I have been reading Hugh Ross’ rather bad statistical guesswork on the probability of life occurring in the universe. Ross lists 154 parameters he says are required for intelligent life to exist in the universe. Then he assigns probabilities to them. Actually he assigns probabilities to 322 parameters (not 154), which is a little confusing. Anyway, he concludes there is less that a 1 in 10282 chance intelligent life would occur without a divine miracle. Poor odds! He doesn’t estimate the probability of a deity spontaneously poofing itself into existence so it is there to create the miracle he claims is necessary. Wouldn’t that probability be zero? If not, surely it must be turtles all the way down?
Ignoring that, where to start with Ross’ figures? The first obvious comment is that he doesn’t (as far as I can tell) say how he arrives at the probabilities of any of his 322 parameters – probabilities that range from 0.00001 to 0.7. In my view his conclusion is meaningless because we don’t have anything like enough information to complete such a task. That fact alone sinks his argument before we even begin.
His article is a variation on the finely tuned universe argument – the idea that since the universe's physical constants are within a small tolerance of what is necessary for life, it must have been designed that way. Usually far fewer variables are quoted, and I have to say I’m skeptical of the number of items Ross thinks are necessary, quite apart from the probabilities he claims he has calculated for each one. But the arguments against both positions are the same.
Fine tuning does not imply design
There are various arguments against the idea that fine tuning means design. First, design rests on the assumption that the only type of life possible is the carbon-based form we are used to here on earth. But there is no reason to suppose this is so: we are restricting ourselves here by our own lack of knowledge and imagination.
Second, and totally fatal to the design argument, is that current theories point to the existence of an infinite number of other universes, each possibly with different sets of physical constants and conditions. Certainly a multi-universe scenario can not be ruled out since no known principle requires that only one universe exists. In fact astrophysicist Victor Stenger states:
No new hypothesis is needed to consider multiple universes. In fact, it takes an added hypothesis to rule them out-- a super law of nature that says only one universe can exist. But we know of no such law, so we would violate Occam's razor to insist on only one universe.
The argument is that an infinite number of universes exist, this universe is the one randomly tuned for life, and so this is the one where we have evolved and where we are here to observe that the universe is fine tuned for life. What else would we expect to observe? This is known as the Weak Anthropic Principle.
A third argument is that a fundamental theory of everything might predict that only a very narrow range of physical constants or even no range at all, would even be possible anyway. In other words, right now we just don’t know what the odds would be.
Talk Reason has a good write-up on why The Anthropic Principle Does Not Support Supernaturalism. In this article they use probability theory to show that the fine tuned argument actually contradicts design. Ross is just wrong.
Little green men?
Scott at the As If blog had another angle on this. He looked at Ross’ site and says Ross’ arguments makes it highly unlikely there are other inhabited planets in the universe. Scott takes Ross’ 154 parameters and assigns 50% probability to each (higher than Ross) and concludes the probability of there being a planet with all of these factors is less than 1 in 2 x 1046. He concludes that not only does the existence of Earth demand an explanation beyond random chance; it is also evidence that we are alone in the universe. See his post What are the odds? for the details. Scott asked for my comments.
Of course, if Ross is right and life demands a divine miracle, then applying probabilities to see if we are alone is meaningless - if God created us against the odds he could have created other life too. That argument didn’t especially appeal to me though. There are enough other reasons to reject Ross’ figures as evidence we are alone.
First, several of Ross’ parameters are for things that he says are necessary for life anywhere in the universe. For example:
6. decay rate of cold dark matter particles
7. hypernovae eruptions
9. white dwarf binaries
10. proximity of solar nebula to a supernova eruption
11. timing of solar nebula formation relative to supernova eruption
86. density of giant galaxies in the early universe
109. decay rate of cold dark matter particles
135. level of supersonic turbulence in the infant universe
136. number density of the first metal-free stars to form in the universe
141. heavy element abundance in the intracluster medium for the early universe
142. heavy element abundance in the intracluster medium for the early universe
144. pressure of the intra-galaxy-cluster medium
I’m not sure but I think there are several more. These, according to Ross, are all necessary for life to form in the universe, and those 12 items according to Scott’s formula would have a combined probability of only 0.00024. But since we know life (ie us) did form in the universe, we know the probability for all of these occurring is actually 1. These parameters, and probably several others, should be removed from Scott’s calculation.
Rolling up probabilities
Another major problem with Ross’ calculations is that many of his parameters are dependent on other parameters, resulting in incorrect multiplying up of probabilities. For example, take these four:
112. rate of nearby gamma ray bursts
119. gamma-ray burst events
128. frequency of gamma ray bursts in galaxy
152. quantity and proximity of gamma-ray burst events relative to emerging solar nebula
All of which if “too many” or “too few” are said to be a problem for life. Scott essentially multiplied these four 0.5 probabilities together to make a probability of all four occurring to be 0.0625. But surely they are essentially different ways of saying the same thing? He is multiplying four probabilities when there should just be one – 0.5.
