[4+ EH] Slush Pool (slushpool.com); Overt AsicBoost; World First Mining Pool

[eleuthria]

2) and 3) is not true until difficulty adjusts - so you earn less in the long run, but until the next retarget everybody just earns more, because due to increased network hashrate, blocks will be found more frequently than every ten minutes.

Sorry I beg to differ. Difficulty is simply a reflection of hash rate - not the other way around. Hash rate is the master - difficulty comes into play in an attempt to make 1 block be found every 10 mins. If someone plugged in a 100 TH miner into the network then it would take a while for difficulty to adjust - but your reward would go down straight away.

Incorrect regarding rewards, at least directly.  You would still see the same rewards until difficulty adjusted.  You may be mixing up reduces rewards per round (which would happen if pool speed increases), but that's the same net reward for the difficulty since more rounds would complete.

The only immediate reduction in rewards would be the result of orphaned blocks and stale shares.  When the network is producing blocks faster than it should, the rate of orphans and stale shares is higher.  This is a marginal decrease though (fractions of a % decrease in rewards for the remainder of the difficulty).

[Trongersoll]

2) and 3) is not true until difficulty adjusts - so you earn less in the long run, but until the next retarget everybody just earns more, because due to increased network hashrate, blocks will be found more frequently than every ten minutes.

Sorry I beg to differ. Difficulty is simply a reflection of hash rate - not the other way around. Hash rate is the master - difficulty comes into play in an attempt to make 1 block be found every 10 mins. If someone plugged in a 100 TH miner into the network then it would take a while for difficulty to adjust - but your reward would go down straight away.

Incorrect regarding rewards, at least directly.  You would still see the same rewards until difficulty adjusted.  You may be mixing up reduces rewards per round (which would happen if pool speed increases), but that's the same net reward for the difficulty since more rounds would complete.

The only immediate reduction in rewards would be the result of orphaned blocks and stale shares.  When the network is producing blocks faster than it should, the rate of orphans and stale shares is higher.

Oh sure, just what we need, someone who actually knows what he's talking about. Who left the gates between the pools unlocked? heheheheheh

[trasla]

There is no "starting all over".
Its not like you calculate the block share by share, and after some time its done.
Each of the 336.000 Hashes your USB Block Erupter calculates per second could be a complete valid block.
Of all those hashes, 5 per minute are relayed to the pool, because they satisfy difficulty one.
Any of those could by chance start with many more zeros, and thats a valid block at current difficulty.

But chance does not care how many valid blocks have been found since the old shares you submitted or something like that.
You always work on the current block, and its always a very small chance per share to be the valid block hash.
And it doesnt matter whether you have worked on the same block before or not.
And it does not matter how fast any other miner does find new blocks, as long as you get notified about it fast enough.
The low number of orphaned blocks say we do get notified fast enough, and thus network hashrate does not matter at all for our income, until difficulty rises.

[nottm28]

Why should my reward go down?
The frequency in which i find blocks with a given difficulty does not depend on the network hashrate in any way.

If that 'someone' suddenly plugged in their 100 TH/s miner into slush's pool, you would see a smaller share of the pot - result - reward goes down. If they plugged it into another pool then slush's pool would find blocks less frequently (more competition) - result - reward goes down.

[EDIT] bottom line is hash rate rules, difficulty follows in order to try and keep 1 block found every 10 mins.

[trasla]

Why should my reward go down?
The frequency in which i find blocks with a given difficulty does not depend on the network hashrate in any way.

If that 'someone' suddenly plugged in their 100 TH/s miner into slush's pool, you would see a smaller share of the pot - result - reward goes down. If they plugged it into another pool then slush's pool would find blocks less frequently (more competition) - result - reward goes down.

1) Reward per block found by the pool goes down, but number of blocks found goes up -> reward stays the same.
2) No, how many blocks slush' pool finds does not decrease when some other pool gets more hashing power, see the post i just wrote before yours.

If there was some mechanism to make our pool find less blocks if some other finds more, we would never need a difficulty adjustment.

There is nothing like "more competition".
You are rolling dice in a casino. You get a dollar on every 5 or 6.
If a thousand other people start doing it too, you still earn the same money.
Only when the casino decides to only pay out on 6 and no more on 5 (difficulty adjust), you suddenly earn less.

[Trongersoll]

Hey nottm, we might as well give up, he just doesn't get it. and we won't convince him that there is more to it than just probability,

[nottm28]

Why should my reward go down?
The frequency in which i find blocks with a given difficulty does not depend on the network hashrate in any way.

If that 'someone' suddenly plugged in their 100 TH/s miner into slush's pool, you would see a smaller share of the pot - result - reward goes down. If they plugged it into another pool then slush's pool would find blocks less frequently (more competition) - result - reward goes down.

1) Reward per block found by the pool goes down, but number of blocks found goes up -> reward stays the same.
2) No, how many blocks slush' pool finds does not decrease when some other pool gets more hashing power, see the post i just wrote before yours.

There is nothing like "more competition".
You are rolling dice in a casino. You get a dollar on every 5 or 6.
If a thousand other people start doing it too, you still earn the same money.
Only when the casino decides to only pay out on 6 and no more on 5 (difficulty adjust), you suddenly earn less.

