Anti solo mining myths debunked

[googlebot1]

IMHO none of the anti solo mining myths are correct except one: pooled mining generates a steadier flow of income (at a price).

Myth 1: Mining in a modern pool produces less stale shares due to long polling, which the regular Bitcoin client doesn't provide for solo miners.

Fact: You don't need long polling for solo mining at all. Just set the askrate to something as 1 second. For solo setups the added overhead is below noise. You loose 0.5 seconds worth of work on average per freshly found block, but even the best pools with LP support cannot beat that in practice, since you seldomly get less than 0.5-1% stale shares (often more). This nonsense is even part of the wiki! There is no 1-2% benefit due to LP, in fact local fast polling is more reliable. Can anyone remove it?

Myth 2: Because difficulty constantly increases, joining a pool improves your chance to get any payback before the next difficulty increase, thus pooled mining has a higher profitability.

Fact: The premise is true, but not the conclusion. Solo mining has a low probability to get a large payout before the next difficulty increase. Pooled mining has a large probability to get a low payout before the next difficulty increase. It sums up to zero. Pooled mining does not increase your chances in a scenario of steeply increasing difficulty.

I think all other myths are self-explanatory nonsense.

There is not one benefit of pooled mining except less variance. You pay for this comfort with pool fees, missed transaction fees, stale shares, generally higher network downtime exposure, the risk of getting ripped off by a pool operator, implicit costs of sophisticated pool hopping (if you aren't the hopper). These costs are real, the costs of solo mining are purely psychological (not for the faint of heart).

[1714156942]

1714156942

[1714156942]

1714156942

[bcpokey]

This is possibly the most confusing topic title i've read to date:

Anti- meaning against, myth meaning fictional story, and debunked to disprove. A triple negative if you will. So if solo mining is the subject, a triple negative would indicate a negative, or be a post against solo mining.


Anywho, that's just a bit of an aside. On the main topic, variance is a cruel cruel cruel mistress. in roughly the last 2 months (2 month anniversary in a week! Better buy some roses!) of mining i have solved... 0 blocks. Clearly i'm a ridiculously unlucky person, but it still stands that i'm probably not alone, and that as difficulty increases the lottery you play becomes that much chancier. The fact is that probability isn't as nice and neat as some people like to make it out to be. Sure if you looked over the course of 10 years of mining my variance might average out to the statistical norm, but in an acceleratingly competitive climate, can one really afford to wait that long?

The fact is that difficulty changes, and with it your chances for hitting a block, you have a "low probability to get a large payout", but it also keeps gettng lower. You can keep missing and missing and missing and so on. Really unless you can put out some serious hash power or have great faith in your luck, anti-solo isnt much of a myth.

[JoelKatz]

bcpokey: In other words, reduced variance is the only advantage of pooled mining.

[Fakeman]

This is possibly the most confusing topic title i've read to date:

Anti- meaning against, myth meaning fictional story, and debunked to disprove. A triple negative if you will. So if solo mining is the subject, a triple negative would indicate a negative, or be a post against solo mining.

That makes no sense. Think of it this way: "Myths against solo mining debunked". There is no triple negative.

[muyoso]

bcpokey: In other words, reduced variance is the only advantage of pooled mining.

No.  Its quite possible for someone to never mine a block solo mining.  Unless I am mistaken, its purely chance and there is nothing that guarantees that if you mine for a certain period of time you are GOING to get a block.  A person with 655Mhash/s like myself could mine a block within the first 10 minutes of solo mining (luck), or it could more likely take a year(or never).  Its much better to just use a pool, get 20 bitcoins in 2 weeks and be done with it.  

Within a few weeks the difficulty is going to be absolutely massive (3million at least by mid july) and I seriously don't think that is going to help anyone in my situation with mining solo.

[JoelKatz]

bcpokey: In other words, reduced variance is the only advantage of pooled mining.

No.  Its quite possible for someone to never mine a block solo mining.  Unless I am mistaken, its purely chance and there is nothing that guarantees that if you mine for a certain period of time you are GOING to get a block.  A person with 655Mhash/s like myself could mine a block within the first 10 minutes of solo mining (luck), or it could more likely take a year(or never).  Its much better to just use a pool, get 20 bitcoins in 2 weeks and be done with it.  

Within a few weeks the difficulty is going to be absolutely massive (3million at least by mid july) and I seriously don't think that is going to help anyone in my situation with mining solo.

Why do you say "no" and then agree with me?

[gmaxwell]

bcpokey: In other words, reduced variance is the only advantage of pooled mining.

No.  Its quite possible for someone to never mine a block solo mining.  Unless I am mistaken, its purely chance and there is nothing that guarantees that if you mine for a certain period of time you are GOING to get a block.  A person with 655Mhash/s like myself could mine a block within the first 10 minutes of solo mining (luck), or it could more likely take a year(or never).  Its much better to just use a pool, get 20 bitcoins in 2 weeks and be done with it.  
Within a few weeks the difficulty is going to be absolutely massive (3million at least by mid july) and I seriously don't think that is going to help anyone in my situation with mining solo.

