Five rounds, one rounded average
In the original daily game, add the five whole-number event scores, divide by five, and round the result. The grade uses that rounded total: S ≥95, A ≥85, B ≥70, C ≥55, D ≥40, otherwise F.
For example, 90 + 88 + 72 + 96 + 81 = 427. The mean is 85.4, the displayed total is 85, and the grade is A. This is a score out of 100, not a percentile or an IQ result.
Enter your five scores to see the calculation and how many additional event points would cross the next grade. The improvement plan raises the lowest unlocked score first; it is a mathematical allocation, not a claim about which skill you will learn fastest.
Each test measures a different error
| Test | Error unit | Normal tolerance | Friday tolerance |
|---|---|---|---|
| Count | |estimate − actual| ÷ actual | 0.20 | 0.32 |
| Track | Internal horizontal field units | 60 | 90 |
| Angle | Degrees | 8° | 14° |
| Tempo | Timing error, milliseconds | 60 ms | 85 ms |
| Split | Fraction of the original line length | 0.05 | 0.10 |
The event score is round(100 × exp(−0.5 × (error ÷ tolerance)²)). At exactly one tolerance of error it is 61, rather than a linear deduction. A larger Friday tolerance allows more error for the same score.
Compare an estimate with its target and see the curve. Track uses the original 800-unit-wide field; do not enter unscaled screen pixels.
Friday changes the target, not just the difficulty
Normal Angle uses the shorter circular difference: 359° and 1° are 2° apart. Friday Angle counts every full turn. A target of 579° is one complete counterclockwise turn plus 219°; stopping at 219° leaves a 360° error.
Split starts at the endpoint named in the prompt. With a left origin, the normalized target is the fraction itself; with a right origin, it is one minus the fraction. A left-origin 8/5 target lies 0.6 original line lengths beyond the right endpoint. Try winding an angle or place a fraction beyond the line.
BPM, offbeats and calibration
A beat period is 60,000 ÷ BPM milliseconds. At 120 BPM, normal continuation taps are 500, 1,000, 1,500 … milliseconds after the final reference beat. Friday offbeats fall at 250, 750, 1,250 … milliseconds; the tempo stays 120 BPM.
Our independent Tempo tool uses the audible reference as its clock. It combines 60% of the absolute mean timing offset with 40% of the population standard deviation of the eight tap errors, then applies the score curve. The original also calibrates using your four warm-up taps. Its final score can therefore differ from this exercise even at the same BPM.
Audio output, Bluetooth delay, touch input and background tabs can affect timing. Use the tool to understand the pulse and error pattern, rather than treating its number as an official result.
Independent Unlimited imports
The analyzer accepts Perceptle Unlimited v1 JSON exports and PU-R1 result codes from perceptle.io. The two formats are independent of the original daily game. An imported run uses its actual completed rounds, including repeated skills and custom sets; omitted skills are not filled with zeroes.
The analyzer regenerates supported seed/configuration targets and recomputes results from answers. It does not trust cached totals. Files remain in your browser. Unsupported versions and malformed records produce a specific error rather than an invented result. See import help.