I’m new with the concept

Mutually exclusive experiments

, can anyone give some real use examples to explain when and how we should use such feature?

You can use the namespaces to be sure that users don't see any other tests within that namespace. One example might be on a pricing page where you have two experiments, one testing a coupon, and the other a pricing change. If they are not run in exclusion, you will get 4 combinations (AA, AB, BA, BB) and when you look at the results in A in isolation you will see both AA and AB - which can increase the noise and chances of incorrect results. Typically, most people overestimate how much experiments will interact, but there are some cases where you want to keep the user sets excluded.

Oh I see. So for your example, we will need to create 2 feature flags and each of them use 50% of the namespace allocation. and it means half of the user will only be in experiment A and another half will be in experiment B.

that's right

👍Thanks for explanation!