The equation
Seven pieces, multiplied
How many civilisations are in our galaxy that we could detect right now? In 1961 the astronomer Frank Drake wrote the question as a multiplication of seven quantities, so that a room of scientists could argue about one at a time. This page is that multiplication, with the range this site uses for each piece and a way to try your own.
N= R* × fp × ne × fl × fi × fc × L
- R*How many stars are born in the galaxy each year (stars per year)
- fpThe share of stars that have planets
- nePlanets per star that could support life (planets per star)
- flThe share of planets that could support life on which life actually starts
- fiThe share of planets with life on which it becomes intelligent
- fcThe share of intelligent species that make themselves detectable
- LHow long a civilisation stays detectable, in years (years)
Multiply the seven and you get N, the number of civilisations in our galaxy that could be detected right now. The first two pieces are measured: astronomers count stars being born and planets around other stars. The last five are not, and the honest answer to each is a range so wide that a single multiplied-out number misleads. So this site does what a 2018 paper by the researchers Anders Sandberg, Eric Drexler and Toby Ord did: each piece carries a range drawn from the published estimates, the multiplication is run many thousands of times with a random pick from each range, and the result is a spread of answers. The middle of that spread is the number on the home page; how the numbers are worked out says the rest.
Try the numbers yourself
move a slider and the answer changesYour N: detectable civilisations in the galaxy
200
Each slider moves in tenfold steps, so a small drag is a big change: that is how uncertain these pieces are. The tinted band under a slider is the range this site draws from when it computes the number on the home page; your position outside it is a position someone has argued, and the piece's page lists who. The slider's ends are only how far it can travel, not estimates.
Each piece, on its own page
7 of 7 written upHow many stars are born in the galaxy each year
The rate at which the Milky Way makes new stars, averaged over the time it takes a star to grow a civilisation.
Range used: 1 to 20 stars a year · 6 published estimates, from 1 stars a year (Frank Drake and the Green Bank meeting, 1961) onward
Piece 2 of 7 · fpThe share of stars that have planets
Of all the stars in the galaxy, what fraction have at least one planet.
Range used: 0.5 to 1 · 6 published estimates, from 0.2 to 0.5 (Frank Drake and the Green Bank meeting, 1961) onward
Piece 3 of 7 · nePlanets per star that could support life
For a star with planets, how many of them could support life: in practice, how many rocky planets of roughly Earth's size orbit where water could be liquid on the surface.
Range used: 0.01 to 1 planets per star · 8 published estimates, from 1 to 5 planets per star (Frank Drake and the Green Bank meeting, 1961) onward
Piece 4 of 7 · flThe share of planets that could support life on which life actually starts
Given a planet where water is liquid and the chemistry is right, how likely is it that life gets going at all.
Range used: centred on one start per planet, fifty orders of magnitude wide · 5 published estimates, from 1 (Frank Drake and the Green Bank meeting, 1961) onward
Piece 5 of 7 · fiThe share of planets with life on which it becomes intelligent
Given a planet with life, how likely it is that some of that life eventually becomes intelligent in the sense that matters here: able to build technology.
Range used: 0.0001 to 1 · 6 published estimates, from 1 (Frank Drake and the Green Bank meeting, 1961) onward
Piece 6 of 7 · fcThe share of intelligent species that make themselves detectable
Given an intelligent species, how likely it is to do something we could notice from here: leak or send radio, light up its planet, change its atmosphere, build things in space.
Range used: 0.01 to 1 · 4 published estimates, from 0.1 to 0.2 (Frank Drake and the Green Bank meeting, 1961) onward
Piece 7 of 7 · LHow long a civilisation stays detectable, in years
Once a civilisation becomes detectable, how many years it stays that way before it dies out, falls silent or changes into something we would not recognise.
Range used: 100 to 10 billion years · 7 published estimates, from 1000 to 100 million years (Frank Drake and the Green Bank meeting, 1961) onward