Estimating Average Treatment Effects in Block Randomized Experiments: A Conversation at the No Rules of Thumb Cafe