The Signal

You are a SETI analyst. Your radio telescope is pointed at a patch of sky that has been quiet for six months. The waterfall display scrolls endlessly — row after row of noise, rendered as a dark heatmap. Somewhere in that noise, there might be a signal. Your job is to decide when to sound the alarm.

The display on your right simulates what you see. Each row is a moment in time. Each column is a frequency bin. The colors tell you what the algorithm decided: green means it flagged something as real. Orange means it flagged noise by mistake.

Crying Wolf

Look at the Detection Threshold slider. It's set to 2.0σ — meaning the algorithm only flags a frequency bin when its amplitude is more than two standard deviations above the noise floor. Drag it down to 1.0.

Watch what happens. Orange floods the display. The algorithm is flagging everything. Most of those orange cells are pure noise that happened to spike. You haven't found more aliens — you've created more false alarms.

Now drag it back up to 3.0. The orange vanishes. But so do some of the green cells. Real signals that were too faint got classified as noise. You missed them.

The Same Tradeoff

This is the same tradeoff you saw in the doctor's office. Lower the threshold and you catch more real signals — but you also catch more noise. Raise it and you get fewer false alarms — but you miss faint signals.

Try dragging the Signal Strength slider. This controls how far the real signal rises above the noise floor. At 0, there is no signal at all — every green cell would be a mistake. At 3.0, the signal is unmistakable.

Now try Noise Level. More noise makes everything harder. The same signal that was obvious at noise level 0.5 disappears into static at 2.0.

The Equation

Toggle the equation overlay. The key number is d′ (d-prime) — the sensitivity index. It measures the gap between the signal distribution and the noise distribution, in units of noise. A d′ of 0 means signal and noise are indistinguishable. A d′ of 3 means the signal stands well clear.

Watch how d′ changes as you drag the sliders. Signal strength up? d′ goes up. Noise level up? d′ goes down. The threshold doesn't appear in d′ at all — it controls where you draw the line, not how distinguishable the signal is.

Drawing the Line

The threshold sets your tolerance for uncertainty. Low threshold: you'll catch every signal, but your inbox fills with noise. High threshold: you'll never cry wolf, but you'll miss the faint whisper that could change everything.

This tradeoff appears everywhere. Next, you'll see it on election night — where the noise isn't random static but systematic polling error, and the stakes are a democracy.