Perceptle Count practice
Count the dots.
Watch for two seconds, then enter your estimate.
Independent normal-mode practice · Best saved on this device
Get a feel for the whole picture.
- Press Start. After a short countdown, a pattern of 45–85 dots appears for two seconds.
- When the dots disappear, enter a whole-number estimate. Submit with the button or Enter.
- See the actual count and your difference in dots and percent. Try again for a new pattern.
A useful way to improve
Estimate small groups rather than chasing every dot. Notice a cluster of roughly five or ten, then use that cluster as a visual unit across the rest of the field. Check whether you tend to overestimate or underestimate.
What the score means
The error is the absolute difference divided by the true count. The normal tolerance is 0.20. Being 10 dots away when there were 50 is a 20% relative error and scores about 61 / 100.
Every score is out of 100. A fresh target appears each round, and your best for this exercise stays in your browser. Read the score curve and grade thresholds for worked examples.
Touch and keyboard controls
A numeric answer works with touch, mouse, or keyboard. The exercise is visual by nature; the true count and error are also provided as readable text after submission.
If you switch apps or hide the page during an active round, it stops without recording a score. Resize the window freely: targets are drawn in the same logical field and saved results remain intact.
Teaching examples · not today’s puzzle or an official answer
Count example: five groups are easier than fifty dots.
Instead of counting one dot at a time, find a group whose size you can recognize, then estimate how many groups fit into the whole pattern. This deliberately tidy example has five groups of ten, or 50 dots. Live practice scatters the dots; you create the groups mentally.
Read one result
Actual: 50. Estimate: 60. You overestimated by 10 dots. The relative error is 10 ÷ 50 = 20%, giving 61 / 100 with the normal score curve.
The same 10-dot difference against 80 dots is only 12.5% and scores 82. Compare the percentage error across patterns, not just the raw difference.
Change one thing next round
If your last few estimates were too high, check whether you counted sparse gaps as though they were full groups. Use one visible cluster as your unit, sweep across the pattern once, and make one final estimate. If errors alternate high and low, keep the same grouping method before adding a numerical correction.
Try Count with a grouping strategy ↑Diagnose repeated Count errors →Another way to trust your instincts.
Try Track