Understand your result
How Perceptle Unlimited scoring works.
Each round turns your error into a score from 0 to 100. Your session result is the rounded average of the rounds you completed. Here is exactly what those numbers mean.
The round formula
score = round(100 × exp(−0.5 × (error ÷ tolerance)²))
A perfect answer scores 100. At half the tolerance, the score is 88; at one tolerance, it is 61; at two tolerances, it is 14. The curve falls smoothly as the error increases, and the final round score is rounded to a whole number.
A tolerance is the scale used in that curve. It is not a pass/fail boundary and does not mean that answers within it all score 100.
What each error measures
| Game | Error unit | Normal tolerance | Friday tolerance |
|---|---|---|---|
| Count | Absolute count difference ÷ actual count | 0.20 (20%) | 0.32 (32%) |
| Track | Horizontal distance in the game’s internal field units | 60 | 90 |
| Angle | Degrees | 8° | 14° |
| Tempo | Mean absolute tap timing error in milliseconds | 60 ms | 85 ms |
| Split | Distance ÷ original line length | 0.05 (5%) | 0.10 (10%) |
Count uses relative error
If there are 60 dots and you answer 66, the difference is 6. The error is 6 ÷ 60 = 0.10. Against the normal tolerance of 0.20, that scores 88. A six-dot miss matters more in a small group than in a large group.
Track uses the underlying field, not screen pixels
The landing marker and target are compared horizontally on an 800-unit internal field. The display scales to your screen. A 60-unit normal-mode miss scores 61; it does not necessarily appear as 60 physical pixels on your phone.
Friday uses the landing point after the trajectory has reflected from the walls. The final distance is still horizontal distance along the landing line.
Angle changes how a full turn is treated
Normal Angle uses the shortest circular difference. A 4° error scores 88; an 8° error scores 61. A target of 359° and an answer of 1° differ by 2°, not 358°.
Friday Angle uses the absolute difference between complete accumulated angles. For a 450° target, 90° is 360° short and 810° is 360° too far. Matching the final direction does not erase a missing or extra full turn.
Tempo uses a fixed schedule
Four audible reference beats establish the interval. The next eight taps are compared with the scheduled continuation times. The error combines 60% of the absolute mean timing offset with 40% of timing jitter. Jitter is the population standard deviation of the eight early/late differences, so alternating early and late taps still carry a penalty.
Normal taps continue on the beat; Friday taps fall halfway between beats. At 100 BPM, the interval is 600 ms. Measured from the final reference beat, normal tap targets are 600, 1,200, …, 4,800 ms; Friday targets are 300, 900, …, 4,500 ms. If every tap is exactly 60 ms late, jitter is zero, the weighted error is 36 ms, and the normal score is 84. Alternating 60 ms early and late gives zero mean offset, 60 ms jitter, and 24 ms weighted error.
The reference does not adapt itself to your warm-up taps. This differs from the original game’s calibration. Audio latency, input latency and browser interruptions can affect the result, so a timing score should not be read as a device-independent measurement of ability.
Split keeps the original line as its unit
Suppose a target is 50% along a line and you select 55%. The error is 0.05, giving a normal score of 61. Friday’s extended answer space does not stretch the measuring unit: an error of half the original line remains 0.50, wherever the target lies.
Direction matters. A fraction of 3/2 from the right is measured leftward from the right endpoint, ending half a line beyond the left endpoint.
Your session score and grade
Add the whole-number round scores, divide by the number of completed rounds, then round the average. In a custom session, each selected round has the same weight. With repeated sets, all their rounds contribute; the result is not just the final set.
For 80, 90, 70, 85, 95, the total is 420 and the average is 84: grade B. If the 70 becomes 73, the average becomes 84.6, which rounds to 85: grade A.
| Grade | Rounded session score |
|---|---|
| S | 95–100 |
| A | 85–94 |
| B | 70–84 |
| C | 55–69 |
| D | 40–54 |
| F | Below 40 |
Make a fair comparison
Use the same seed, mode, difficulty, selected skills and number of sets. Easy and Hard alter targets or exposure rather than creating a universal difficulty conversion. Friday has different mechanics and wider tolerances. Scores across modes, devices or different skill selections answer different practice questions.
Your lowest average identifies the lowest-scoring skill in that session. It does not diagnose eyesight, attention or a medical condition. We do not calculate population ranks or claim your results match an official daily puzzle.