Topic: What about bad pool luck? Or maybe it's something else? (Read 2076 times)

The luck was pretty good on BTC Guild before this bad streak. It does suck but I will ride it out and hope that it swings back the other way just as far.

Well BTC Guild's performance over the past week has been pretty decent, so if there was some sort of withholding attack it is being mitigated by some very good luck right now.

Great stuff. Gotta say. Keep them coming.

When bad luck strikes for a long period you start thinking changing pools
I did start with Slush then another pool and then Ghash where am mining now
I even made a comparison put 4 antminers on Guild and 4 on Ghash
There was a difference small but it was there.Plus that Guild has fees and usually bad luck
Ghash are not saints they forget to confirm block they have connection problems
and more
But its a jungle...from one side of the river and from the other side is the chinese asic companies
most probably will be eaten from the wild beasts......

Well if you look right now they're 10% over expected earnings for the past day.  If they get another block before next shift they'll jump up close to 20% over.

Bitcoin mining statistics only really make any sort of useful sense over a period of many weeks, if not months. The normal statistical variances even over a period of 2016 blocks are surprisingly wide. I've published a lot of data on hashing stats over at http://hashingit.com.

Being able to switch between Windows and Linux variants was necessary for GPU mining.  For ASICs you can use a low end windows laptop or even a Pi.

You don't.  I'm not sure what bozo333 was getting at with that.

I hope you have a lot of spare time on your hands because the whole thing is pretty complex, yet fascinating! Prepare for a steep learning curve in a bunch of related areas like electrical wiring, linux programming, networking, juggling various operating systems, types of pools, trading currencies etc. Google is your friend. And I would suggest you start reading in forums....a lot... no point in posting totally basic questions IMHO :-)

I'm new to this and just trying to find out how Bitcoin works.  When you say "Block header + nonce" does that mean that the two things are concatenated in this order: block header first then nonce.  How is the nonce incremented?  Is it done in some random way?  How is the the work divided up in pool mining?  How do you coordinate things so that people don't try the same nonce?  These things may be simple to you folks, but I'm thinking of trying to mine and want to know how this thing works before I invest $$$$ in it.

Thanks for putting that into scientific terms. 

was wondering the same thing.

BTCGuild payouts are complete garbage right now.

Is it possible? Yes.  Is it likely? No, and the proof is in the math.

Now obviously the chips and mining equipment aren't 'equal'.  Some will be faster (meaning it can make more attempts per second), some will be more energy efficient (meaning each attempt takes less energy), and some will have fewer hardware errors (meaning, well, fewer errors), and if you're lucky you've got one that is all three.

But for any single attempt, a BitFury is just as likely to get a good share / find a block as an Avalon, or a Bitmain, or Minion, or RockerBox or insert-chip-here.

The reason for that is that the output of sha256(block header + nonce) is unpredictable.  I.e. you can't tell the chip to only generate results that are within a particular region.  It just gets 'a' result and either that's below the threshold, or not (thus no shortcut).  If a chip somehow failed to return results that are below a threshold, the chip would need to actually have pathways in it that specifically single those out, either by design or by some very peculiar quirk (thus the 'possible, but unlikely').

If a chip can't scan all possible nonces (e.g. BitFury's hard coded nonce bits), then at some point well into the future there may well be a block that such a chip cannot solve.  Thankfully, by then, you probably shouldn't be using such a chip anymore anyway - if for no other reason than that you're likely no longer alive

There could also be the possibility that there is equipment out there can't actually solve a block now? What proof is there that all equipment is capable of achieving a best share in the billions and all chips and mining equipment is equal? We are told it can? But can it really?If so where's the proof?We are just left with X companies word?Many companies could of very well just made short cuts because no one would be able to prove they did and chalk everything up too luck? sha 256 only uses 128 bit memory bus @ 2^128 which is limited to about 16 Billion as a 32 bit address is just approximately only 4 billion @ 2^32 so  it is mathematically impossible for it too use every nonce numbered 1 to 16,800,000,000 at full capacity already it can't use every nonce so a short cut is already made somewhere?.
 I have read that there is equipment that is unable to use every nonce? This would make sense as a limiting factor?This might be the limiting factor on much of the equipment today? As earlier mining equipment was built with most likely a use of a limited amount of nonce's as it only needed to solve a best share in the millions and not expected billions?This very well can be the deciding factor of who solves a block or not? and why there is a lack of block finding all around? Because it seems to be clear cut across the network and not just a single pool.
 We should have a list of all the mining equipment and compare all companies and  the Best Share achieved on each equipment and which has solved a Block or not and with what if they did? I think then we will get some interesting  clear answers and find out some equipment might have never solved a block or achieved a best share in the billions?There might be companies that sold faulty equipment or just made short cuts on the nonce's the equipment can use?So now at this difficulty it is incapable of solving a block with difficulty so high? Something to think about it? As could very well be the possibility it's not the pools but the equipment we use?

Well yes it sucks, but what can you do other mine somewhere else.  You can't get mad you bet on red and the roulette wheel does black 8 times in a row (essentially another pool find more of the 2016 blocks then your pool).

You can't eliminate variability.  The only issue is whether o not something malicious is occurring at the pool either knowingly or unknowingly.  Most pools have gone through this at one point or another.  Back when difficulty was 9 million Ozcoin had a block that took 81 million shares, 9 times difficulty!!!  That was what caused a lot of people to leave Ozcoin, but the pool still survives despite a theft of almost 1000BTC at one point.

Eleuthria has nothing to gain really if his pool has bad luck - he makes more if it has good luck.  I'm sure he would love to see the number go back to neutral or above, but you can't will it.

Some people will jump onto the pool thinking it has had so much bad luck that it is due for a really good luck streak.

Thanks for the correction...
But even making 20% less than expected for 2 weeks sucks, or does it not? While I am new to mining, I have 20 years experience with a hedge fund. Will talk to our risk management guys about this. I know this much, that we would go through our trade systems with a fine tooth comb if our performance swayed that much.

Again, you are incorrect.  Please look at the table correctly.  2 week luck is at 80%.  1 month is at 89%.  The pool is now down to 6% of the network.  The pool is run by Eleuthria who actually limited/stopped his growth when he got close to 40% just so as not to scare people with 51%.  Ghash.IO hasn't bothered to inhibit its growth at all.

BTC 1 month luck is 80%....really not acceptable in my opinion. Can make the diff between making or losing.

Well I switched that hash power over to slush alone and my payouts are more in line where they should be.

Anybody already come up with the 'optimal' allocation between pools?

The bolded part is not true.  Do the math yourself or check with Organofcorti.  It falls within a normal confidence interval.

That being said yeah it's not pleasing.  If you flip for who pays lunch and you lose 2 out of 3 times do you say the coin is rigged?

There was a known withholding attack that ended a month ago and that messed up the 3 month stats.  It is possible that somebody is repeating the attack but it is almost impossible to detect until after the fact.

What's funny is that when BTCGuild had very good luck last year people were suspecting they were somehow cheating.

If you're not happy you can try Slush or Bitminter.  You luck may or not follow, depending on how superstitious you are.