You pick a small placebo group to compare to a new treatment arm, but you may have millions with the affliction taking the comparable treatment. maybe the placebo ends up being all NFL players, using the extreme to make a point., as in the case with Covid. Then results for control arm end up being better than the potential treatment, let’s say it’s B. We all ready know the stats for Covid patients, why do we need a control arm? It introduces potential error in the study. There is no power of thinking that can influence the equation, no placebo needed, which was the point I was making.
Market Data 
Markets 
AI
