S9v1
Yep

I was asleep for that one

Good to know im not the only one who find a block here on kano with S9v1

looks like all those "bad" miners still able to find blocks

There's no doubt they find blocks, they've found 95 so far.
The issue is the total number S9v1 has found is well below what it should have - CDF[Erl] = 0.991261 with that last one.
(They're expected to have found almost 25 more)
...
Had all those machines stayed they could have hit those blocks.
...
Yes, and they 'could' have hit more, less, or even none.
Doesn't matter what they 'could' have done, since that's
supposed to be random.
Unfortunately, history suggests otherwise.
You altered the history when you told the S9v1 to leave without a long enough sample size. Had they stayed the variance would of caught back up. Any sample size should be bare minimum one year.
Time isn't a factor in determining block luck statistics.
The number of hashes and the number of blocks are the only 2 numbers that matter.
I've no idea where you got 'one year' from.
But they have also been hashing on the pool since I asked people to switch, just not as many.
Time is a factor to figure for block luck statistics. You might have no blocks one day and ten the next. If you calculate the statistics in a short amount of time. You will not have a real number for what the statistics will be over a longer amount of time.
But they have also been hashing on the pool since I asked people to switch, just not as many.
Yes and that is part of my point. Had the majority of the s9 stayed. They could have gone on a lucky streak and come back even with the expected number. But we will never know now since you asked them to leave based on your statistics taken over a way to short of a time frame.
At the same time you asked the s9 to leave. Someone asked about the expected amount for the new avalons. Which you replied there had not been "enough time to get a good idea of the performance of the avalon." Funny how time is a factor when it comes to the avalon. But not the s9.