>>Somehow a trial enrolling sicker patients has a blended predicted survival (halabi) that is better than 01 or 02A<<
Gold has publicly said several times that when they checked the additional criteria for men to be accepted into 9902b, their symptoms did not correspond with an increase in death rates on the halabi. You evidently disagree with the analysis of these symptomatic patients against the halabi by their biostatistician. But, I think it is only fair to state that your disagreement on mortality is somehow a different biostatistical result from adding those symptomatic patients to the halabi than his is .