Blocks are paid across the 10 open shifts, in order to smooth out variance across the shifts/shares.  So each shift gets *approximately* 10% of the block reward.  It varies slightly because the cutoff is checked every 20 seconds, so one shift might be 2.00b shares even, another might be 2.07b shares.  The reward is weighted so that the block reward given to each *share* is the same across those 10 shifts.

The luck figures don't factor in the current round.  Additionally old rounds will roll off the luck figure sometimes leading to an inflated percentage being displayed (such as when there was a 22 hour block one day and the luck was showing something absurd like 190% because there were only 2 short blocks actually being used to calculate the 24-hour luck).

The pool uses variable difficulty with the same target rate as pretty much every pool.  Setting a minimum difficulty is useful for high speed miners (I'd consider anything above 100 GH/s to be "high speed" these days).  Some miners don't smoothly accept work restart messages, so when the difficulty is being ramped up by vardiff you might lose a little speed for the first 30 seconds to a minute or so.  It should be negligible, but I know some ASICs used to have issues when using default difficulty settings on pools because of how fast they were having to communicate share submissions.

If your worker is specifically for the 200+ GH/s miner, I'd recommend you put your minimum to 128.

BTC Guild is no longer 25% of *hash rate*.  It occasionally has days where it is above 25% of blocks, but nothing can be done to prevent luck.  Even a 10-15% pool can have a day where it shows up above 25% if they get a decent burst of luck.

You realize only one pool (GHash.io) is actually above 25% of the hash rate these days, right?

Don't forget that no exchange will actually send your funds instantly, they all wait days/hours, and large transactions are manually processed.  Additionally, a double spend is not only publicly viewable but OBVIOUS once it has successfully happened.  You'd never get any money from the exchange.

Yeah.  I think it's worse when you have a solid run of good luck then a really bad run of luck *and* a difficulty change.  It exaggerates the drop of earnings for miners who weren't aware that the previous 24-hours were above what they should expect.   Going from 100% to 50% of earnings for a day sucks.  Going from ~140% to 50% sucks more.

It was two really bad rounds (~4 hours each) in close proximity.  That is not unheard of, and it can be quite a bit worse.  The blocks were 7/8x difficulty worth of shares.  Pools have had 10x difficulty worth of shares to solve blocks before.  And as stated for nearly a whole page in the pool thread:  There is absolutely nothing wrong with the pool.  Luck cannot be controlled, and when you have two bad rounds close together it makes the luck for the nearby shifts take a complete dive.  At least the shifts on BTC Guild are long enough that even with those two huge rounds, no shares went unpaid, whereas on ghash.io and BitMinter it is possible to have a share get paid 0.00000000 BTC if rounds like these were encountered.

They're not separated from normal blocks since you don't know they're orphans when you first find them, and they're awarded to miners either way, so it doesn't exactly matter to the miner if its an orphan or not.

Some of the recent script updates have made it easier to match the orphans after the fact though, so I'll look into maybe italicizing the block number if it was an orphan.

UPDATE:  Orphans will now show up as 'Orphan' instead of a Block Number in the Block History (Pool Stats page).

Our orphan rate is the same/lower than other pools, and we've not had any of them in the last 48 hours.  Bad luck is completely out of anybody's control.  Additionally, the pool is overall lucky since luck began getting tracked beyond a single difficulty period.

People lose their heads when luck is bad, but nobody asks if the pool is broken when it's paying out 50-60% above what you'd expect, which is just as likely/unlikely as 50-60% below what you'd expect.

Anybody looking at luck figures to decide joining a pool doesn't understand what luck stats mean.

While technically variance is normal...this is more of the opposite of lottery-winning variance.  This is hit by a truck while witnessing your dog get struck by lightning variance.

Everything is running fine, it's just really shitty luck.  2x 4(+) hour blocks in close proximity pretty much nukes the blocks per shift for most the day.  A 4 hour block is (roughly) a 7x difficulty block.  These are not extremely rare, but it is quite rare to have 2 of them very close together.

I suppose this is payback for when Guild found 11 blocks in a row the other week.

