Vigilance Decrement: A Signal-Detection Radar Watch (2D)

Watch Controls

Event rate 15 / min

Signal salience 0.65

Shift length 3 min

Your Watch (click to flag)

Elapsed

00:00

Events seen

Sensitivity d′ (measured)

—

Criterion c (measured)

—

Hit rate (last 20)

—

False-alarm rate

—

Click the bright ring when you see it. Watch not started.

How it works

This is the 2D, actually-playable sibling of the 3D vigilance model: instead of an internal random-response draw deciding hits and misses for you, a real radar-style scope shows a stream of blips and you click whenever you think you see the rare bright target ring among the frequent dim distractor dots. Every response is real, so the sensitivity and criterion readouts below are measured from your own clicks, not read off a formula.

Each event has a short response window:
  click during it  → "yes"
  no click          → "no"

Hit           = target present & you clicked "yes"
Miss          = target present & you said "no"
False alarm   = target absent  & you clicked "yes"
Correct rej.  = target absent  & you said "no"

Behind the scenes a classic vigilance-decrement curve still runs on simulated time — sensitivity d'(t) decays and criterion c(t) drifts upward exactly as in the 3D model — but here it drives the target's on-screen visibility, fading the ring toward the background as the shift wears on:

d'(t) = d0 · exp(−3.5 · p)      p = elapsed / shift length, 0..1
c(t)  = c0 + 2.2 · p
target visibility ∝ d'(t) / d0
d0  scales with Signal salience

Event rate — how often a blip (target or distractor) appears; higher rates leave less time to notice each one.
Signal salience — how bright/large the target ring starts out; higher salience keeps it visible longer into the shift.
Shift length — how long the whole watch runs; the fatigue curve above always completes exactly once over this length, so a 1-minute shift decays just as visibly as a 10-minute one.

Because the fading is real and your responses are real, the d′ and criterion readouts below are computed from your actual last-20 hit/false-alarm rates using the standard z(H) − z(F) estimator — you should see your own measured d′ fall and criterion rise as the target gets harder to see, the same signature Mackworth found in real radar operators.

Hit Miss False alarm Correct rejection

Click anywhere on the
scope when you see
the bright target ring