Reginald Williams in ‘The Case Against Theism’ (Think Autumn 2011) argued that the ‘empirically verifiable’ psychological need ‘to believe that good things exist when in fact they don't’ offers ‘the best reason anyone should expect’ for endorsing atheism over theism. My article outlines six objections to his thesis, questioning how empirically verifiable the evidence he adduces is, and pointing out various logical fallacies such as illicit use of generalizations and begging the question. It concludes that atheism needs defending on stronger grounds.