Blink Reaction Test
Play one battle to measure your fastest clean blink response — free, on-device, and without an account.
What does the test measure?
BLINK BATTLE measures the delay between ECHO-01's eyelids starting to close in a true two-eye blink and your own eye closure starting. A battle report shows your fastest clean reaction after the round.
What is excluded?
Closing before the opponent's eyelids move is anticipation, not reaction, so negative timings are excluded. Winks, half-closes, and off-window blinks are gameplay mistakes rather than valid reaction samples.
Why calibrate first?
The game spends about eight seconds learning your personal eye-open and eye-closed thresholds. Calibration adapts detection to your eye shape, camera angle, and current lighting before timing begins.
Does the camera feed leave my device?
No. MediaPipe Face Landmarker runs in a Web Worker inside your browser. Frames stay in device memory and are not uploaded, recorded, or used for identity recognition.
Is this a medical test?
No. It is a casual game measurement affected by display timing, camera frame rate, lighting, eyelid travel, and model inference. It must not be used for medical, neurological, or fitness assessment.
How should I read my result?
The battle report shows your fastest valid response in milliseconds. A lower number means your eyelids started closing sooner after the opponent's true blink began. Use it mainly to compare your own rounds on the same device and in similar lighting; it is not a population percentile or a universal reaction-time score.
How is this different from a click reaction test?
A click test times a visual cue to a finger press. BLINK BATTLE times an animated two-eye blink to the start of your own eye closure, so camera frame rate, eyelid movement, and on-device detection are part of the measurement. The millisecond results from the two test types should not be compared directly.