<Please get your facts straight.>
Somebody told me about this thread. As usual, Dew asked others to get the facts straight while fudging them himself. This will be my only post here to set the facts straight. I have no wish to resume a discussion with Dew about PSB or anything else.
<What I argued was not that the IMPACT trial would necessarily fail, but rather that neglecting to take into account program-survival bias was causing some posters ...>
First, your original prediction of HR 1.20 as well your later revised 1.25 would have guaranteed that IMPACT failed. That is a fact!
Second, nobody really knows what the heck PSB is from a computing point of view. As far as I could ever tell from all your past posts, PSB is a make-up thing for you to arbitrarily discount anything by whatever you could pull out of your biases.
<If I recall correctly, Ocyan initially predicted a hazard ratio* of more than 1.50 and later reduced his prediction to something closer to 1.40.>
You recall incorrectly. I have never explicitly called for a particular HR value. I know enough about statistics to not try to predict a particular value for something that should be thought of as a range while running simulations.
What I did do was to discuss the question of how Dendreon could assert constant HR for the range of events from 240 at the interim to 304 at the final and showed a plausible model of what they might have done in their sims. But that is a guess of "what Dendreon did", not what I actually did myself. Here is the post on investorvillage.com that talked about that:
www.investorvillage.com/smbd.asp?mb=971&mn=252189&pt=msg&mid=6914132
News 
Market Data 
Discover 
AI
