I was wondering the same, but I guess the advantage is being able to safely-ish run tests in production against production data and other live services the sor

I was wondering the same, but I guess the advantage is being able to safely-ish run tests in production against production data and other live services the sort of stuff you usually mock out in unit tests. and may not be feasible to have a whole copy of your prod 'big data' and services just for testing on with Scientist you get a report not just of mismatched results and exceptions (stuff good tests should usually be able to catch) but also if the new code is slower