This is not the best analogy, but it is the best I can come up with to dispel the myth that an AI is better than a SERM for gynecomastia. This is not to say an AI is not useful to treat gyno, but it is far from the first choice. As in any choice, clinical context is paramount.
Here is the analogy. Actually, there are studies that basically use this paradigm. Imagine there is a castle with 100 gates to enter. And, you have the job of protecting the castle against marauders. The marauders in this case is E2! There are two options: (1) Protect ALL of the gates (SERM); or (2) Protect NONE of the gates, but kill 50+% (even say 90%) of the marauders (AI). Which of these options will best protect the castle?