Theists, Atheists, and Pattern Recognition
The answer cannot be as simple as "my eyes send me the information." What is incoming from my retina is not a description of the world, with pens and keyboards, but a pattern of colors, the contents of my visual field. Figuring out that the reason part of that pattern is red and part brown is that I am seeing a red object of particular shape and size sitting on a brown surface requires my brain to analyze that pattern and deduce from it what I am seeing. As A.I. researchers discovered when they tried to program computers to make such deductions, it isn't easy. We are able to do it only because evolution has provided our brains with very sophisticated pattern recognition software, probably incorporating a good deal of information about the nature of the world around us, hence the likely meaning of the patterns we see.
An analogous process occurs when we use all of the information available to us to form a picture of the world--not merely what is where in the visual field but what the universe is like and why. We are trying to construct a pattern, a picture of reality, which makes a reasonably good fit to the available facts. The fit is unlikely to be perfect, both because we may not get the pattern quite right and because some of the "facts" we are fitting may be wrong. And the process of constructing the pattern involves nothing as simple as formal logic. Just as in seeing, we are using pattern recognition software created by evolution and incorporating beliefs about the nature of reality—true or false—that led our ancestors to reproductive success.
Pattern recognition need not give an unambiguous result. In one familiar example, the same black and white picture can be seen either as a vase or as two faces. In another familiar example, a paranoid may have a picture of the world that fits all of the data available to him, with apparent inconsistencies explained by the plots of his enemies.
Which gets me back to the discussion of religion in my earlier post. Some people, trying to make sense of the world around them, construct a pattern that includes some sort of god. Others construct a pattern that doesn't. Neither pattern is the result of rigorous deduction from the data or anything close, so it isn't surprising that atheists cannot prove theists wrong, nor theists prove atheists wrong.
That does not mean that logic can tell us nothing at all about the subject. Some patterns are inconsistent with enough data to make it very unlikely that they are correct, or close to correct; one can climb Mount Olympus and observe the absence of the Olympians. But I think it is clear from a very large number of arguments conducted by many people over many centuries that one cannot, on that basis, reject either all versions of a universe with a god or gods, or all versions of a universe without.