Quote from: WBF1 on March 03, 2018, 12:51:34 AM

Quote from: kano on March 02, 2018, 09:37:12 PM

Quote from: WBF1 on March 02, 2018, 08:08:17 PM

...
- All else being equal, on a long enough time scale, the ONLY thing that makes one pool pay more than others is a lower pool fee.
...

Yes, but the "All else being equal" also includes pool hash rate to some extent.
If the pool gets too small - as I've been trying to avoid over the last couple of months and will hopefully have some more hash rate here soon again - the diff changes also can affect your reward dramatically.

This is why I mention on occasion "as long as we are expected to find a number of blocks each diff change"
The issue is that if you're mining for multiple diff changes per block, then your reward loses around half the diff change from start to finish.

A simple example is that since 13-Jan difficulty has increased by 55.73% (yeah that's a scary number)
So if you were mining from say 10-Jan to now and then found a block, you'd actually lose over 25% of your expected reward!
All your shares from then to now have different PPS values, but your reward is based on 'now', the much worse PPS value.

This also affects us here to a lesser extent due to the fact we don't expect to find many blocks per diff change.
Thus why I'm trying to get that hash rate back up again.

Small pools like to pretend this doesn't happen

So basically this is a bad side effect of the N in PPLNS stretching over difficulty changes?

No, not exactly.
It's the bad side of blocks stretching over multiple difficulty changes.
For the N in PPLNS, it affects people ramping up or down.

Once you've ramped up, your reward 'matches' your hash rate when the block is found, since that is (expected to be) your hash rate for the last 5Nd also.
The variance the 5Nd deals with is during changing hash rates.

If you have the same hash rate (forever) then the 5Nd makes no difference to you after you've ramped up, until you ramp down.