NorthWest Biotherapeutics Inc (NWBO)

[exwannabe]

Senti, looks almost perfect.

Leprecon tricked you though, that is what they do after all

199 was just a pair of mislabeled points and all the rest were correct. You had 196,198,199,199.

I took your latest dataset and encoded it into a K/M calculation. As it is missing one censor I tossed it in where it most likely is (that thicket in the middle).

Doing that the numbers are almost dead on.

Senti 24 months: 46.36% vs Pub 46.2%
Senti 36 months: 25.51% vs Pub 25.4%

I would guess the very minor error is due to where I dropped the phantom censor and/or me getting a censor/event pair out of order.

I suspect you might never find the missing censor. There should be about 1 case of 2 censors on the exact same day of the meth curve. And I doubt it shows up much if any. Not really any point in this any more as I think we now have "the answer".

I would never have guessed anybody would have succeeded at this task.

What really sucks btw is that one can save the output of a graph as vector graphics and embed that in a PDF. Much better this way than as a pixel image as they did. For starters, you can expand it HUGE and it won't pixilate. But even better one can get the raw numbers direct w/o having to read the chart. Painless and fast.

Hope your eyes are OK