If I run two A/B Tests 1) and 2), will about 50% of the users in 1)'s A be in 2)'s A and B ?
Or are Users always split the same way and when doing two A/B tests at the same time A and B are always the same Users?
Even if they are not the same users, is it safe that the users picked in two A/B tests are so random, that there is no skewing towards one or the other (e.g. 40% in 1)'s A are in 2)'s B