One brut way to find the cause of a regression change

Lets say we have N number of committed changes. A new defect surfaces. One brut way to find the cause of the defect is to revert the code to the point where the most dubious code change and see if the defect still arises. If it does, the most dubious code change or one of its successive commits are the culprits. If not, the most dubious code change is not the culprit.