1) Only if the 100TH/s miner was an existing miner somewhere else (i.e. moved his processing power from somewhere else to slush). If he was a 'new addition' then your share decreases.
2) I don't see that - other pools gaining hash power and slush's pool remaining the same = less blocks found (by chance).

[trasla]

Hey nottm, we might as well give up, he just doesn't get it. and we won't convince him that there is more to it than just probability,

This time, you are partly correct, at least.
You would not convince me.


Really, guys, you did not completely grasp how it works.
If you want to, reread the explanations, or ask, or just try to explain to me just why you think it does matter.
If you feel like i am just a dumb guy, that okay, i can live with it.
My girlfriend has objections anyway: http://xkcd.com/386/

[nottm28]

Hey nottm, we might as well give up, he just doesn't get it. and we won't convince him that there is more to it than just probability,

Gave it one more go - look for organofcorti's posts - he really explains things well

[nottm28]

Hey nottm, we might as well give up, he just doesn't get it. and we won't convince him that there is more to it than just probability,

This time, you are partly correct, at least.
You would not convince me.


Really, guys, you did not completely grasp how it works.
If you want to, reread the explanations, or ask, or just try to explain to me just why you think it does matter.
If you feel like i am just a dumb guy, that okay, i can live with it.
My girlfriend has objections anyway: http://xkcd.com/386/

No worries my friend believe what you will - I like a strong mind.

[trasla]

1) Only if the 100TH/s miner was an existing miner somewhere else (i.e. moved his processing power from somewhere else to slush). If he was a 'new addition' then you share decreases.
2) I don't see that - other pools gaining hash power and slush's pool remaining the same = less blocks found (by chance).

Okay, lets do this step by step.
I dont get why you think the hashrate of some other pool should influence how often we find blocks.

Finding a block means calculating a hash of some data containing the past block header and new transactions since the last block, plus a random number (nonce), and the hash has to meet certain conditions, like starting with 19 million (difficulty) zeros. Do we agree?

EDIT: I have to correct myself, it has to start with a certain number of zeros, so its value is less than the target, but that number isnt 19 million (the hash is "only" a 256 bit number, and only has that many digits).

[nottm28]

1) Only if the 100TH/s miner was an existing miner somewhere else (i.e. moved his processing power from somewhere else to slush). If he was a 'new addition' then you share decreases.
2) I don't see that - other pools gaining hash power and slush's pool remaining the same = less blocks found (by chance).

Okay, lets do this step by step.
I dont get why you think the hashrate of some other pool should influence how often we find blocks.

Finding a block means calculating a hash of some data containing the past block header and new transactions since the last block, and the hash has to meet certain conditions, like starting with 19 million (difficulty) zeros. Do we agree?

Step by step is good for me...

I dont get why you think the hashrate of some other pool should influence how often we find blocks.

Let's say slush's pool combined power is 10% of the total network hash rate. One block (by design) is found every 10 mins. So, on average slush's pool should find a block approx every 100 mins. (could be 10 mins for one block, 20 mins the next, 5 hours the next - but the average is 100 mins).

Now, if the total network hash rate suddenly doubled and nothing was added via slush's pool - our combined percentage of the whole network would suddenly drop to 5% - so one block found every 20 mins (not 10).

Do you see this?

[trasla]

One block (by design) is found every 10 mins.
Do you see this?

There we go, thats the point.
This "by design" - how do you think it works?

It does via adjusting the difficulty. If an increase in hashrate lets the network find blocks more often, difficulty is increased to a value which leads to one block every ten minutes on average, with the current hasrate. If the hasrate drops, and the networks finds blocks too slowly, difficulty is lowered. So "every ten minutes by design" does not work instantly, but is periodically adjusted via the difficulty value.

[nottm28]

One block (by design) is found every 10 mins.
Do you see this?

There we go, thats the point.
This "by design" - how do you think it works?

It does via adjusting the difficulty. If an increase in hashrate lets the network find blocks more often, difficulty is increased to a value which leads to one block every ten minutes on average, with the current hasrate. If the hasrate drops, and the networks finds blocks too slowly, difficulty is lowered. So "every ten minutes by design" does not work instantly, but is periodically adjusted via the difficulty value.

Ok trasla, no worries, I tried

[trasla]

You want to give up?
If you dont agree, just explain to me how the network makes sure one block is found every ten minutes.

[nottm28]

You want to give up?
If you dont agree, just explain to me how the network makes sure one block is found every ten minutes.

Ok, difficulty is set to 'try' and maintain one block find every 10 mins. If suddenly the total network hash rate was doubled (due to a massive sudden addition of asics say). Then difficulty woud stay the same for a few days. Blocks would suddenly be found every 5 mins not 10 for a few days. Until the guys who develop the code say - hey 'we need to increase difficulty' - 'blocks are being found too often'. In the mean time, your reward will have dropped. I'm trying to explain that hash rate is the driver - and difficulty is the compensator...

[Trongersoll]

One block (by design) is found every 10 mins.
Do you see this?

There we go, thats the point.
This "by design" - how do you think it works?

