Chris and I explain what the long term mechanistic data will always outweigh that of outcome data, and all the ridiculous examples of where outcome data has led people astray