Sixty years of SETI silence is only evidence against ETI when P(detection | ETI) is high enough that detecting nothing is surprising. This workflow computes that probability, the Bayes factor K, and the updated posterior P(ETI | null) from your survey parameters.
The COSMIC/VLA Sky Survey (2023–2025) covered ~75% of the sky commensally with VLASS, targeting 950,000+ objects across L and S bands. Frequency coverage is partial: L+S bands span a small fraction of the plausible technosignature window (1 MHz–100 GHz).
Set fsky and ffreq to reflect both angular coverage and EIRP completeness. A survey that covers the whole sky but can only detect Kardashev-II transmitters has an effective ffreq limited by what fraction of broadcasting civilizations exceed that EIRP floor. The SKA mid-array (projected 2030s) will push both parameters substantially higher.
Omega Centauri specific: OC lies at δ = −47°, which puts it in a downgoing direction for IceCube but accessible to MeerKAT. For SETI, the VLA/COSMIC survey covered OC within fsky ≈ 0.75 and L+S band frequency coverage.
N (communicating civilizations) and δ (signal duration) together determine how much of the Galaxy's search space is filled with detectable signals at any given time. The Civiletti (2025) geometric factor min(1, 0.6δ/RMW) captures the fraction of the Galaxy that an outgoing broadcast sphere has already filled.
Sandberg, Drexler & Ord (2018) showed that propagating log-uniform priors across Drake parameters gives a prior with substantial probability mass at N = 0. The Drake (1961) point estimate gave N ≈ 20,000. The range from N = 1 to N = 10⁶ covers the full debate.
Signal duration δ matters geometrically: only signals currently in transit toward Earth are detectable. A civilization that broadcast for δ = 100 yr and stopped may have sent signals that already passed us. The factor 0.6δ/RMW captures this — for δ = 1,000 yr it is only 0.012, reflecting how thin the detectable shell is.
P0 = P(ETI exists) is your prior before the survey result. The Bayesian update is:
P(ETI | null) = K·P₀ / [K·P₀ + (1 − P₀)]
where K = 1 − P(detect | ETI).
When the survey is weak (K ≈ 1), the prior passes through almost unchanged — the null result is uninformative. When the survey is powerful (K ≪ 1), the null result strongly downweights the probability of ETI.
Reference priors: Agnostic: P₀ = 0.50. Sandberg (2018) range: P₀ ≈ 0.05–0.20. Rare Earth (Ward & Brownlee 2000): P₀ ≪ 0.01. Copernican ("life is common"): P₀ ≈ 0.99.
—