The Fine Tuning Fallacy

Fine tuning arguments are ubiquitous among proponents of theistic belief.  They are the ultimate "God Did It" argument.  That is to say, they appeal to ignorance.  It boils down to this:  How did the state of affairs in which we find ourselves come to be?  I don't know.  Therefore God Did It.  Now, of course, theists will object to that statement of the problem.  It's based on probability, they will tell you.  It's based on the fact that the probability (of physical laws and constants being what they are) is so small that we almost certainly wouldn't have found ourselves in this state of affairs without divine intervention.  So divine intervention is the most likely case.  But I'm here to tell you that this theistic argument based on probability is bogus.  And I'll explain why.

First, let me define a few terms.  In statistical analysis, there is something called "prior probability", and something called "posterior probability".  Prior probability is the known distribution of possibilities before any additional facts come to light that would affect the outcome.  For example, we might know that there is a prize (a shiny new Cadillac) behind one of three doors in a game show.  So our chances of selecting a door and winning the prize based on the prior probability is 1/3.  But what if we have some different information about the prize?  Let's say the there are a thousand objects and only one of them is the Cadillac.  And then three of those objects are chosen at random to place behind the doors.  This certainly affects the possibility of winning the prize, which is now reduced to 1/1000.  But that's still the prior probability.  It's what we know before any additional information comes to light.

Now, let us consider posterior probability.  We already have two scenarios described above.  Let's add an additional piece of information that can affect the outcome.  Monty Hall, the game show host, opens one of the three doors to reveal that the prize isn't there.  How does this change things?  In the first scenario (where the prize is known to be behind one of the three doors), we now know one of the three that is no longer a possibility.  The posterior probability of winning (given that we know not to choose one of the three) then becomes 1/2.  In the other scenario, we still don't know that one of the doors contains the prize, but we do know that our odds of driving home in the Cadillac have improved (slightly) to 1/999.  So the posterior probability reflects additional information - an improvement on our odds, while the prior probability remains unchanged.  But it is important to understand that knowing something about the outcome (like the prize is not behind door #3) doesn't imply that we know the prior probability. Either distribution scenario might still be true.

Let's consider a third scenario.  Let's say the game has been rigged - a Cadillac is behind each of the doors.  In this case, it doesn't matter which door we choose.  We are guaranteed to win.  The prior probability is 1.  In this case, any additional information is useless.  The posterior probability will always be 1. 

So now we have three different probability distribution scenarios.  Let's call them S1 (with prior probability 1/3), S2 (with prior probability 1/1000), and S3 (the rigged game with prior probability 1).  But we can only say what the prior probability is if that information is available to us in the first place.  What if Monty Hall opened door #3 and it was not the Cadillac?  One thing we could say is that S3 (the rigged game) is out of the question.  But is there any way of knowing whether S1 is true or S2 is true?  No, there isn't.  If the only thing we know is that the prize isn't behind door #3, then there is absolutely no basis to say that S1 is true, or S2 is true, or some other scenario that we haven't considered.

But what if Monty Hall opened door #3 and revealed the Cadillac?  We can say that our posterior probability is 1, and in this case, all three scenarios are still in play.  Knowing the outcome tells us nothing about the prior probability.  Is there any particular reason to suppose that it must have been a rigged game?  No, there isn't.  It might have just as well have been S1 or S2.  Without any additional information about the probability distribution beforehand, we just don't know what scenario might have been true.  All we know is that we are happily driving the Cadillac.  If we look at probability considerations alone, and we lack any prior information about the distribution, there is no basis to say that the game must have been rigged.

Just to enhance our illustration, let's think about a different game, with 1000 pennies on a table.  And once again, there are three different distribution scenarios.  In S1, half of them are heads up.  In S2, only one of them is heads up.  And in S3, all of them are heads up.  Now the game is to blindly choose one penny (without regard to whether S1 or S2 or S3 is true), and if you get tails, you die.  Now let's say we have played the game, and we find ourselves alive.  The only thing we know at this point is that we were lucky enough to live, but we don't know which distribution scenario might have been in play.  It is still entirely possible that many other players weren't so lucky.  What if we had not picked a heads?  We wouldn't be alive to know it.  Maybe S2 was the case, and a thousand people played the game, but we are the only one who survived.  That might make us feel special, but if we were not a survivor, we wouldn't be around to ask whether the game was rigged for our benefit.  It could just as well have been S1 or S2, and we have no way of knowing, and no reason based in probability to think it might have been one or the other.  The fact is that we have absolutely no reason to say it must have been a rigged game.

But this is exactly what theists do with their Fine Tuning argument.  First, they present a false dichotomy between S2 (the low probability scenario) and S3 (the rigged game), without any consideration of S1 (a higher probability scenario).  With regard to the probability of physical laws and constants supporting life, they make it seem extremely unlikely, but the truth is that their numbers are nothing more than speculation.  Nobody knows what the real probability distribution is, and anyone who tells you they do is lying.  And all this serves to set the stage for the second part of their deception, which is to say that with this dichotomy, we are safe in assuming that the rigged game scenario must be the case.  But that's just denying the rules of probability.  Even if it wasn't a false dichotomy, there is still no basis to make that assumption.

Just like any other theistic argument, Fine Tuning depends upon a distortion of logic.  If you're trying to argue for a God that doesn't exist in reality, this is what you're reduced to.  You can't make an argument that is valid and sound, because that could only result in truth, which isn't consistent with your theistic objective.