When I started out in science, I ran gels. It takes a day or two to master and then you just do it hundreds of times. If you fail, it costs your time only & you try again.
CryoEM takes months to master and a typical PhD student will solve <5 structures. How do we train?