It's also quite possible for a pool to never mine a block (or to mine far under the expectation), it's just less like just as it is less likely for the pool to do superbly well.  Worse, because of downtime, fees, cheating, etc. the pool is expected to average less overall. The the reduced variance is a real and useful result, but don't exaggerate the contrast.

[bcpokey]

This thread seems to be full of people using negatives when they mean positives.

@Joel if you say reduced variance is the only benefit that makes it sound like a trivial thing, try turning it around, what are the benefits of solo mining? No fee pools, pools that have been around for a long period of time (why would they scam YOU?), etc. Stale shares isn't even a real thing, as every share that does not complete a block is worthless in solo mining, so there is no comparison. Basically every "benefit" of solo mining is either addressed or purely hypothetical.

@Everyone telling me you can not mine blocks, hey thanks for telling me exactly what i said already. Now on the other hand, statistical probability makes it entirely improbable for you to NEVER mine a block with certain amounts of hashing power (i have had > 5GHash for some time now). if you want to quibble with probability you probably don't want to learn that that is what makes our universe run more or less. But you can run a ridiculous bad streak while difficulty mounts making it less and less likely for you to outgain what you would have gotten in pooled mining.

And that's just it, the real benefit of pools. There is a real tangible benefit to returns NOW rather than possible returns later. The two are not equal in an accelerating growth model.

[googlebot1]

And that's just it, the real benefit of pools. There is a real tangible benefit to returns NOW rather than possible returns later. The two are not equal in an accelerating growth model.

That's a bit rhetorical. Just as gmaxwell wrote: there is no sure return "NOW" in a pool vs. an uncertain return in a distant future for solo mining. There is just a trade-off between probability and payout. There indeed is a risk to never see a block before difficulty hits the space age. But there is also a chance to find multiple blocks during the current difficulty and get a large payout. A pool doesn't pay you better in the latter case, solo mining does.

Having instant benefits now instead of later might be a huge benefit during accelerated growth of a whole economy. But Bitcoin is a currency and does not earn you interest. Getting 50 BTC now or later is of no difference if it is BTC's value itself that is growing. If mining would return another currency, say Batcoin, and your goal was to participate in Bitcoin's growth, instant payout of Batcoins would be important, so that you can convert them to Bitcoins before the price (Batcoin/Bitcoin) increases. But the payout is denoted in Bitcoin already, you participate in the growth of its value, even if you aren't cashed out immediately, because the return is always a nominal 50 BTC.

It would be different if the price of Bitcoin was falling and you always sell your coins immediately to earn interest in another currency. Then an instant payout would be a benefit.

[PandaMiner]

...But the payout is denoted in Bitcoin already, you participate in the growth of its value, even if you aren't cashed out immediately, because the return is always a nominal 50 BTC.

Aren't the blocks some time next year going to be 25 BTC from then on?

[Vladimir]

Pools are the tax authorities, which levy tax on ignorance and lack of knowledge of elementary theory of probability. Kind of like weird reverse lotteries.

[bcpokey]

And that's just it, the real benefit of pools. There is a real tangible benefit to returns NOW rather than possible returns later. The two are not equal in an accelerating growth model.

That's a bit rhetorical. Just as gmaxwell wrote: there is no sure return "NOW" in a pool vs. an uncertain return in a distant future for solo mining. There is just a trade-off between probability and payout. There indeed is a risk to never see a block before difficulty hits the space age. But there is also a chance to find multiple blocks during the current difficulty and get a large payout. A pool doesn't pay you better in the latter case, solo mining does.

Having instant benefits now instead of later might be a huge benefit during accelerated growth of a whole economy. But Bitcoin is a currency and does not earn you interest. Getting 50 BTC now or later is of no difference if it is BTC's value itself that is growing. If mining would return another currency, say Batcoin, and your goal was to participate in Bitcoin's growth, instant payout of Batcoins would be important, so that you can convert them to Bitcoins before the price (Batcoin/Bitcoin) increases. But the payout is denoted in Bitcoin already, you participate in the growth of its value, even if you aren't cashed out immediately, because the return is always a nominal 50 BTC.

It would be different if the price of Bitcoin was falling and you always sell your coins immediately to earn interest in another currency. Then an instant payout would be a benefit.

How is it rhetorical if no one seems to accept it. There is a sure return now because pools of a certain size do not go full difficulties without block solving. There *is* no sure return of any kind with solo mining. if difficulty were constant perhaps, but as it jumps up up up your time to hash keeps increasing. Getting 50 BTC now or later does not make a difference except that the longer it takes you to actually GET that 50BTC the less likely you are to command similar returns to pool payout. This is why its not rhetorical?

The idea that the lottery will reward you more handsomely in the future is just wishful thinking. Yeah you could find like 20 blocks in a row at 20mil difficulty with 500MHash/sec, but you really aren't going to.

