Thank you for pointing this out, this is my understanding of stratified sampling as well. However, I'm very surprised by the huge ~25x (!) reduction in the vari

Thank you for pointing this out, this is my understanding of stratified sampling as well. However, I'm very surprised by the huge ~25x (!) reduction in the variance found by the author's simulation between random and stratified sampling. I feel there must be something wrong with the simulation setup, but at first glance nothing seems amiss...