Can we use probability distributions to work out the likelyhood of only 1 planet having life?

[johnnoble90]

Hi. I'm a new member so if i have completely misunderstood this then please don't laugh at me.

 

I was having a discussion on another forum about the likelihood of life on other planets, and i thought about applying the poisson distribution to workout the highest probability that only 1 planet would have life on it (i.e us). I had to assume the mean was 1 and it came out at 36.7879%. Now this only shows the probability of only 1 planet forming with life, prior to knowing that 1 planet has formed with life, but it would at least show that if we are alone, it is a pretty big coincidence because even having an average designed to make it as unlikely as possible, it is still almost a third of a chance.

 

Can someone please explain whether this is correct or explain where i have gone wrong?

Edited May 24, 2011 by johnnoble90

[Chalky]

You said it yourself: you assumed that the mean was 1. Is it?

 

If we only consider the Earth, then yes. But what about other planets?

 

Consider Mercury, Venus and Mars as well. We currently have no evidence for life on them. Assuming that this is true, then your mean is already down to 0.25.

 

What about all the other planets in the solar system? The galaxy? The universe? Our mean potentially becomes very small. There's also an equation called the Drake Equation which may be of use to you.

[DrRocket]

Hi. I'm a new member so if i have completely misunderstood this then please don't laugh at me.

 

I was having a discussion on another forum about the likelihood of life on other planets, and i thought about applying the poisson distribution to workout the highest probability that only 1 planet would have life on it (i.e us). I had to assume the mean was 1 and it came out at 36.7879%. Now this only shows the probability of only 1 planet forming with life, prior to knowing that 1 planet has formed with life, but it would at least show that if we are alone, it is a pretty big coincidence because even having an average designed to make it as unlikely as possible, it is still almost a third of a chance.

 

Can someone please explain whether this is correct or explain where i have gone wrong?

 

In order to apply probability theory you need two things; 1) A probability space and 2) rationale for why that space applies to the issue at hand. You have only the second..

 

No one has any quantative model for abiogenesis.

 

Qualitatively it would seem that in a universe as large as this one life ought to arise elsewhere. But to make quantitative arguments without any basis is just fooling yourself.

 

Note that in your model you get the same probability for life nowhere. Does that seem reasonable ?

[johnnoble90]

You said it yourself: you assumed that the mean was 1. Is it?

 

No. The mean of 1 is the value where the likelihood of 1 planet having life is the highest. We have no evidence for the actual mean.

 

In order to apply probability theory you need two things; 1) A probability space and 2) rationale for why that space applies to the issue at hand. You have only the second..

 

No one has any quantative model for abiogenesis.

 

Qualitatively it would seem that in a universe as large as this one life ought to arise elsewhere. But to make quantitative arguments without any basis is just fooling yourself.

 

Note that in your model you get the same probability for life nowhere. Does that seem reasonable ?

 

Assuming the mean is 1 then yes it would be equally likely we "should have" had 0 planets with life. There is no evidence for what the actual mean would be though. A mean of 1 in this distribution was simply a worse case scenario.

 

All i was trying to do with this was show that the law of small numbers makes it unlikely that when life forms by chance, it is unlikely to form only once, rather than never or more than once. It is impossible at the moment to work out an actual probability of life on other planets.

 

It is possible i don't know enough about this distribution but i don't see how it wouldn't apply, as it seems to apply to everything else with a probability and we use it in our own logic as well considering you yourself stated that life seems likely on other planets, despite the fact that we don't really understand much about how life formed.

Edited May 26, 2011 by johnnoble90

[pantheory]

Hi. I'm a new member so if i have completely misunderstood this then please don't laugh at me.

 

I was having a discussion on another forum about the likelihood of life on other planets, and i thought about applying the poisson distribution to workout the highest probability that only 1 planet would have life on it (i.e us). I had to assume the mean was 1 and it came out at 36.7879%. Now this only shows the probability of only 1 planet forming with life, prior to knowing that 1 planet has formed with life, but it would at least show that if we are alone, it is a pretty big coincidence because even having an average designed to make it as unlikely as possible, it is still almost a third of a chance.

 

Can someone please explain whether this is correct or explain where i have gone wrong?

 

No John, the use of statistics for this reason is wrong. The reason is that there is no "normal distribution" involved which is the basis for normal curve or Poisson statistics. The exo-planets that we have detected so far are generally too large and/ or too close to its star to support life as we know it. Although we might be on the verge of discovering such planets thought to be within the so called Goodilocks zone (habitable zone). Also there are no statistics at all concerning other water planets so there could seemingly be no basis to assess the likelihood that water planets will produce life of any kind other than it happened once in our case. Statistics in general are based upon population information and it will probably be millennia before we could use statistics in this way (if ever). Right now a modified Drake Equation seemingly might be the best means to estimate probabilities of exo-planetary life in general.

Edited June 18, 2011 by pantheory

[Ophiolite]

I believe I have posted this in every thread in every forum on which this topic arises: you can't make a meaningful extrapolation from a sample size of one.