The network difficulty is generally at or around the minimum.
That means that each of your 4.6 CLAM outputs has a 2^20 / (4.6 * age_in_days) chance of staking each second.
The distribution was around 72 days ago, and so:
>>> 2**20 / (72 * 4.6)
3166
each 4.6 CLAM output will typically take about an hour on average to stake.
When it stakes, its age is destroyed, so you won't expect to see it stake again for a while.
And the 0.1 you saw is pseudo-random, and can be as much as 1000 CLAMs if you're really lucky.
99% of the time it's 0.1 though.
Ok, so forgive me if this has been addressed already, but I'm just curious if I understand this correctly. If I have 50 clams, odds are I am going to hit the same number of PoS blocks in a year no matter if those clams are in one pile or spread out in little piles over 100 different addresses. EXCEPT for that long confirmation time of 510 blocks. If we view PoS in clams as a lottery, the guy that has all his clams in one pile effectively doesn't have a ticket for 510 blocks. So the guy with lots of little piles will end up hitting more blocks, as he is always going to have a ticket. So it is advantageous to spread your clams out as much as possible. Do I got this right?