We can measure tolerances and understand variables that we couldn’t in the past. This awareness makes us want to manipulate these variables in an attempt to control outcomes.
Measuring a variable does not mean we understand its impact on the outcome when it is confounded. Furthermore, we have limited capacity to manipulate all of the variables.
How many improvements have you seen implemented at work that did not lead to the expected outcome? How do we determine improvements are the improvement rather than just noise we are taking credit for?