Newb question about pool mining vs. Solo

[CubedRoot]

I have a setup that is running about 1 Ghash/sec right now.  I have a 5870 on its way to push the right to somewhere around 1.4 Ghash/sec.

I am currently mining over at BTCGuild and donating at 2.5% just to take advantage of the features the pool offers.

Is it worthwhile to mine solo?  I have read a few of the threads and wiki's but am still confused.  I calculate that 1.4 Ghash/sec at current difficulty should get around 2 to 2.5 BTC a day average (once you divide out the block you find across the days in the month).  Are those numbers off?

Is anyone solo mining with a comparable hash rate (around 1.4 Ghash/sec)?  If so, what are you averaging at current difficulty?

Thanks for the input.

[kiwiasian]

It's really just your choice...

Pooled mining = you're pretty much guaranteed to get paid in small amounts...in the long run you lose a lot of potential BTC
Solo mining = you are taking the risk yourself, you might find a block and you might not...provided you are successful, in the long run you get a lot more BTC.

It's up to you.

Most of us here pool.

[bcpokey]

Solo mining is a lot like pooled mining except that you will have a much larger variance in block generation. On average you can expect to find 1 block in 15.5 days. Over a long enough timeline you will average out to make a little more than you would donating to a pool, but not by that much, and it's a bit nerve wracking waiting 15-30 days to see any result. It's of course a personal choice.

EDIT: One thing I'm unsure of, maybe someone else can answer. If you are working on a block of Difficulty A, and you do not solve it before the network implements a raised difficulty, do you still work on the block at its original difficulty or the new difficulty? That could be a bummer for a solo miner.

[CubedRoot]

Good points.
If the return is only marginal compared to pool mining, I believe I will stick with pooling, if nothing else to have the nice web interface that lets me check stats and get email notifications.

[Graet]

same here....no updates in years...

has anyone got any advice/threads concerning solo mining on a 70gh avalon asic?

<Graet> ;;gentime 70000
<gribble> The average time to generate a block at 70000.0 Mhps, given difficulty of 21335329.114, is 2 weeks, 1 day, 3 hours, 38 minutes, and 4 seconds
could easily be 5 days or 5 weeks too - basically how much variance can you handle?

on a pool before fees you would expect to average
<Graet> ;;genrate 70000
<gribble> The expected generation output, at 70000.0 Mhps, given difficulty of 21335329.114, is 1.65000764516 BTC per day and 0.0687503185485 BTC per hour.

of course difficulty will go up 10%+ in ~8days making it harder to solo and pool mine

the rule of thumb I have heard a lot is : if you cant expect to generate a block within a difficulty period (just under 2 weeks currently) you should pool mine

good luck with it

[redtwitz]

i have 2 x 70gh machines and figure that if im ever going to solomine it should be now as the difficulty is updating and increasing soon.

with regards to actually setting up for solo mining, i now have about 6 threads open looking for clues.

Here's a clue: Don't.

Your 140 Ghps will produce on average 140,000,000,000 ÷ 2³² = 32.596 diff 1 shares per second, which gives 32.596 × 60 × 60 × 24 × 30 = 84,488,832 diff 1 shares per month.

Thus, you should create on average 84,488,832 ÷ 21,335,329 = 3.96 blocks per month, which gives a total of 99 BTC.

In a mining pool, that will be pretty close to your actual reward. Whether you sacrifice a few percent and go for PPS or mine in a pool with low fees, that pays tx fees and maybe even does merged mining, your rewards should be within a few percent of those 99 BTC.

In the meantime, if you mine solo, there's a e^-3.96 = 0.0191 = 1.91 % probability that you won't earn a single BTC in the first month.

The decision is obviously up to you. Just making sure you know what you're getting yourself into...

[Trillium]

Thx graet,

i have 2 x 70gh machines and figure that if im ever going to solomine it should be now as the difficulty is updating and increasing soon.

The fact that you have seemingly spent (how many thousands?) of $ or BTC on Avalon hardware but you can't answer your own question is highly concerning to me.

Why would you take such a huge investment risk diving into a area you seemingly don't know much about? 

The answer to your question can be googled in 30 seconds or less, why have you not done this before investing in this hardware? Likewise the tutorials and guides to setting up solo mining on either SHA-256 coins or scrypt coins are just a forum search / google search away... it would take you days to read them all.

Even without looking at it quantitatively or statistically, the short answer to your question is "it doesn't really matter" because your hashrate is so high you will still see returns in either case in a fairly small short frame (weeks-month).

I have a setup that is running about 1 Ghash/sec right now.  I have a 5870 on its way to push the right to somewhere around 1.4 Ghash/sec.

I am currently mining over at BTCGuild and donating at 2.5% just to take advantage of the features the pool offers.

Is it worthwhile to mine solo?  I have read a few of the threads and wiki's but am still confused.  I calculate that 1.4 Ghash/sec at current difficulty should get around 2 to 2.5 BTC a day average (once you divide out the block you find across the days in the month).  Are those numbers off?

Is anyone solo mining with a comparable hash rate (around 1.4 Ghash/sec)?  If so, what are you averaging at current difficulty?

Thanks for the input.

