Manual withdrawals are your full balance.  Auto withdrawals are exact amounts.  This is pretty clearly stated in the support section.  Either mine above 0.01 and do a manual withdrawal for free, or above 0.001 and pay a fee.

Shifts are based off a certain number of shares.  When enough shares have been submitted to a pool, the current shift completes.

The 10 most recently completed shifts are "open".  Any time a block is solved, the reward is split evenly among all the shares contained in the 10 open shifts.  That is why the warning at the top states that the number are not final, all shifts in the top table will increase in earnings if a block is solved while they're still in that table.

When a new shift completes, the oldest shift is now closed, meaning it will not receive any additional rewards if a block is solved.

I'm not referring to raw network growth in this case, but specifically *where* it shows up.  I'm still expecting BTC Guild to shrink as a percentage of the network (same with Eligius, slush, BitMinter, etc.).  As posted earlier, "home mining" has become a very specific niche within the overall mining ecosystem.  By Jan 1st, I doubt most original pools combined will even be 15% of the network.  As it stands today, BTC Guild + Eligius + Slush + p2pool + BitMinter + Ozcoin + Eclipse combined are only ~16.5% of the entire network [based off the last 2016 blocks].

GHash had it happen a lot (not sure if it's changed) because they used a pitifully small 'N' value.  I know BitMinter has it happen during their more painful blocks as well.  It's the unfortunate part of PPLNS.  You have a tradeoff of a smaller N value for faster share maturation, or a larger N for slower maturation.  Statistically, it makes no difference in the long run.  However, if you sampled large N shifts vs small N shifts, you'd find that large N shifts were more likely to end up close to expected value on a per-shift basis, while small N shifts had a much larger range.  The average would be the same though.

Depending on how the network grows over the next few months, I may be doubling our shift times to target 2-hours per shift roughly instead of 1-hour.  This would be done by doubling the current 'N' value, which would put us back to the point we used to run at, where 'N' was approximately 10 times the network difficulty.  At that value, the odds of a 0-block shift are incredibly small.  You'd basically need a record breaking block (shares submitted > 10x network difficulty).  I think only one pool (Deepbit) has ever had a round break the 10x mark.

3, all within the last month.  2 of them were back to back (which is somewhat expected, it was a single bad block after all).  It was only recently that the network difficulty vs shares per shift was at a point where you can expect the occasional 0-block shift.

Hopefully not, we still have about an hour until a shift closes @ 0.  But considering the luck run this last week (20% above expectation), I'm not too worried.  We're still on track to probably get the 3-month luck back over expectation as well soon, unless we have a full week of pretty bad luck.

I've never pulled my punches when it comes to mining viability or the grim outlook for the "home miner" as we continue to move forward into the industrialization of Bitcoin mining.  I just hope we end up with enough players in the game to prevent full consolidation of mining power.  As far as I can tell, the future of Bitcoin mining will likely end up in a Cold War/Mutually Assured Destruction type arms race between a handful of ASIC manufacturers.

I'm still not sure how I feel about that.  I've always anticipated it, ever since I got into BTC in early 2011 and read up on how it works.  I just didn't expect it to take place for decades, mostly because I never anticipated exponential growth in Bitcoin's value, starting just months after I entered the game.  My only fear is that the rapid growth is taking us into that "end game" mining industrialization before enough companies have tried to establish themselves in that space.  The only major ones I can name are Spondoolies, KNC, Bitfury, Bitmain, and ASICMINER.  There's other companies out there making hardware, but I believe those are the only ones actually producing the chips, the rest just buy raw chips from those companies and assemble their own boards for them.  I didn't include BFL because they are perpetually a year behind the others, and a running joke at this point.

BTC Guild's pool software is 100% custom.  It is not related to slush's python stratum server in any way.  BTC Guild did not push a competing PROTOCOL against Stratum, because it was not worth fighting over a couple extra bytes of efficiency when comparing the protocol I was working on vs Stratum, when we were moving from the awful inefficiencies of getwork.


However, "home mining" was going to die in the end regardless.  You would have to be blind not to realize that since the inception of Bitcoin.  The entire system functions based off a zero-sum game of chance.  The more power one person obtains, the less everybody else will get as a result.  Once enough money was on the table, there would ALWAYS be large entities manufacturing their own hardware for pennies on the dollar and putting it online in the cheapest power locations in the world.  Home mining is a hobby, not a business or investment.


