A few comments in reply to my thesis on whether or not we can know that god exists were of the form, "You admitted at the start that you can't know god exists, but then you argue that you can. WTF?"
I should have made this distinction more clear then; I was speaking initially of mathematical certainty and arguing at the end for scientific knowledge.
Some things can be absolutely known within a mathematical framework. "If (if A then B) then (if not B then not A)" is a tautological truth within the framework of mathematical logic. On a different note, beginning with the assumption of "B" then it's trivially simple to prove "If A then B". Within mathematics, some things can be proven with absolute certainty. In that sense, we cannot be absolutely certain on the existence of god because it's not a mathematical question.
Scientific knowledge is a different matter. We begin with facts, build from there to laws, and thence to theories. At any time, a contradictory fact could bring the edifice crashing down. Biologist J.B.S. Haldane said, according to popular myth, that rabbits in the Precambrian would explode the theory of evolution. In a sense, scientific knowledge is a mass of supposition resting precariously on the admission that it hasn't been proven false yet. However, the longer a scientific theory lasts without being proven false, the more certain we can be that it is true.
At this point an English major might be tempted to butt in with, "But everything ends up getting proven false! We used to think the Earth was flat! Einstein proved Newton wrong! Neener neener!" At this point I gleefully crib from a delightful essay by Isaac Asimov. Yes, older models have been proven incorrect, but they weren't entirely incorrect and they weren't equally incorrect. The Flat Earth model breaks down after just a few miles when you disappear under a horizon it simply cannot predict. Nevertheless, I've taken advantage of the Flat Earth model on each of my dozen or so cross-country drives by referring to a book of maps, all of them printed on flat paper without losing my way (at least not because of the maps). It's wrong, yes, but it's not entirely wrong. And the spherical model that replaced the flat model is also wrong! The Earth is an oblate spheroid (flattened at the poles) thanks to Newtonian physics, a fact ably demonstrated by the much more oblate Jupiter. And yet the oblate model is incorrect as well because, thanks to odd internal geography, our planet is slightly larger on the southern hemisphere. However, to say that the oblate spheroid model is as incorrect as the spherical model and as incorrect as the flat model is just wrong. One model is more wrong, and none of them are entirely wrong.
Each model accounts for certain observed facts (the earth is apparently flat) and is replaced by a better model that accounts for those and for others (the shadow on the moon is round and we disappear over the horizon), which is replaced in turn by another that takes further facts into account (the laws of motion). In each case, our knowledge is never scrapped or replaced entire, but upgraded and patched. Einstein didn't replace Newton, for Newton's observations and mechanics are nearly perfect for the observations he was capable of making; rather, Einstein expanded upon Newton by developing a mechanic for observations made in the centuries that followed, and Quantum Mechanics did the same a few decades later in the other direction. In other words, scientific knowledge has never truly been one of scrapping a false system for a true system, but of replacing a system with a more accurate one. Lose the burlap sack and put on a prom dress, lose the prom dress and put on a tailored suit.
So why do I believe there is no god on the basis of scientific evidence? Because, after thousands of years of observation, there is no evidence. The old canard "absence of evidence is not evidence of absence" can be discarded because absence of evidence, when one would reasonably expect to find evidence is evidence of absence. The longer we go with nothing more than feelings and hopes to confirm the alleged existence of a deity, the more firmly we can say, "Sorry, you've not only failed to convince me you're right, you've convinced me you're wrong." Yes, evolution could be toppled by a Precambrian bunny, but a century and a half of repeated attacks have failed to disprove it; we can say we know evolution is true, in spite of its precarious "not false yet" scientific status. After millenia of abject failure to provide proof or evidence for the existence of a god, or any supernatural phenomenon, why does religion still get a free pass in the opposite direction?
The alleged deity is such a large phenomenon that I feel safe saying I know there is no god because the evidence of such an overwhelming entity should be equally overwhelming. A natural high underwhelms those who don't experience it.