“In my belief that a large acquaintance with particulars often makes us wiser than the possession of abstract formulas […]”
from William James, preface, The Varieties of Religious Experience (1902).
Colin McGinn is asked what some of the best reasons are for not believing in God, and responds with (transcribed from here, at approx. 20:00):
“Well, the classic argument against [God] is the problem of evil. Even religious people find this one very uncomfortable. So the argument is simply, God is meant to be a being who is all-knowing, all-powerful, and all-good. So, how come there is suffering and pain in the world? Why does God allow it? God, if he’s all-good, thinks it’s bad that this should occur, would rather it didn’t occur, like any decent person, and yet he lets it occur. Now, that would be okay if he didn’t have the power to change it, but he’s all-powerful, we’re told by religious people he intervenes all the time in various ways. So, why doesn’t he intervene to prevent the death of a child or the torture of a prisoner? He doesn’t do it. So, you don’t want to conclude from that that God is actually quite a bad person – that’s a conceivable conclusion you might draw. But what you conclude from it is that the combination of these two characteristics is inconsistent. He’s all-good and he’s all-powerful – you need all-knowing too, because he has to know what’s going on – but it’s essentially the conflict between all-good and all-powerful and the existence of evil.”
Philosophers like easy ways of showing that something is absolutely a yes or a no based on conceptual arguments, instead of developing probabilistic arguments that rely on various strands of empirical evidence of varying degrees of strength.
For example, most contemporary theists’ conceptions of God probably aren’t centrally that of an omniscient-omnipotent-omnibenevolent being – rather, their concepts of God are centralized around various experiences, anecdotes, and so on, which fit in with daily practices (and so on) that seem to work, which they then might fit with theories they’ve heard or thought up based thereupon.
Arguments, then, against the existence of “God” that attack a particular theological conception must reckon with the fact that, if successful, all they’ve done is knock a somewhat artificial theological add-on off of an empirical body. It is as if I were to show that a contemporary theory of gravity is incoherent, and then declare “Gravity doesn’t exist!” The problem is that objects still fall when I release them, and so on, and these phenomena are what’s important and what give rise to various theories of gravity.
Central idea: what sustains and motivates theism is people’s experiences, not theological constructs like the above which to a large extent supervene on the experiences, et. al.