> there is a chance of something other causing the instant relief AND matching the moment of cannabis use You forgot about selection bias here The point is no

> there is a chance of something other causing the instant relief AND matching the moment of cannabis use You forgot about selection bias here The point is not it's impossible to deduce. The point was that without randomization even simple passage of time gets equivalent support from that experience as cannabis. People make that mistake all the time with common cold, cause they drink some special tea on day 7, get the final relief, and think that was the tea.