It does via adjusting the difficulty. If an increase in hashrate lets the network find blocks more often, difficulty is increased to a value which leads to one block every ten minutes on average, with the current hasrate. If the hasrate drops, and the networks finds blocks too slowly, difficulty is lowered. So "every ten minutes by design" does not work instantly, but is periodically adjusted via the difficulty value.

Ok trasla, no worries, I tried

heh, this is like tag team.

I think that the problem is that we are looking at long term and you are looking short term. Yes, until the difficulty increase the time to find a block would decrease. And you would continue to find them at a constant rate. But, when he difficulty does go up, it will more than compensate for the previous "gift". At least i think it will. there are those that speculate that the date the last Bitcoin is mined is slowly moving closer.

[nottm28]

One block (by design) is found every 10 mins.
Do you see this?

There we go, thats the point.
This "by design" - how do you think it works?

It does via adjusting the difficulty. If an increase in hashrate lets the network find blocks more often, difficulty is increased to a value which leads to one block every ten minutes on average, with the current hasrate. If the hasrate drops, and the networks finds blocks too slowly, difficulty is lowered. So "every ten minutes by design" does not work instantly, but is periodically adjusted via the difficulty value.

Ok trasla, no worries, I tried

heh, this is like tag team.

I think that the problem is that we are looking at long term and you are looking short term. Yes, until the difficulty increase the time to find a block would decrease. And you would continue to find them at a constant rate. But, when he difficulty does go up, it will more than compensate for the previous "gift". At least i think it will. there are those that speculate that the date the last Bitcoin is mined is slowly moving closer.

I think the 'end date' is coming closer Trong. There's loads of debate about increasing the life span of btc - some complicated maths the likes of organofcorti and a few others understand...

[EDIT] trasla - please pm me - we need to take this offline. I don't wanna bloat this thread any worse - sorry guys

[trasla]

I think that the problem is that we are looking at long term and you are looking short term. Yes, until the difficulty increase the time to find a block would decrease. And you would continue to find them at a constant rate. But, when he difficulty does go up, it will more than compensate for the previous "gift". At least i think it will. there are those that speculate that the date the last Bitcoin is mined is slowly moving closer.

I completely agree - but thats exactly what i said before. If someone else adds hashrate, my reward stays the same _until difficulty adjusts_.


nottm28:

Okay, we are getting near.
Lets say we have two pools, each with half the hasrate (say, 10 GH/s each).
Both find a block every 20 minutes, totalling for one block every 10 minutes.
Now one pool adds 20 GH/s. Network hashrate is doubled, blocks are found every 5 minutes.
One pool has 30 GH/s, tripled hashrate, finds 3 blocks in 20 minutes.
The other pool stays at 10 GH/s, finds one block every 20 minutes.
In total, 4 blocks in 20 minutes -> one block in 5.

One pool gets same number of blocks with same hasrate, so same income as before for all its miners.
The other pool gets three times the blocks, but every miner in it now has only one third of the relative hashing power, so he gets one third of his usual rewar per block -> same income as before.

I agree with you, only "in the meantime, your reward will have dropped" i cant see.

EDIT: Well, yes, might be good not to spam the thread, of course you can reply via pm.

[nottm28]

I think that the problem is that we are looking at long term and you are looking short term. Yes, until the difficulty increase the time to find a block would decrease. And you would continue to find them at a constant rate. But, when he difficulty does go up, it will more than compensate for the previous "gift". At least i think it will. there are those that speculate that the date the last Bitcoin is mined is slowly moving closer.

I completely agree - but thats exactly what i said before. If someone else adds hashrate, my reward stays the same _until difficulty adjusts_.


nottm28:

Okay, we are getting near.
Lets say we have two pools, each with half the hasrate (say, 10 GH/s each).
Both find a block every 20 minutes, totalling for one block every 10 minutes.
Now one pool adds 20 GH/s. Network hashrate is doubled, blocks are found every 5 minutes.
One pool has 30 GH/s, tripled hashrate, finds 3 blocks in 20 minutes.
The other pool stays at 10 GH/s, finds one block every 20 minutes.
In total, 4 blocks in 20 minutes -> one block in 5.

One pool gets same number of blocks with same hasrate, so same income as before for all its miners.
The other pool gets three times the blocks, but every miner in it now has only one third of the relative hashing power, so he gets one third of his usual rewar per block -> same income as before.

I agree with you, only "in the meantime, your reward will have dropped" i cant see.

EDIT: Well, yes, might be good not to spam the thread, of course you can reply via pm.

Before:

Pool A, 10 GH/s, 20 mins per block, 25 coins earned
Pool B, 10 GH/s, 20 mins per block, 25 coins earned

Me in pool A, 1 GH/s, 20 mins per block, 2.5 coins earned

After: (difficulty not increased)

Pool A, 10 GH/s, 30 mins per block, 25 coins earned
Pool B, 30 GH/s, 10 mins per block, 25 coins earned

Me in pool A, 1 GH/s, 30 mins per block, 2.5 coins earned

In your scenario, my own earnings jump down from 2.5 coins per 20 mins to 2.5 coins per 30 mins...