with the browser's frame clock while the dot is on screen, so it is the actual time the dot was visible, not just the planned one. On slow devices a frame or two of error is possible, about a hundredth of a second, which does not change the score in a noticeable way.
What it trains
Duration trains interval timing, the ability to judge how long something lasts. We use it constantly without noticing: how long a traffic light has been red, how long we have been talking, when the pasta is done, or how long a pause in a conversation feels. It is one of the oldest topics in experimental psychology.
Playing regularly also teaches you about your own tendencies. Many people overestimate short intervals and underestimate long ones, and the bias plot on the Stats page shows where your own balance lies.
Tips for a higher score
- Use the sense of a beat that is already inside you: your resting pulse is about one beat per second, which gives a rough clock for the one to five second range.
- If you decide to count, count steadily and in a fixed rhythm such as 'one thousand one, one thousand two'. It will be consistent, but not perfect, since people speak faster when excited.
- For very short dots, under about half a second, do not count. Go by the feeling of a blink, which is roughly 0.3 to 0.4 seconds.
- Avoid anchoring on your previous answer. Each round is independent and the duration is drawn freely from the range for your level.
- Look at the bias feedback after a run. If you are always a bit short, push your answers up by 15 percent until the feeling is recalibrated.
The idea behind it
The idea that the error in judging time grows in proportion to the time itself goes back to Weber's law and, for time specifically, to scalar timing theory developed by John Gibbon and colleagues in the 1970s and 1980s. In animals and humans alike, the spread of responses around the correct value is roughly proportional to the interval, which is why the game scores ratios and not differences.
A second well-known finding is Vierordt's law, described by the German physiologist Karl von Vierordt in 1868: short durations tend to be overestimated and long ones underestimated, so that estimates are pulled toward a typical middle value. Attention matters too. Time seems to pass more slowly when you are watching the clock and faster when you are absorbed in a task.
Frequently asked questions
Is counting cheating?
No, but it changes what is measured. Counting uses your sense of rhythm in speech; not counting uses the inner clock. Both are good skills, and you can try each and compare your scores.
Why is the slider logarithmic?
Because the error in time judgement is proportional to the duration. A log slider gives you the same precision at 0.5 seconds as at 10 seconds.
Related games
Free, no account. Your best scores are saved on your device.