Go / No-Go Test: Tap or Hold Back

How to play the go/no-go test

The go/no-go test is a classic psychology-lab task, and it plays like Simon Says: act on one signal, stay frozen on the other.

1. Tap Start and wait for the countdown. Pictures of Coach Mochi then appear one at a time in the middle of a big tap pad. A legend under the pad reminds you which is which.
2. GO is Mochi lifting a barbell on a round green plate, tagged GO. Tap anywhere on the pad as fast as you can.
3. NO-GO is Mochi asleep on a red stop-sign octagon, tagged with a stop hand and SHH. Do nothing. Keep your thumb still and let him sleep.
4. Each picture stays up for 0.6 seconds, and a tap still counts until 0.9 seconds after it appeared. Then comes a blank gap of 0.7 to 1.2 seconds before the next one.
5. Watch the feedback: a green check and a cheering Mochi for a hit, a red ✕ and a startled Mochi if you tapped a No-Go (plus a short buzz when Vibration is on), and a gray Missed if a Go slipped by. From 3 in a row, a star counter shows how many right calls you’ve strung together, taps and holds alike.

There are two modes, and every run lasts 30 seconds:

• Test mode has fixed rules. One picture in four is No-Go. Use it for your personal record.
• Training mode has eight levels. As you climb, pictures show for less time (0.7 down to 0.4 seconds), the gaps get shorter, and the No-Go share rises from 25% to 40%. You move up a level when you catch at least 90% of Go pictures and tap no more than 10% of No-Go ones. You drop a level if false alarms reach 30% or hits fall below 70%.

What it measures

The main skill is response inhibition: stopping an action your brain has already started to prepare. Because three out of four pictures are Go, tapping turns into a habit within seconds. The No-Go picture asks you to cancel that habit on the spot. A second skill rides along: sustained attention, keeping your focus steady for the full set instead of drifting into autopilot.

Researchers use go/no-go tasks, and close relatives such as the SART (Sustained Attention to Response Task), to study exactly these two skills. In one of the largest studies, 10,430 people took a go/no-go-style attention test. The ability to stay on task peaked in the early 40s, while people’s strategy became steadily more cautious with age (Fortenbaugh et al., 2015).

Some honest limits:

• The No-Go share sets the difficulty. When No-Go is rare, people make far more mistakes than when it shows up half the time. Numbers from different versions can’t be compared.
• One set is a small sample. Judge your trend over a week, not a single run.
• Phones add delay. Touchscreen web apps register taps roughly 58–70 ms late (Pronk et al., 2020). That lowers the speed part of your score, not your accuracy.
• It is a game, not a diagnosis. A higher score means you got better at this task. Research on brain training finds that practiced tasks improve, while evidence that the gains carry over to everyday life is thin (Simons et al., 2016).

Averages and ranges

Published go/no-go numbers exist, but only for other versions of the task:

Our setup sits between those two, so we don’t show a percentile. Peer comparison may come later; for now, the target is your own record.

Here is how the score works, so you know what a good session looks like. The result screen shows one number, your control score in points:

Score = 100 × (hit rate − false-alarm rate) × speed factor. The speed factor is 1.3 minus your median Go reaction time in seconds, kept between 0.6 and 1.2. The score never drops below 0.

The lesson: accuracy moves the score more than speed. Being 50 ms slower costs about 5 points, but a set only has a handful of No-Go pictures, so a single false alarm can cost 10 points or more. Phone delay matters too: 60 ms of extra touch latency is about 6 points on a clean run, so compare records from the same phone.