How about these three:
20. parent star age
21. parent star mass
23. parent star color
But aren’t the color, mass and age of a star interdependent? For example, massive hot blue stars burn more quickly and therefore have shorter lives.
Look at these:
71. Kuiper Belt of asteroids (beyond Neptune)
94. total mass of Kuiper Belt asteroids
I would be grateful if anyone could explain the difference between those two.
Then there are these three:
105. distance from nearest black hole
15. parent star distance from center of galaxy
132. parent star distance from galaxy’s corotation circle
… any of which if “too close” or “too far” are claimed to mean life could not exist. But aren’t these essentially the same thing, or at least very similar? But there’s more – the above three are also linked to:
8. supernovae eruptions
16. parent star distance from closest spiral arm
24. galactic tides
110. ratio of inner dark halo mass to stellar mass for galaxy
117. star formation rate in solar neighborhood during past 4 billion years
118. variation in star formation rate in solar neighborhood during past 4 billion years

The reason is that these nine parameters (and maybe others), are all related to what is known as the Galactic Habitable Zone (GHZ): an area of space with a safe environment and access to the chemical materials necessary for building Earthlike planets. Scott’s 0.5 probability for each of these nine would yield a combined probability of 0.00195 (0.2%). However, a paper published in the journal Science in January 2004 estimated as many as 10% of all stars in the Milky Way galaxy would be within the GHZ – over 50 times what Scott would calculate.
Here’s another problem. The following numbered parameters are all to do with plate tectonics: 14, 54, 92, 153 and 154. Combined, with a 0.5 probability of each, they suggest the probability of plate tectonics arising as 0.03125. But SETI state:
There is nothing miraculous about tectonic activity, and in fact there is some evidence that it has occurred on both Mars and Venus.
Plate tectonics are probably fairly common, possibly even the norm, despite the 3% probability suggested by Scott using Ross’ parameters.
There are so many others like this it would be tedious to list them all, but I think the point is made: Ross did not allow for interdependencies – he incorrectly multiplies interdependent probabilities.
More errors
Then there are some parameters are just wrong. For example:
23. parent star color
* if redder: photosynthetic response would be insufficient
* if bluer: photosynthetic response would be insufficient
But SETI states:
Indeed, even stars somewhat brighter than the Sun, which because of their heightened ultraviolet production might be thought dangerous, could be the loci of life. Jim Kasting has shown that the enhanced ultraviolet radiation from F-type stars would produce so much atmospheric ozone that planetary surfaces could be very well shielded.
Also, Astrobiology Magazine states:
Of course, the development of photosynthesis on Earth did not just rely on the sun's wavelengths. It also was dependent on various factors specific to our planet, such as cloud cover, atmospheric composition, amount of land, and the depth of the ocean. Since so many factors played a role in the evolution of photosynthesis on Earth, perhaps photosynthesis could evolve on other worlds orbiting less-than-ideal stars.
In addition, other planetary factors perhaps could speed up the rate of photosynthesis development. There does not seem to be a particular reason why photosynthesis must develop at the same pace as it did on Earth.
As long as planets have enough liquid water oceans and cloud cover to protect against UV radiation, say Wolstencroft and Raven, planets orbiting most stars should be able to develop photosynthesis. One key to finding such planets is to look in the star's habitable zone.
(My bold.)
Ross’ probabilities ignore the habitable zone phenomenon: Earth like temperatures can be found around most stars, but at different distances. In addition, many of Ross’ parameters refer to Jupiter or Neptune, or to other factors relating just to our solar system. Again, Ross assumes that the formation of life requires a solar system like ours. He ignores the possibility of differing evolutionary paths that could adapt to colder/ hotter stars and planets, more/fewer massive planets, and countless other differing environments where life could still evolve. Ross also assumes that the time it took for intelligent life to evolve on Earth is the time it will take in all cases. Making these conclusions based on a statistical sample of one planet is absurd, and yet that is precisely what many of his parameters assume.
There are so many dubious parameters in Ross’ list it would be tedious to list them all, but I think the point is made: Ross’ list is bogus. Talk Reason has a very critical evaluation of Ross, covering some of these points, and others.
The Drake equation
The more usually quoted calculation for this kind of thing is the Drake equation devised by Dr. Frank Drake in the 1960s. The Drake equation estimates N: the number of civilizations in our galaxy with which we might expect to be able to communicate. Historical estimates of N range from 0.05 to 5,000, depending on the assumptions. From the link it seems like the current thinking is N = 0.0000008316. If there are 200 billion galaxies, I think that would make the potential number of civilizations with intelligence in this universe to be around 166,000.
Is that a reasonable estimate? In my opinion it is also based on many assumptions that we cannot be sure of – in other words I don’t know. But I’m pretty sure it’s a better guess than using Ross.
(Note: Pictures are from NASA and the Hubble)