Cel Sci Corporation (CVM)

[lightrock]

Where is your work to back up your statement?

That statement being something like this...

I think the imperical data will show something in-between what we've discussed. Something more similar to a static (negligible) benefit initially and then an osi that increases at an increasing rate with additional time. This would look like survial curves that start nearly coincident and then after a period of time the divergence begins at an accelerating pace.

And, interestingly not entirely wrong. The fix is that it converges again if you are trying to look between a date 1 and a date 2 as opposed to the whole thing at 298 events. The 'fault' is the enrollment how many exactly when = exactly when is it physically possible to see how much difference (day to day).

Y'all may remember this monte carlo... which says pretty much the same thing.



To me the biggest problem is we should be "empirical" unless we're talking about Darth Vader which would be "imperial".

:)