I re-read Chapter 4 on Confounding/lurking variables couple of times. it most relevant to @patternscience Really understood why RCT are referred as the gold standard everywhere especially on the rule of exchangibilty (more on this later) and RA Fishers work