Philip Hampsheir
First published 1st April 2026. Updated 24th April 2026.
Introduction
There is a criticism of the Drake equation so well-worn it has achieved the status of received wisdom. Mention the probability of life emerging on a habitable planet, and someone will note, quite correctly, that we are extrapolating from a sample size of one. One planet. One data point. One example of life in the cosmos, from which we are expected to derive universal probabilities. It is, as criticisms go, entirely legitimate — and almost entirely wrong.
The sample size was never one. It was three.
The oversight is so straightforward it is almost embarrassing. Our solar system contains not one but three planets that were, by any reasonable scientific definition, habitable at some point in their history. Earth, obviously. Mars, with its ancient river valleys, lake beds and the rich catalogue of geological evidence being assembled by the Curiosity and Perseverance rovers. And Venus — where recent modelling by NASA researchers suggests temperate surface water may have persisted for billions of years before the planet’s runaway greenhouse event erased it entirely.
The key to the Venus argument is the Faint Young Sun Paradox. Four billion years ago, our sun burned at roughly 70 to 75 percent of its current output. Under those conditions, Venus — closer to the sun but receiving correspondingly less radiation than it does today — sat comfortably within what planetary scientists call the circumstellar habitable zone. If we accept, as the geological record on Mars increasingly demands, that the early solar system supported liquid water and clement conditions on multiple worlds, then Venus becomes a very strong candidate rather than a speculation. We have three candidates.
One survived.
What the Ratio Tells Us
This matters because it fundamentally changes the evidential character of a long-standing debate. The ‘n=1 problem’ has been used, legitimately, to argue that we cannot meaningfully estimate the frequency of life in the universe because we cannot distinguish between a cosmos in which life is commonplace and one in which it is vanishingly rare. A single data point sits neutrally between those positions.
Three data points do not sit neutrally. A 33 percent retention rate within a single optimised system — a stable, middle-aged G-type star, a Jupiter-class outer planet intercepting potential impactors, multiple candidates distributed across the habitable zone — tells us something empirically real about the severity of what astrobiologists call the ‘hard filters’: the mechanisms that terminate habitability before complex life can establish itself.
It is worth emphasising what the argument does not require. It does not require that Mars ever developed life. It does not require that Venus did either. The hard filter argument operates independently of origin: even granting early habitability on all three worlds, two lost that habitability before the timescales associated with complex life could elapse. The operating window on both Mars and Venus closed, by current estimates, within the first one to two billion years. Given that the transition from prokaryotic to eukaryotic life on Earth alone consumed approximately two billion years — and complex, multicellular organisms required a further billion — the implications are sobering.
In a system with every apparent structural advantage, the window slammed shut on two of three candidates before the evolutionary clock had barely started.
Expanding the Denominator
There is a counterargument, and it is a good one. If we expand our definition of habitability beyond the classical surface-water model, the denominator grows rather than shrinks. Europa, with its vast subsurface ocean, and Enceladus, with its active geothermal venting confirmed by the Cassini mission, represent a category of candidate worlds entirely outside the traditional habitable zone. Titan adds a further and chemically exotic case. If any of these bodies hosts life — even microbial, even simple — the entire conceptual framework around stellar luminosity and orbital distance requires revision.
But this expansion of the denominator does not weaken the original argument; it complicates it productively. A confirmed detection of life in the subsurface ocean of Europa would drive the estimated value of fl — the Drake equation’s term for the fraction of habitable worlds that actually develop life — dramatically upward. It would also confirm that hard filters operate at origin rather than persistence, since subsurface ocean worlds are far more stable over geological time than surface-ocean planets vulnerable to atmospheric collapse.
Either result is interesting. Both results are available from our own solar system, without travelling a single light year.
The Mars Complication
Mars is currently the most interesting object in this discussion, and not only because it is the most accessible. The September 2024 analysis of the Cheyava Falls sample from Jezero Crater — leopard-spot morphology, organic compounds, vivianite and greigite in proximity, all consistent with microbial metabolism on Earth — places the biological hypothesis squarely in ‘null territory’: neither confirmed nor excluded, with the needle persistently straining toward confirmation.
If Martian life is eventually confirmed — whether extinct or, in some protected subsurface niche, extant — the n=3 argument gains a further dimension. The trajectory and ejection dynamics of material between Mars and Earth, under early solar system conditions, favour Mars-to-Earth transfer over the reverse. If life predates Earth’s own biogenesis on Mars, the question of life’s origin on our own planet shifts address. The implications are not merely statistical.
The Loose Case
This paper has proceeded on the tight case: three surface-water worlds, classical habitable zone, early solar system, like-for-like comparison. That is the cleanest evidential ground and the most defensible basis for the retention rate argument. But the honest number is N = 3 + x, where x is a matter of the reader’s tolerance for including subsurface ocean candidates. In rough order of habitability confidence, the x population runs: Europa, Enceladus, Titan (with caveats on organic transfer to the water layer), Ganymede, Ceres, and — at the speculative tail — Tethys and others. Include none of them and the tight case stands on its own terms. Include all of them and the distribution becomes considerably more robust, the argument considerably harder to dismiss. The core claim — that our solar system is a small but real statistical sample of habitability outcomes, not a singleton — strengthens as x increases. The reader may proceed as far as their evidential standards permit.
Correcting the Record
The standard critique of the Drake equation’s sample size is not wrong in principle. Extrapolating universal probabilities from a single example is methodologically suspect and anyone applying rigorous Bayesian reasoning to the problem has said so at length. The error is in assuming the single example is Earth.
The example is the solar system. Within it, we have three independently habitable worlds, two hard filter events of confirmed severity, a retention rate of one in three under optimal structural conditions, and a growing portfolio of candidate bodies that may extend the sample further. That is not one data point. That is the beginning of a distribution.
The answer to the question of whether we are alone in the universe may be unknowable at our current technological reach. But the answer to the question of whether we have enough data to say something meaningful about the filters that govern that probability — that answer, it turns out, has been sitting in our own backyard for four billion years.
Philip Hampsheir is a journalist and broadcaster with nearly three decades of experience at BBC, Bloomberg and TRT World. He produces the YouTube history channel A Piece of the Past.