Using a wink as a command requires left/right discrimination and is infeasible for some users

What it is

A wink treats the two lids as two keys: a left closure one command, a right closure another. That requires hardware that can stably tell which side closed, and a user who can actually close only one side. A substantial fraction of people cannot produce a clean unilateral closure—the other side moves with it, or one side will not close at all. For them a wink is not hard; it is infeasible. It is also not a long blink: long blinks are usually a bilateral duration channel.

Why it happens

Unilateral closure depends on independent facial-nerve control of orbicularis oculi, and the sides are not symmetric. Many people can wink only one side; the other co-contracts; some co-contract on both. The tracker must still report gaze from the open eye while one is closed, and must not confuse the closed side with tracking loss. Shadows in sun, hair covering one eye, weaker IR on one headset camera, all mix “closed” with “lost.”

Feasibility stacks with social meaning. A deliberate wink is a flirt or a joke in many settings, and as a command it has a social cost in public. Facial palsy, post-stroke, and some congenital asymmetries permanently remove one side.

Studying it

Screen capability first, with no machine: can the person close only left, only right, and how much does the other side co-move. Then measure unilateral detection accuracy on the target hardware against tracking loss. Do not write success among people who passed screening as product usability—that is a conditioned sample. Report how many of a random recruit cannot produce a separable unilateral closure at all. Facial-nerve disease, glasses, and bangs belong as strata.

Where it stops holding

People who can wink cleanly and who use the system in private can gain bandwidth from left and right as two keys. People who can close only one side can use that side as the only key; do not require both. If a headset often loses one pupil, take wink out of the design and use bilateral long closure instead. Children’s unilateral control is less stable and should not be the default.

Applying it

• Make wink an optional channel, with a left/right capability check on first enable; do not assign a command to a side the person cannot produce.
• Default wink off in public, or offer a confirm that does not look like a wink.
• Verify in an unscreened sample the fraction who cannot wink unilaterally, and on a headset that often loses one eye confirm that tracking loss is not counted as a wink.

Related

• Same group: C8.11.1 Spontaneous physiological blinks and voluntary blinks must be distinguished · C8.11.2 Voluntary blink commands are usually recognized by lid-closure duration, not a single blink · C8.11.4 False triggers of blink input are easily confused with fatigue and dryness
• Adjacent: C8.04 Gaze-plus-confirm combinations · C3.24 Accessible alternatives to gestures
• Search: wink · unilateral blink · facial nerve