Which stats page?  The PPLNS Stats page can be a bit load heavy due to the Luck Graph and 210 shifts of dat being loaded.  The Shift Detail view of PPLNS Stats is extremely heavy on the browser due to having over 11,000 rows (full shift breakdown by user) in the table that's being loaded then being paginated via JavaScript.

Dashboard and Pool Stats haven't changed and shouldn't be causing you any kind of unusual load times.

It was a 4 hour block.  Not common, but nothing wrong with it.



BTC Guild pays all orphaned blocks.  There is no difference between a stale block and an orphan block.  To the miner they're the same thing.  In both cases, the pool was not working on the dominant blockchain.  In both cases it was due to the pool not being aware of another block on the network at that same time.  You could argue an orphan might be slightly different because another pool might submit one after yours and that one ends up being dominant.  However, for a MINER, it shouldn't matter.  In both cases the pool solved a block and the miner should have been paid.  A "stale" block is just a sign of a poorly configured pool, and the pool mining a "stale" block is just as bad as an orphan in terms of lost income.

If you didn't use the pool's actual 24 hours for your numbers, you're WAY off, by nearly a factor of 2.  144 blocks * 25% = 36 blocks/day expected, which is 900 BTC.  90% of that is PPLNS and 10% is PPS (rough numbers), or 810 / 90.  810 x 2% (accounting for orphans) = 16.2 BTC.   90 x 6.5% (accounting for orphans) = 5.85.  Meaning if the pool is functioning normally, at perfectly neutral luck, it's ~22 BTC.  I did not include tx fees the pool collects because those are not a noticeable increase, and those are shared with miners.  Additionally, the pool *pays* transaction fees for all payouts which is not included.  Right now the numbers are a bit higher due to network growth.  If you want to factor that in, add an extra 20% or so.  Of course that assumes that the pool grows in unison with the overall network.

A percentage of that income is also required to be set aside as a buffer for bad luck (PPS), as well as the chance of a security breach, because unlike some pools *cough*50BTC*cough*, BTC Guild will actually honor its debts in the event of a compromise.

And his calculation was on one of the luckier days.  If he did the same calculation today he'd see I've *lost* money since last night.

This is actually doable on your own currently.  Make two workers, one for PPLNS, one for PPS.  Then when you setup your miner (cgminer/bfgminer), set up two pools for BTC Guild.  Setup one for the first worker, one for the second worker.  Set the management strategy to 'load balance'.  This will end up splitting your shares 50/50 between PPLNS and PPS.

BitMinter has had way more than 3 in the last 1600.  I'm curious now what they use to differentiate between "stale" and "orphan" considering those mean the same thing when it comes to blocks.  They've had 3 stale+orphan in the last 100, 5 in the last 200, 7 in the last 300.

Smaller pools will have more orphans than larger ones purely because they're less likely to be able to confirm their own orphaned block than a large pool.  It's also very common to see random events (1 in 100) show up in clusters, meaning a small sample size can show a much larger (like the 3/100, 5/200, 7/300), or smaller (0/300) occurrence just due to the distribution of random events.


UPDATE:  Guild has had 4 orphans in its last 600 blocks, 7 in the last 850.  That's not far off from the approximation of 1%, which is what has been estimated by most pools for years.

Subtract 1% from the 3% fee for orphaned blocks, which are roughly 1 in 100 blocks.  And PPS is barely over 10% of the pool now.


And don't "imagine what it was 2 difficulty retargets ago", because BTC Guild was not 2700 TH/s 2 retargets ago.

If you have earned 1 BTC, then I have made 0.03 BTC(ish, orphans screw that up a little since I pay you for those and end up losing a block worth of coins).  So you're off by a factor of 10 there.

The entire network solves 144 blocks a day, not BTC Guild.  BTC Guild is only about 25% of the network.  Additionally, BTC Guild pays orphans which means while I keep 3% of each block, about 1% ends up going back to the users when the pool has an orphaned block.   When the network is growing, it's a bit more than 144 blocks/day.  BTC Guild has to grow at the same percentage as the network in order to maintain a 25% share.