EDIT:  At the current stage, I'd define "home mining" as anything using regular consumer computer parts, or very cheap ASICs.  If you haven't invested a large amount into your setup, you're unlikely to ever turn a profit.  The only ASICs with a *chance* at turning a profit required significant upfront costs to get a reasonable $/GH value, which normally not only means buying the most expensive hardware, but buying in bulk in order to get further discounts.

No, actually Slush is pretty much the ONLY pool with an actual *penalty* to people that leave, by constantly degrading old submissions so they rapidly become worthless.  With PPLNS, the expected value of your share is the same whether you stop mining the second you submit it or you keep mining.  With PPS, your share has a pre-determined value.  With Eligius' CPPSRB, it's the same.  Those are really the 3 dominant payment methods used by all major pools these days, and all 3 are anti-hopping by being unhoppable, not by trying to penalize users if they leave the pool.

EDIT:  And for the post below:  Slush's score system is hoppable.  It's just not hopped much because the profitability of pool hopping is virtually 0 when there aren't multiple large hoppable pools, so nobody bothers anymore.

You need to look at mining hardware/mining support forums.  Errors are a local problem, they have absolutely nothing to do with the pool.

There's no such thing as progress towards a block.  The current block we're mining is whatever the most current block on the network is...


EDIT:  This pool is PPLNS.  You're constantly asking questions under the assumption that we're paying out based out on single blocks/rounds.

We did.  Last time it was 2x 0 shifts in a row due to a block that was ~6.5x difficulty.  This one was ~6x.  Luckily we've been having pretty good luck overall the last week, to the point that a 0-block shift didn't even take our 3-day average under 100% (121% as of this post).

No, that will not be happening.  Most sites with "Pending" balances mean they're showing you blocks which haven't confirmed yet.  BTC Guild credits you as soon as a block has 2 confirmations instead of waiting for 100 or 120.

If you're referring to 'Estimated' amounts...I'm not adding that because many people are too stupid to understand what 'Estimated' means.  BTC Guild solves a block on average every 2-2.5 hours.  Sometimes we get 3-4 in an hour.  Sometimes we take 8-12 hours on a really bad one.  Even a record breaking bad luck block would not take a full day.  If that is causing you too much anxiety, you shouldn't be mining at all.

Then you should turn off auto withdrawals and just manually withdrawal instead.  Auto withdrawals assume the miner is staying on the pool (thus the need for automated payments), in which case fractional BTC is of no concern, they'll just get their next payout that much faster.

No.  The filtering server *does not record shares*.  You only connect to it for a single submission, after which point you're redirected to real servers.  This old cached work on the filtering server had no impact on stats.



Only the filtering server had the issue.  All the *real* servers were on the current block the entire time.  You "saw a new block" once you submitted a valid share and were redirected to the proper servers.

The filtering server was having some bitcoind restarts the other day.  When the filtering server's bitcoind goes offline, it loads older cached work.  Since the filtering server is only there until you submit a single share, it's not exactly a big issue if it is giving out old work, the whole point is just to make sure you're a legitimate miner before sending you to the real servers.

Shift lengths have been increased slighlty since the pool has gained back some speed (they've gone from 10b -> 11b a few days ago, and now 11b -> 12b shares).  As usual, this makes no change to what you earn, it simply makes it take a little longer for a share to go from open shifts to a closed shift (so longer ramp-up time, and a longer wind-down time when you stop mining).  They should still mature in a little under 12 hours.



Earnings per block is a completely pointless number since the larger the pool, the smaller your earnings per block but the more frequent they come in.  Just use any BTC calculator for 1600 GH/s for an idea of what your earnings would be at neutral luck, and subtract 2% for the pool fee.  Don't look at your 24h earnings until you have mined for > 36 hours, since your first 12 hours are paid less due to ramp-up time (and when you stop you get paid for 12 hours even though you stopped, so it ends up offsetting the ramp-up).

A little over 3 weeks left before the servers disappear for ScryptGuild.  Again, if you haven't already, make sure you either withdraw your coins or click the donate button.  After the servers go down, any unclaimed coins will be lost.

In other words, their built in stratum controller doesn't work with BTC Guild because they didn't implement client.reconnect, joy.  It works just fine if you direct connect to the redirect URL?

Time Completed is based off the timestamp of the block, which is not always the actual time it was found due to nTime rolling and the fact that work is only pushed out every 30 seconds (and older work is still valid unless a new block is solved).  Blocks are also only checked for once per minute, so a block found at 4:59:05 AM CST (but with a timestamp of 4:58:37) would get caught in a shift that ended at 5:00 AM CST instead since it would not be seen until after the shift close.