Sure, I'll toss in my wild-ass guess.

I'm assuming we get to 90 deaths on 393 (262/131) patients by Jan 13

Discussion has ranged from 28% (36.7) to 35% (45.9) deaths in placebo group,

So worst case is a 37/53 split in deaths, 28.2% vs. 20.2%, a 28.4% reduction, p= .085 two sided. Ouch.

First significant case is 39/51 split in deaths, 29.8% vs. 19.5%, a 32.1% reduction, p=.028 two sided

Best case is a 46/44 split in deaths, 35.1% vs. 16.8%%, a 52.2% reduction, p= .0001 two sided.

Personally, my over/under is 42/48, 43% reduction, p=.004 2-sided