Quote from: jhoren on April 13, 2018, 09:42:46 AM

Quote from: kano on April 12, 2018, 11:25:52 PM

Quote from: jhoren on April 12, 2018, 10:38:19 PM

Quote from: wavelengthsf on April 12, 2018, 10:18:03 PM

Quote from: jhoren on April 12, 2018, 09:08:08 PM

Thank you ccgllc. I think my main question might be how ramp down/up with five hours of off time might affect it. I don't think it would be noticeable but I wonder how I might be able to calculate that.

If you turned your miners off for 5 hours every day, they'd be mining 19 hours. That's roughly 20% less hash rate. You can take your hash rate and subtract 20% from that and run your calculations. The ramp up and ramp down won't really matter.

Thank you. I ran my calculations basing it on the total hash rate of the pool for the week and then of my % of that.

With a single machine there's about a $2.25 difference per week at the current BTC-USD rate. Maybe worth it if I get more machines running.

I ran it also if I ran it during the very expensive portion of the peak time and I would lose $.25 per week there. So, I could probably talk myself into giving the pool that hash and not stop it. But I'd still be saving using the peak/off-peak instead of just the straight cost.

Also to work out what to expect to lose from mining 80% of the time, as stated above that's 20% less expected reward.

How the ramp down and up works is that instead of losing 20% on the next block, and of course the luck of single block finding has very high variance, you instead lose it over 5Nd - i.e. you expect that 20% loss spread out over 5Nd after the outage - or 4% each 100% expected block
If it's a daily thing, then you should see the rewards settle close to the expected 80% each block reward as the daily 4%s overlap.

Thanks Kano. And it isn't a true 20% as it was only M-F, I would still be up 143 of 168 hours. But those 25 hours are so miniscule it may not be worth the hassle.

So your expected reward is 143/168 = ~85% of the 100% expected reward for your hash rate.

'Ramping' up and down isn't from having your reward reduced, it's from spreading it out (over 5Nd)
So you really do expect whatever fraction of the time you are mining as your expected reward %

This actually also works a lot better than on slush.
On slush the variance due to reduced mining or changing your hash rate (up or down) is actually higher than here due to the fact that the N (in PPLNS) is so short or slush.

Of course our overall variance is higher here due to the lower pool hash rate, but the variance affect on your reward due to changing your hash rate here is a lot lower than on slush.