No OP, you definitely DON'T want to solo mine if you are just running a couple of GPU's. You could mine for years and not find a block, and those power bills won't make it worth it.

If you are really interested in the idea of solo mining I would suggest you look at the altcoins that have much lower difficulties where you might be able to find blocks every few hours/days.

[redtwitz]

No OP, you definitely DON'T want to solo mine if you are just running a couple of GPU's. You could mine for years and not find a block, and those power bills won't make it worth it.

The OP is from June 1, 2011. Back then, the difficulty was at 434,877. The 1.4 Ghps GPU rig back then was roughly comparable to a 70 Ghps ASIC right now.

[Trillium]

No OP, you definitely DON'T want to solo mine if you are just running a couple of GPU's. You could mine for years and not find a block, and those power bills won't make it worth it.

The OP is from June 1, 2011. Back then, the difficulty was at 434,877. The 1.4 Ghps GPU rig back then was roughly comparable to a 70 Ghps ASIC right now.

Ughh this guy has necro'd so many threads with this question I've lost count.

[edgar]

so redtwitz says Dont Do It, as there is 1.91% chance of not earning anything...and presumably a 98% chance that i'd earn around 100 (in the same time frame)

 and you say it doesnt matter.

assuming that you have both done the relevant research why would your answers not be the same?

your 'very heightened concern' is appreciated, thank you.

to answer your seeming rhetorical q's,

i spent 110btc, which cost me almost nothing(they were worth thousands at the time but that is irrelevant to me)

the reason i 'invested' without doing your level of research is the batches were only available for a very short period of time, i missed out on batch 1 and regretted it, and decided with a bit of encouragement that batch 2 would be worth investing in (IF you had the funds and could risk them without worry/selling my house)

had i known id be waiting as long as i have i may well not have invested, yet here we are (and i wish id bought some B3s as well)

so, solo mining is worth it, maybe, if you have enough G/h's.....

i am confident 140-160 will be enough for me to get 'lucky' at least for the next 3-4 difficulty adjustments.

plus i only need to raise my initial coins back anyway and that shouldn't take too long.

and BTW i have over 12 pages of how-tos & tutorials open, some of them with 'tons' of pages, so asking for a pointer or 2 shouldnt be of much concern to anyone other than those with too much time on their hands.

I highly appreciate any and all of the decent responses i have had to any & all of my posts. Thanks.

[Trillium]

It all depends on how long you want to wait between payouts. Statistically you might not find a block for some days/weeks if you solo mine and financially this might not be acceptable to you. If you are more impatient and would like daily or every-few-daily payouts then a pool is a better option as you will be able to very quickly meet any pools auto-payout thresholds and these payments will occur reliably.

[redtwitz]

so redtwitz says Dont Do It, as there is 1.91% chance of not earning anything...and presumably a 98% chance that i'd earn around 100 (in the same time frame)

Well, no.

First of all, you cannot earn around 100 BTC with solo mining. You can only find an entire number of blocks, so you'll earn 0 BTC or 25 BTC or 50 BTC or 75 BTC or 100 BTC, etc.

The number of found blocks in a given time frame follows a Poisson distribution, where the parameter is the expected number of found blocks in that time frame.

With 140 Ghps at 21,335,329 difficulty, the probability of finding k blocks in 30 days is 3.96^k × e^-3.96 ÷ k!. For reasonably small values of k, that gives:

Code:

Earnings | Probability
---------+------------
  0 BTC  |   1.91 %
 25 BTC  |   7.55 %
 50 BTC  |  14.95 %
 75 BTC  |  19.73 %
100 BTC  |  19.53 %
125 BTC  |  15.47 %
150 BTC  |  10.21 %
175 BTC  |   5.78 %
200 BTC  |   2.86 %
225 BTC  |   1.26 %
250 BTC  |   0.50 %
275 BTC  |   0.18 %

Thus, with probability 19.53 %, you'll earn 100 BTC (within a few percent of earnings when mining in a pool); with probability 44.14 %, you'll earn 75 BTC or less; with probability 36.33 %, you'll earn 125 BTC or more.

Whether that's acceptable for you or not depends entirely on yourself. Personally, a probability of 44.14 % of losing at least 24 BTC in expected earnings by solo mining for a month would be too high for me.

assuming that you have both done the relevant research why would your answers not be the same?

Since when do two people on the Internet agree over anything?

[Trongersoll]

 OP You should definitely solo mine. It is obvious that paying a pool runner 2-3% is going to keep you up at night wondering how you can get that back. You will obviously make more solo mining over the long run. Why should some pool operator get rich off your effort. You obviosly want every fraction of a BTC that you can get and won't be comfortable until until you get it. Greed is good. 

[Trillium]

Thus, with probability 19.53 %, you'll earn 100 BTC (within a few percent of earnings when mining in a pool); with probability 44.14 %, you'll earn 75 BTC or less; with probability 36.33 %, you'll earn 125 BTC or more.

I'd take those odds. But now plug OP's 1.4GHash into your spreadsheet and see how poor it looks. The problem with this thread is that we've got OP with next to no hashing power (on SHA-256 at least) at 1.4 GH/s and then we've got a guy with two Avalon rigs which is the complete opposite case.

[polarhei]

Unless you have 10Thash/s your own, otherwise, you better to work together