Primal probability in percents

This is not a calculation formula this indicative prediction since according to Poisson approximation snapshot is 150 steps and cannot be count as 1/400.
P=sum----0…n(1/150)/S where S is some adjustable blizzard variable depending on some unknown conditions --im gonna believe in that, but 1/400 is completely bullshit, sorry )

Just thought I’d pop an update in this thread as the PTR is currently underway. Seasonal PTR heroes enjoy the benefits of a double bloodshard community buff. That means, if you do a GR90, you’ll end up with 800-900 bloodshards per run.

There’s an NPC in town which sells PTR caches containing legendaries / set items. As an example, there’s a Demon Hunter armour cache, for 50 bloodshards. Those caches contains every single piece of all five DH set armours, i.e. 31 set items. This means that for every GR90 run, you can buy 16-18 of these caches.

It’s easily possible to run GR90s in 3-4 minutes each, but let’s say 4 minutes to be safe. So, in an hour, you could easily get 15 of those runs done. At an average of 17 caches per run, that’s 15 * 17 caches = 255 caches. At 31 set items per cache, that’s 255 * 31 set items = 7905 set items.

So, farm for a few hours on the PTR and record your results. Let’s say you did 3 hours of farming, that’d be 23715 set items, which is a much better sample size than 400 or 1600.

The Expected Value of ancients would be 23715/10 = 2317-2318.
The Expected Value of primal ancients would be 23715/400 = 59-60.

Go ahead. Test it.