Not just false, but pretty damn stupid.  Our sensory organs don't really have much to do with it.  There are two issues here:

1) Are our senses reliable?

Clearly, they are only reliable to a certain degree.  However, based on induction, and the literally billions of bits of data that we have consistently received through our sense organs, we can reasonably conclude that our senses are reliable within their limitations. 

2) Are all things equally probable?

A very basic understanding of probability will demonstrate that this is clearly not so.  If we flip a coin ten times, the odds are not even that the division will be 5:5.  If your friend believes this, tell him I'll bet him $100 on each of 100 series of ten coin flips.  I get everything but 5:5, and he gets 5:5.

He's partially correct.  All the gods are equally probable, which is to say that they're so staggeringly improbable as to be discarded as legitimate possibilities.

You see, for us to calculate probability, we must have evidence of the existence of the thing in question.  For instance, if I wanted to know the probability of a three legged frog being born, I could theoretically examine frog DNA, calculate reproduction rates within a given population, and predict such a thing with some degree of accuracy.

Gods are not properly defined, so there's no way to even begin calculating their probability.  Based on prior experience with things for which there is zero evidence, we can then say that 1) based on the fact that there's no definition of god, and 2) In the history of science, there's never been an example of a thing without identity existing, and 3) In the history of science, the actual existence of things for which there was previously no evidence has been extraordinarily rare, 

There is an astronomically small chance that any god exists.

 

Basically, he's trying to reduce you to nihilism, which is a losing battle.  We know for certain that we exist, or to be more precise, any conscious, sentient being knows that it exists, for existence is axiomatic.  To question existence, we must grant it.

From existence, we can deduce a couple of other axioms.  From these axioms, we can begin induction.  While induction is not 100% certain, it does encompass varying degrees of certainty -- which can be deduced with certainty.  In other words, I can deductively prove mathematical concepts, and math is the foundation for probability.

 

Nihilism is the last refuge of apologists who have nothing productive to say.

In essence, they're trying to turn the tables on atheists.. have their cake and eat it to.  Out of one side of their mouth, they will say, "You can't disprove god, so he might exist."  Out of the other side, they say, "You can't prove anything, so god might exist."

In other words, they are trying to create a philosophical loophole through which god might fit, but in doing so, they have denied the very foundation of philosophy, and can say nothing of anything.... In still more words, if they are truly to hold to nihilism, then their own words are admittedly meaningless, so their arguments and beliefs are completely unjustified.

 

 

To be precise, all gods who are described as supernatural are equally improbable, for supernatural precludes any further discussion.

Without bringing the lack of ontology into the picture, your statement is potentially better. However, we have to remember that some known quantities are necessary before we can make even rough probabilistic predictions.

Well.... there's logically impossible (like married bachelor), physically impossible (like perpetual motion) and biologically impossible (which depends on the time line of evolution as well as the limitations of DNA, environment, etc...)

Impossible is still impossible. If we are to use the word literally, any impossibility is still impossible, so the odds are technically 0.

Even accounting for the somewhat dubious use of Bayes Theorem to get the odds that "the sun will not rise tomorrow," we're still so close to zero as to be indistinguishable... more digits in the numerator than atoms in the universe.

Deduction is the only completely certain conclusion. Literally anything else is at least slightly uncertain.

Conclusive, on the other hand, happens relatively often. Anyone who's made it out of first year philosophy can demonstrate that we must act (and some would say we must believe) that the universe we observe really exists. If it does, then it is external. If it is external, then we can make observations about it that will be consistent with the observations of others, yada, yada, yada.

When something is 99.999999999999999999999999999999999% certain, it's conclusive. That happens pretty often in life, and relatively often in science.

 

No not really. I see where you are coming from I really do but there is a difference between logical inconstiency and invoking bullshit terms like supernatural. A logically inconsistent God is not possible in any concievable world let alone this one. A supernatural God could concievably exist in some world. Its just that all the evidence points to the world simply not being like that as such we can disregard supernatural "explanations" as being exceptionally unlikely to be a true account of the world.

I'd agree that calculating the probability of an unknow property is impossibel but we could beging by looking at the way the world actually is and form an estimate of probability that way. The fact that all "supernatural" explanations thus far proposed have actually turned out to have entirely natural causes is itself a rather strong inductive argument against supenatural explanations in general. We could use this to form some kind of porbability against.

For all practical purposes I would agree that Zues throwing thuder bolts (physically impossible) is as unlikely as an omnipotent being (logical impossability) BUT there is a technical difference between the two.

Indeed

Only if the deduction rests upon premises which are themselves certain. We can have these certain deductions within systems such as mathematics. But when it comes to discussing stuff that is out there in the world then this does not happen much which was my point. As some stage prety much all of our deductions about the external world rest upon empirically gleamed facts as such they are open to error also.

We are playing with sematics here. My use of the word conclusive above meant 100% certain. I agree that if we take it to mean very very likely then it is relativly common. But that 0.0000001% makes all the difference. But we are actually in agreement here.

Punch him in the face really hard. When he gets mad at you, and he will, tell him that by his logic it's equally probable that he got punched by George Washington, or God, or a leprechaun, or the entire original broadway cast of "Cats". Then ask why he's getting mad at you when he has twelve billion other potential suspects.

I know how frustrating it can be to argue with people that constantly resort to red herrings and other bullshit.

Probability has been dealt with, but I see a couple more problems. He's taking the limitations of sensory input (that we experience "light" rather than the fundamental particles of it, for instance) to mean that anything goes. There are conclusions that can be drawn from evidence, and those can later be falsified by further information, but at no point did the topic jump the rails into a probabilistic clusterfuck. The likelihood that an '89 Chevy will start on a cold day competes with the probability it won't. Without knowing the details (as in being able to experience the mechanical parts and know if they're hitting their mark at the time), it's a black box we can only deal with probabalistically. But it doesn't equally compete with the probability of it morphing into a sparkling unicorn when the key is turned.

Even if the argument were made, say, for "all possible universes," it doesn't translate into a suspension of the brute dynamics at work in our world. Even if we find that the fundamental nature of reality is counterintuitive, it doesn't affect our ultimate experience of it.