[ausbah]

I think you're forgetting that the "replication crisis" also occurs in your so called "100% deterministic fields". CS has the issue of ML models, medicine drug studies, etc.

The bottom line is science is hard to do, and even harder to do right. I think all fields do try to follow the scientific method to the best of their ability, but for some fields it is simply harder to do because of the sheer complexity of the subject matter being hard to model, experiment in, and understand.