ATL, many thanks. I’ll take a look at them.
One quicker way to do this will be to do a one-tailed test around one year survival. Not a sure thing, but if the populations to one year survival rate don’t test out within a strong p-value, then that’s another signal. I don’t think that binary checkpoint requires a parametric distribution. But I’ll check. My only pause is the long-standing debate on start point (surgery v randomisation) on here. My belief that a 3 month offset is inappropriate but that 6 weeks might be necessary. Whilst that doesn’t materially change the stats over 20-40 month timelines, it might upset a 12 month snapshot survival rate.
Juggling a lot of stuff this weekend - so I may not get to it. I’m already sufficiently long that I don’t think I can convince myself to buy in more this week. Those of you holding 100ks of shares got in much earlier than I did.
It’s telling that someone came in at the last minute Friday for 36k shares. I don’t know if that’s a single block but it sure looks like someone dropped 50k usd into the shares at the last minute. Feels pretty suspicious to me - especially as there were some other big buy windows at the end of the day (iirc 20k at around 20 minutes left in the session). If something is oozing out, then that would be a pretty tell for it.
Market Data 
Markets 
AI
