giszmo
Legendary
Offline
Activity: 1568
¡ɥɔʇɐʍ ʇsnɾ 


June 24, 2011, 03:29:55 PM 

Hi, I've been working on that question and have a result that is far off the result previously found in the wiki: My assumption: network growth is 2%/day My result: 70% to never mine a single block ever. Old assumption: difficulty growth is 10%/iteration (which is far more optimistic than my assumption) Old result: 96% to never mine a single block ever. (which is a magnitude more pessimistic than my result) Who's interested in proof reading? Who has another, maybe simpler solution to provide? The Wiki link is here: https://en.bitcoin.it/wiki/Why_pooled_mining#Alternative_Perspective






Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.




HideousBeastManGuy
Jr. Member
Offline
Activity: 56


June 24, 2011, 03:38:45 PM 

There's simply no reason not to join a pool, especially with such a low GH/sec.
Hell, if you had even 50GH/sec you should probably join a pool since many are feefree. That said, if I did have 50GH/sec, I'd probably balance it across several pools for downtime issue and to iron out the small amount of variance.




Adam


June 24, 2011, 03:42:42 PM 

It sounds reasonable under your parameters, but there's no way I see 2% growth continuing longer than a few months, but I'll go with it. I just did some really quick math and here's what I got:
Chance of solving with 1 hash today: 1.68814E16, *500mhash = 8.4407E08 of solving in 1 second * 60*60*24 = .007293 chance of solving a block today.
I multiplied the odds of solving with 1 hash by .98 in excel, did the same calculation, and continued out for 50,000 days, although excel stopped recording values after 33000 days because at that crazy growth rate you would have a 9.9x10295 chance of solving a block in a day.
The sum of all the chances of solving a block in an individual day was 36.46%. There's probably a more elegant solution to this, but I just ran the calculation for another purpose and already had the spreadsheet.




giszmo
Legendary
Offline
Activity: 1568
¡ɥɔʇɐʍ ʇsnɾ 


June 24, 2011, 03:54:41 PM 

There's simply no reason not to join a pool, especially with such a low GH/sec.
Hell, if you had even 50GH/sec you should probably join a pool since many are feefree. That said, if I did have 50GH/sec, I'd probably balance it across several pools for downtime issue and to iron out the small amount of variance.
No reason to not use a pool?? Pardon me? How about pools having the potential to attack the bitcoin network as a whole? see https://en.bitcoin.it/wiki/Talk:Weaknesses#Pools for details. It sounds reasonable under your parameters, but there's no way I see 2% growth continuing longer than a few months, but I'll go with it.
Well ... my 2%/day are simply how performance was for 2 years now. Sure there are limits and at some point we will approximate Moore's law (that is when all PC's do mining I just wanted to take somewhat equal assumptions as the other authors of the article. The sum of all the chances of solving a block in an individual day was 36.46%.
although I would not vouch for your way of getting there, your solution is also much more in favour of mine. Thanks for looking into it.




TraderTimm
Legendary
Offline
Activity: 1792


June 24, 2011, 03:59:52 PM 

Although this should be asked in the 'Mining' subforum, I'll bite.
If you believe in reinforcing the structural components of bitcoin, you'll mine solo.
If you have no patience, you'll mine in a pool, but sacrifice the peertopeer robustness that solo mining provides. You're essentially at the mercy of pool downtime, not to mention contributing to a potential 51%+ attack scenario, or being a victim of a pool DDoS attack from an outside source.
Choose wisely...

fortitudinem multis  catenum regit omnia



HideousBeastManGuy
Jr. Member
Offline
Activity: 56


June 24, 2011, 05:08:49 PM 

I stand corrected, but I do not see that as a big enough threat to avoid mitigating a gross amount of variance if you have only Mh/sec. And as I explained, even if you have more computing power than that you could also spread yourself over several feefree pools to iron out variance even more, as well as mitigate risk for the chance of the described scenario.




