Topic: [1050 TH] BitMinter.com [1% PPLNS,Pays TxFees +MergedMining,Stratum,GBT,vardiff] - page 232. (Read 837101 times)

New beta version of BitMinter client is out, with BFL ASIC support.

Here's a video of it with a BFL Jalapeno: https://plus.google.com/u/0/107253423193517853418/posts/PQuAMcQnq6g

Eh what?

The network hash rate affects your expected earnings? How? Details please ...
Difficulty obviously does, but the network?

 thank you,    I can see that my idea correlates with method number 2. Although your explanation is quite clear compared to my effort to express the same thing.

 Yeah variance in a short time period can be quite a lot. I have seen 1 block in 14 hours and then 14 blocks in the next 14 hours.

Two ways to calculate your expected earnings per second

1. (best method) yourhashrate / 2^32 / difficulty * bitcoinreward
2. yourhashrate / networkhashrate * network blocks per second * bitcoinreward

You were forgetting how blocks per second changes when the network hashrate increases and difficulty stays the same. And don't forget the part variance plays.

 


while typing my answer.   you mention hashrate and difficulty as the kings.

   not to disagree  but I don't still understand why our total network share does not affect our blocks.  

 seems to me 6 out of 70 will do better then 5.9 out of 80.  I  still fail  to understand and have tried to understand why this ratio does not apply in some way.  

Bitminters luck was better than median this week, and although the previous three weeks were worse median they weren't as bad as BTCGuild or Ozcoin.

Not sure if your idea is correct.
 A lot of people myself included struggle with this:

 1)the current difficulty is set still for 11 days.
2) the current network hash moves up and down constantly
3) bitminters hash moves up and down constantly

 I have been told by some heavy hitters that bitminters  Hash to current difficulty is what matters in assessment of luck. The rest of the network does not affect what bitminter should find in blocks in a day.  I don't quite understand this.  Right now we are hashing 5.5 out of the networks 74. and the difficulty is 8.9 mill.  While the diff has been at 8.9 the network hash had as much as 85Th hash and bitminter has has as high as 6.05th hash.

  5.5 divide by 74 is  .0743   our share of the network. difficulty is 8.9 mill

 6.05 divide by 84 is  .072   our share of the network. difficulty is 8.9 mill  

  so looking at the two examples I think the way the luck is calculated right now uses our hash against the difficulty. this would mean we should get more blocks with the 6.05  since it is larger then 5.5 and the difficulty is  constant.     While I see the consistency in this I think it is not accurate  since our network share is better  with the smaller  hash I would think we would get more blocks with   that number.

 Look at it from another viewpoint.  Lets say BTC guild the biggest pool had a technical issue crashes and the crash lasts 1 day. lets say it even affects the miners in it and they are all locked out for a day.  so for this day we are at 6th the network difficulty is set at 8.9 mill   but the network now has  40Th not 74TH.  so our share is 6 divide by 40 = .15     For this day I feel we would get more blocks then the day before when the network had 74Th and we had the same 6th.  I have been told this is not true since the hash of us is still 6 and the difficulty is still 8.9 mill.

But my training is accounting  not a programming or statistical degree.   I can say this 2 or 3 times when the network had a big drop off in hash and we did not drop we had great luck.

You know the theory, so you've answered your own question. Resetting work (or bitcoind, or the server itself for that matter)  has no effect on the probability of finding a valid block. The chance to find a block depends on only 2 things: hashrate and difficulty.

I was thinking about bitminter's recent bad luck. Could it be because only on bitminter LP resets the work, every time a new nmc block is found, for bitcoin work as well? For example, on ozcoin you could have an option to ignore LP on namecoin blocks and continue working with "old" work until new block is found on bitcoin network. In theory, this shouldn't affect the chances, but in practice bitminter has a negative luck for some time now...or always below expected if you look at the reward graph on bitminter's website.

 the almost 14 hour  block# 232949 time   13:57 =    1 block

  the next almost 14 hours   blocks # 232957 to 233037  time  14:26 =   13 blocks     


   
This comes to 14 blocks in 28 hours and 23 minutes.  Just about 1 block every 2 hours and 2 minutes.  We should be around 1 hour 46 minutes to 2 hours a block based on our hash against difficulty of 8.9 mill.

My point is it averages out both sets of blocks   are way off one is very bad luck one is very good luck. When averaged very close to normal numbers.

Can you give us ETA for Litecoin pool and miner . July - August - September ?

Mining bitcoins is like going through billions of lottery tickets. There's nothing to abandon, we're not staring at the same lottery ticket all this time. We just had some bad luck playing the lottery today.

Each hash your computer does is one lottery ticket. There is no progress towards completing a block. If your lottery ticket is not a winning ticket, then you throw it away, it's completely useless.

Pool hash rate was probably much lower then?  At any rate, I agree that it is what it is: variance.

Quit panicking, this is variance. We've seen blocks last much longer in the past. There is nothing wrong with the pool. Miners of old can tell you about blocks lasting 48 hours.

There's no point. You have the same chance, in the next second, to solve a block that you have already been working on for 13 seconds as one you've been working on for 13 hours. Or 13 days for that matter. The chance of finding a solution within a given time window depends solely on pool hashrate and current difficulty; nothing else. 'History' plays no part.

Now, if something was suspected to be wrong with the server, then that is a different story and a restart can make sense.

Can a pool abandon a block?  13 hours and counting, we're missing out on lucky and unlucky blocks of shorter duration...

no it can't hit 100%    99.99999999999999...... and on.

ever hit 100% before on the CDF? 

If pool hash rate doesn't increase proportionally to difficulty increases, there will be longer and longer blocks.

M

what's going on now, just really bad luck?  block has been 10 hours