How not to be afraid to put flexible experiments in an inflexible context