It would not surprise me if the placebo rate was lower than predicted, as this has been a recurring and frustrating feature of prevention trials. One possibility is that participants may instinctively alter their behaviour (diet, exercise, etc) when they enter a study. Another is that investigators may be too optimististic (subconsciously) when doing the power calculations for a trial. If this is what happened in BETonMACE, I agree it reduces the likelihood of their being a significant Lipitor/Crestor difference. Time will tell!