Milliseconds, completed level, or remembered digit length.
Scoring & methodology
From one run to one readable score.
Every Human Arena result follows a visible path: record the challenge metric, convert it to a 0–100 value, then combine all five skills into one local Human Score.
01 · Score pipeline
Three steps, with no mystery layer.
Human Score is calculated in the browser from results stored on the current device.
Each skill uses its own fixed linear scale and clamps values to the 0–100 range.
The final number unlocks only after every axis has a saved result.
Human Scoreround((Reflex + Precision + Memory + Perception + Focus) ÷ 5)
The radar chart keeps the five normalized values visible, so the average never hides the shape of the profile.
02 · Five fixed scales
Each challenge keeps its natural unit.
Values between the displayed endpoints are scaled linearly. Results beyond an endpoint are capped at either 0 or 100 points.
01Lower is better
Reflex
Reaction Time
- Saved result
- Median of 5 valid rounds
- Scale endpoints
- 160 ms → 100 pts
520 ms → 0 pts
02Lower is better
Precision
Aim Trainer
- Saved result
- Average across 30 targets
- Scale endpoints
- 350 ms → 100 pts
1,400 ms → 0 pts
03Higher is better
Memory
Sequence Memory
- Saved result
- Highest completed level
- Scale endpoints
- Level 12 → 100 pts
Level 0 → 0 pts
04Higher is better
Perception
Visual Memory
- Saved result
- Highest completed level
- Scale endpoints
- Level 12 → 100 pts
Level 0 → 0 pts
05Higher is better
Focus
Number Memory
- Saved result
- Maximum digit length
- Scale endpoints
- 10 digits → 100 pts
0 digits → 0 pts
03 · P-values
A useful estimate, not a live world ranking.
Labels such as P75 are calculated from fixed reference curves built into the site. Higher P-values represent stronger results on that curve. They are designed to make personal results easier to read, not to claim that you beat an exact percentage of all people.
Every curve is capped between P5 and P95. There is currently no global participant dataset, demographic adjustment, or verified leaderboard behind these labels.
Reference rangeP5 → P95
- Reflex
- 160–520 ms
- Precision
- 360–1,500 ms
- Memory
- Level 1–12
- Perception
- Level 0–12
- Focus
- 1–10 digits
04 · Measurement conditions
Your setup is part of every browser result.
The most meaningful comparison is repeated play on the same browser, display, and input device.
Browser timing
Timing challenges use performance.now() to measure elapsed time. System load, background work, power-saving mode, and browser scheduling can still move a result.
Display and input latency
Refresh rate changes when a visual signal becomes visible. Mouse, keyboard, trackpad, stylus, and touch devices can introduce different input delays.
Why repetition matters
One attempt may be unusually fast or slow. Reflex uses the median of five valid rounds, while the records page helps compare repeated attempts over time.
05 · Interpretation
A game score, not a definition of ability.
Human Arena can help you notice patterns in your own attempts. Results are not medical, neurological, diagnostic, IQ, brain-age, or universal ability measurements.