Arena — Focus

Flanker — answer the center arrow with ← → keys (or tap the pads), ignore the rest. Reaction time uses the pointer event.timeStamp correlated to stimulus onset; the seed-derived run is re-scored server-side. Same-setup — not a lab-grade replacement for the tests.

The science, playfully

Eriksen flanker · selective attention

The science behind it

The flanker task is one of psychology's cleanest tests of selective attention: a center arrow flanked by others pointing the opposite way. When the flankers disagree, everyone slows — that congruency effect is how much irrelevant noise leaks into your decision.

Reading your score

You race on your speed when you are correct on the hard, incongruent trials; the interference number (incongruent minus congruent) shows how much the distractors cost you. Lower interference means better filtering.

The honest limits

Interference is the gap between two noisy measurements, so it is extra jittery and needs many trials to settle — treat one round as a rough read. Same browser-observed ± and same-setup framing as everything else.

Per-action telemetry (stimulus onset, response event.timeStamp, sub-frame cursor path for aim modes, per-trial responses for decision modes) is re-derived from the seed and re-scored on the server. Ranks are provisional in beta.