This isn't rocket science. You have a probability curve that models your expected time to solve a block that keeps getting shifted away from where you are as time passes and difficulty increases. This is what bugs me about people saying "oh anyone CAN find a block at any time!" Yes, and anyone can win the lottery, it is still a stupid investment if the odds are stacked against you. Spread your timeline out, and you will see the advantages that grabbing as much as possible for yourself as soon as possible will yield. it's simple.

Again, i will use myself as an example to make this less abstract for you:

i have been mining for 2 months more or less. in that time i have solved 0 blocks. i have earned roughly 350BTC in that time. this means going forward i would need to solve > 7 blocks just to catch up to myself. the next difficulty (occuring in about 2 hours) will be about 1.4mil. The average time to find a block at this difficulty is 14 days at my current hash rate (> 1 difficulty change). Coin toss @ 9.75 days. The chances of me dropping 7 blocks in a row in this time period are astronomical against. Obviously i'm an unlucky case, but i'm also a real case. Again increasing difficulty isn't just some boogeyman, it's a real consideration and why you need to consider the coinage you can gather now as superior to ones you imagine getting in the future.

[PandaMiner]

Difficulty has increased.
http://www.btcserv.net/bitcoin-calculator/

[googlebot1]

LOL!!!!!!!!!!!!!!!!

After my last post my lonely HD5870 (940/300) has just generated a block! Can't see, which one it is, yet. It includes 0.0715 BTC fees and has only 23 confirmations so far.



Compare that to a measily 0.4 BTC a day, is that really worth your time? All or nothing!  

[nathanrees19]

Stale shares isn't even a real thing, as every share that does not complete a block is worthless in solo mining, so there is no comparison.

No. Stale work has no chance of producing a block. Current work has a non-zero chance of producing a big reward.

[kokojie]

I have 3M+ shares mined on btcguild, yet Blocks Found: 0

So if it were not for pools, my 4 GH/s would have produced nothing for me so far if I mined solo.

[Adam]

Solo mining is undoubtedly better mathematically, but you could mine for years without ever finding a block and I just can't see that sitting well with most people.  This is a bit of a simplification because I am not taking difficulty increases into account and because I am assuming every day as an individual trial instead of every share, but the average time to solve a block with a hash rate of 1ghash is 68.5 days at the current difficulty.  If you convert that to a percentage you have a roughly 1.46% chance of generating a block in any given day, or a 98.54% of not solving a block today.  

Chances of solving at least one block, at constant difficulty and 1ghash/sec:

1 day : 1.460%
10 days : 13.676%
50 days : 52.064%
100 days : 77.022%
200 days : 94.720% (1 chance in 19 of having a run this bad with no blocks)
365 days : 99.534% (1 in 214)
500 days : 99.936% (1 in 1,561)
1000 days : 99.99996% (1 in 2,436,681)

[Meni Rosenfeld]

The utility of money is concave, logarithmic is a good model. It's expected utility you need to maximize, not expected money.

This means that variance is not just "psychological", there is a real economic cost to variance. These days you might get away with saying that if you're a Vulcan you're better off mining solo, but in the future when mining will be one in a million chance to get $100M even that much won't be true.

And we're not Vulcans. Trying to mine solo caused me to lose sleep. Plus, some pools have convenience and monitoring features, and in the future end users won't be able to run a full network node.

Solo mining is undoubtedly better mathematically

Only if you think mathematics is about extremely oversimplified models completely detached from reality.

[googlebot1]

Solo mining is undoubtedly better mathematically, but you could mine for years without ever finding a block and I just can't see that sitting well with most people.

The alternative is heating and noising up your house for a wimpy 0.72 BTC a day on the same assumptions. That's really not worth the hassle.

On the other hand, I have switched 1 week ago from pooled to solo mining, after earning 3 BTC in total, the hard way. Today I have 53 BTC in my account, that's really worth the effort. The big profits make you smile. IMHO, a trickle profit of <1 BTC really isn't much more satisfying than 0 payback and the chance for big cash. That might be interesting for kids converting their parents' electricity to cash, but not for me.

[toast]

Would you rather:

1) Pay me $1,000. I generate a random number from 1 to 1,000. If that number is 523, I will pay you $1,000,000,000. You can keep playing as many times as you want, but after you play a few times the reward drops from $1,000,000,000 to $500,000,000

2) Take $1,000 I give you for free right now.

See the problem? If you are unlucky before it drops to 25BTC/block, you CAN NOT expect things to average back out in the long run. You COULD get lucky in the short term too, but then your EV is the same as (not better than) solo mining. It's a question of whether you think the implicit fees you mentioned are worth the fact that you might make losses that will not be smoothed out by the law of large numbers after block payout changes.

EDIT: Wait, or would the probability rise to 1 in 500 at the same time (in situation 1)? In that case, nevermind, but:

The logical extension of your mentality about this is that you shouldn't buy health insurance because you would pay more per month on health insurance than you would on average spend every month on paying for your own healthcare when you get badly injured.