Mousepotato


June 24, 2011, 07:09:48 PM 

It sounds reasonable under your parameters, but there's no way I see 2% growth continuing longer than a few months, but I'll go with it. I just did some really quick math and here's what I got:
Chance of solving with 1 hash today: 1.68814E16, *500mhash = 8.4407E08 of solving in 1 second * 60*60*24 = .007293 chance of solving a block today.
I multiplied the odds of solving with 1 hash by .98 in excel, did the same calculation, and continued out for 50,000 days, although excel stopped recording values after 33000 days because at that crazy growth rate you would have a 9.9x10295 chance of solving a block in a day.
The sum of all the chances of solving a block in an individual day was 36.46%. There's probably a more elegant solution to this, but I just ran the calculation for another purpose and already had the spreadsheet.
NEVER TELL ME THE ODDS!  Han Solominer

Mousepotato



wareen


June 24, 2011, 07:23:28 PM 

I stand corrected, but I do not see that as a big enough threat to avoid mitigating a gross amount of variance if you have only Mh/sec.
What's with all you trying to avoid every bit of variance  it is the very essence of what makes life interesting If I had to feed my kids from my mining income then I'd probably think differently but I hope nobody of you is in that position. I for my part like to be surprised once every few days weeks months, therefore I mine solo




Sukrim
Legendary
Offline
Activity: 1974


June 25, 2011, 12:19:11 AM 

My result: 70% to never mine a single block ever.
If you take "ever" as in "in eternity", you have a probability of 100% to mine a block, hell, you're even 100% likely to mine 1000 blocks with only 1 hash per second! Please specify a REAL timeframe like "within the next 5 years", not infinity!

https://bitfinex.com < leveraged trading of BTCUSD, LTCUSD and LTCBTC (long and short)  10% discount on fees for the first 30 days with this refcode: x5K9YtL3ZbMail me at Bitmessage: BMBbiHiVv5qh858ULsyRDtpRrG9WjXN3xf



fitty


June 25, 2011, 12:30:48 AM 

Hi, I've been working on that question and have a result that is far off the result previously found in the wiki: My assumption: network growth is 2%/day My result: 70% to never mine a single block ever. Old assumption: difficulty growth is 10%/iteration (which is far more optimistic than my assumption) Old result: 96% to never mine a single block ever. (which is a magnitude more pessimistic than my result) Who's interested in proof reading? Who has another, maybe simpler solution to provide? The Wiki link is here: https://en.bitcoin.it/wiki/Why_pooled_mining#Alternative_PerspectiveThe bigger issues is how many coins would you get (50? 25? 10?) and what will the value be in 6 months, 12 months, 24 months. If you're worried about 50%+ pools, just join a smaller pool. Doesn't that balance out the danger? As long as you're not in the biggest you're a counter balance to them in terms of the network.




TheVirus


June 25, 2011, 12:39:57 AM 

My result: 70% to never mine a single block ever.
If you take "ever" as in "in eternity", you have a probability of 100% to mine a block, hell, you're even 100% likely to mine 1000 blocks with only 1 hash per second! Please specify a REAL timeframe like "within the next 5 years", not infinity! Not true as there's a finite number of bitcoins.

Donations welcome: 1H8Pj3qYqfzxqgHzMLLsL6hWGEN88QLdkb



Quantumplation


June 25, 2011, 12:41:03 AM 

My result: 70% to never mine a single block ever.
If you take "ever" as in "in eternity", you have a probability of 100% to mine a block, hell, you're even 100% likely to mine 1000 blocks with only 1 hash per second! Please specify a REAL timeframe like "within the next 5 years", not infinity! Not true as there's a finite number of bitcoins. But not a finite number of blocks. After we stop minting coins, there will still be new blocks with transaction fees.




TheVirus


June 25, 2011, 12:44:17 AM 

My result: 70% to never mine a single block ever.
If you take "ever" as in "in eternity", you have a probability of 100% to mine a block, hell, you're even 100% likely to mine 1000 blocks with only 1 hash per second! Please specify a REAL timeframe like "within the next 5 years", not infinity! Not true as there's a finite number of bitcoins. But not a finite number of blocks. After we stop minting coins, there will still be new blocks with transaction fees. True, but I was under the assumption he meant a block that produces a BTC prize, not a fee.

