<<Your post is non-responsive insofar as I was talking about management's spin on the 9902b trial design, not the valuation of the company.>>
Iwfal's post seems quite responsive to me. The reason a 2% decrease in percentage of goals saved is material is that difference translates on average to 6/10ths of a goal per game, assuming 30 shots on goal per game. And with frequent ties and one goal wins in hockey, that difference will frequently matter to the outcome of the game, way more than 2% of the games. (please excuse my nonscientific use of the "average"). With the Dendreon phase III trial in question, a successful trial is the whole game -- and if we take the power statistic seriously, the chance of winning the game is 2% less.
And of course, that assumes a degree of precision for the power computation that is not real world -- 2% is likely noise, as Iwfal said, given all of the other factors that affect whether the trial will be successful in the real world.
