thehappybunny SG could we discuss the chaos theory?if not, how about hypothesis testing, sampling and estimation?would that be alright?