Before comparing piles, testers of a rec-center no-show model spell out use cases, users, sites, expected mixes, and hard edge days — using desk experts, an older booking system, and a trusted public benchmark — then take a stratified sample so every subgroup is in the baseline. Which representativeness step is that?

Before comparing piles, spell out use cases, users, sites, expected mixes, and hard edge days, then take a stratified sample so every subgroup is in the baseline. That is defining the target population and a stratified reference baseline. Fitting first, a Chapter 6 A/B method, and skipping the baseline all start in the wrong place.