Donations welcome: 1H8Pj3qYqfzxqgHzMLLsL6hWGEN88QLdkb



cloud9


June 25, 2011, 12:48:50 AM 

Current total network hash from http:// is 11.11 THash/sec. Your share is 500 MHash/sec.
So your odds should be 11.11 x 1e6 : 500 , every 10 minutes or so.
That is 11,110,000 : 500 or 22,220 : 1 or 0.00453% , every 10 minutes or so. If there are 6 x 10 minutes in an hour, you should probably strike lucky after 22,220 / 6 / 24 = 154.3 days if the total network hash remains unchanged, and you do not have more than average bad luck?!?

Disclaimer: Postings of Cloud9 are only individual views of opinion and/or musings and/or hypothesisses. On a nonauthoritative, peertopeer public forum, you do not need permission from Cloud9 to derive your own conclusions or opinions, so please do. Calculations and assumptions to be verified.



Quantumplation


June 25, 2011, 12:49:31 AM 

My result: 70% to never mine a single block ever.
If you take "ever" as in "in eternity", you have a probability of 100% to mine a block, hell, you're even 100% likely to mine 1000 blocks with only 1 hash per second! Please specify a REAL timeframe like "within the next 5 years", not infinity! Not true as there's a finite number of bitcoins. But not a finite number of blocks. After we stop minting coins, there will still be new blocks with transaction fees. True, but I was under the assumption he meant a block that produces a BTC prize, not a fee. You don't get charged a fee for generating a block in the future. YOU get awarded the fee, so it's still a BTC prize.




Sukrim
Legendary
Offline
Activity: 1974


June 25, 2011, 12:53:32 AM 

My result: 70% to never mine a single block ever.
If you take "ever" as in "in eternity", you have a probability of 100% to mine a block, hell, you're even 100% likely to mine 1000 blocks with only 1 hash per second! Please specify a REAL timeframe like "within the next 5 years", not infinity! Not true as there's a finite number of bitcoins. Not true, as the 8 decimal places are just a convention but can easily be extended, so miliSatoshis etc. are being mined and probably even generated. There will never be 21 Million BTC (or more) in notinfinite time frames, only 20999999.999.....

https://bitfinex.com < leveraged trading of BTCUSD, LTCUSD and LTCBTC (long and short)  10% discount on fees for the first 30 days with this refcode: x5K9YtL3ZbMail me at Bitmessage: BMBbiHiVv5qh858ULsyRDtpRrG9WjXN3xf



TheVirus


June 25, 2011, 01:06:49 AM 

My result: 70% to never mine a single block ever.
If you take "ever" as in "in eternity", you have a probability of 100% to mine a block, hell, you're even 100% likely to mine 1000 blocks with only 1 hash per second! Please specify a REAL timeframe like "within the next 5 years", not infinity! Not true as there's a finite number of bitcoins. But not a finite number of blocks. After we stop minting coins, there will still be new blocks with transaction fees. True, but I was under the assumption he meant a block that produces a BTC prize, not a fee. You don't get charged a fee for generating a block in the future. YOU get awarded the fee, so it's still a BTC prize. I know, it was a poor choice of wording on my part.

Donations welcome: 1H8Pj3qYqfzxqgHzMLLsL6hWGEN88QLdkb



iya


June 25, 2011, 01:14:13 AM 

My result: 70% to never mine a single block ever.
If you take "ever" as in "in eternity", you have a probability of 100% to mine a block, hell, you're even 100% likely to mine 1000 blocks with only 1 hash per second! Please specify a REAL timeframe like "within the next 5 years", not infinity! Not true as there's a finite number of bitcoins. Not true, as the 8 decimal places are just a convention but can easily be extended, so miliSatoshis etc. are being mined and probably even generated. There will never be 21 Million BTC (or more) in notinfinite time frames, only 20999999.999..... Not completely true, either. At least according to the current code, the first block which does no longer award even a single satoshi, will be block 6930000 = 210000 * 33, mined in ~2140.




giszmo
Legendary
Offline
Activity: 1568
¡ɥɔʇɐʍ ʇsnɾ 


June 25, 2011, 03:28:09 AM 

If you're worried about 50%+ pools, just join a smaller pool. Doesn't that balance out the danger? As long as you're not in the biggest you're a counter balance to them in terms of the network.
How do you know the smaller pool is not under the control of the same mastermind?




