Topic: [1050 TH] BitMinter.com [1% PPLNS,Pays TxFees +MergedMining,Stratum,GBT,vardiff] - page 11. (Read 837101 times)

Actually it is multiplicative not additive.  Thus it is one in 10,000.  Which if you extrapilate out at the same difficulty means it should happen about once in 25 months.  I.E. hardly a once in the life of the universe event.

I didn't mean anything bad by not taking the 29 NMC for advertising. I am going to advertise and the pool does have coins for this. But if you want to donate something extra for advertisements, send BTC to:

14uS36iDCjXPB5c1XHSCWTWpH7RysqwkoZ

NMC can be sold for BTC at btc-e.com or other exchanges.

If you are good with graphical work you can make a banner for the pool (contact me). If you have a blog or other website you could mention the pool and link to it. Word-of-mouth is one of the most important things. If someone is looking for a pool or looking to start mining, you can suggest this pool. And obviously, mine at Bitminter yourself. If you get any other ideas, send me a message or post it here.

That's a few things for those who want to help the pool grow. All help is greatly appreciated.

I am also working on a few ideas to improve the pool. More on that later.


To address a few things that has come up in the last couple days:

No the pool is not broken. Bad luck happens, just like good luck. We are now below 1% of the global hashpower though, so unlucky rounds take longer than before.

You don't earn more the bigger a pool is, that's not how it works. The payouts may be further apart now, but they are also bigger. Variance is higher, but average income is the same. Your income is not dropping towards zero because the pool is smaller, making you unable to cash out. That is incorrect.

Yes, sometimes there is a shift with zero pay. But shifts are often paid multiple times as well. You can see this at https://bitminter.com/stats/rewards which shows the payouts during the last 500 shifts. Sometimes the green line on the top graph hits the bottom (zero pay for a shift), but also notice how it goes up and down on certain "levels" depending on how many times each shift is paid (which again depends on how many blocks we found).

From the bottom graph you can see we had plenty of good luck lately. But unfortunately the bad luck during the last week has cancelled out most of that at this point. We have to expect this to happen, good and bad luck cancelling each other out. But I can understand the last week has been rough.

The yellow line (expected average) dropping in the top graph is because of the difficulty increases.

And this'll be why I bet-spread.

If you look at Doc's graph you'll see the peaks & troughs. It averages over a month, not a week.

But...

Every pool has good and bad weeks. The thing is big pools don't really wobble if they get a bad week. Where-as little pools do.


Both my larger miners are set to 'quota balance'. Both share hash between one large & one small pool. All I have to do is adjust the quotas depending on how scary said pool is getting. If I get cash issues I simply move hash to the safer pools ( usually the bigger ones ).

If a small pool is having a good run I'll transfer some hash there, as the payouts will be better. I can't predict the future accurately, sure; but at least I can cover The Bills.

*shrugs* So far so good.

The mistake is to have all your eggs in one basket.

Minter is fun, but even I've dialed down the hash here now. Bills bills bills. Hopefully we'll be back on an even keel in a couple of weeks, and I'll be able to dial things back up again.

No, you wrote:


You are clearly indicating that if p = 0.99, then the quantile is infinity. This is incorrect. The exponenential quantile is infinite only if p = 1.0 .

The CDF can go to 99.9999999... but never reaches 100% it will get very very close but never there so is infinitely long.

I think the point is that if you find a pool (or pools) that are fair and you trust, in the long run, you get the same amount.   So, you need to find somewhere that you think is fair, and then only compare over the long run.
My kids' account is hashing over the next 10 years (either it will be a college fund or a father's foolish hobby), so I guess I should just solo mine for them, but that has costs that add up if you do not hit a block for a very long time so not caring how much it goes up each week is easy to do.    The business side has 20k in electricity to pay each month.   It is harder to not care but I believe in BTCGuild and in Bitminter and can live with a week with only two blocks.   It will not be like that for the long run.  

OK so what I think your saying is take a guaranteed smaller payout on a larger site or have a chance at getting better "luck" and benefiting from it on a smaller site.

I'm just thinking that the more block you can hash before the difficulty goes up the better. 3 days between blocks is horrible luck.  I'm really hoping something else isn't gone wrong.

Fahlcor

Because there is a symmetrical chance of getting twice the expected payout as there is of getting 0 payout.  If you mine for 24 hours and get twice what you should have, and then mine another 24 hours and get nothing, you end up right where you should be.

The chance of low payouts during a given timeframe is greater the smaller your pool is.  At the exact same time, when a smaller pool gets lucky it pays out *significantly* better.  The expected earnings is the same regardless, and so is the long term average.  It's just a matter of how far you can deviate from expectations during that time frame (and the smaller the time frame, the greater that deviation can be).

---

Adding on that previous thought, if you want to attempt to refute it:  Pooled mining as a whole works on the exact same principal.  Every hash is equal, even though a user with 1 GH/s is statistically unlikely to ever find a block for the pool, and will siphon off a small piece of each block they participate in, they are still worth having.

You will have users who have never mined a block for the pool (similar to your picking and choosing periods of time with no payment).  You will have users who have mined significantly more blocks than expected for their hash rate.  Average it all together over enough time and you should end up with the same number of blocks as you would have mined if it was just a single user mining with the pool's hashrate.

Looks like your main problem is that you don't properly understand how PPLNS works.

I'm not really sure what you mean, but by definition the lower tail 0.99 CDF occurs when an event is expected to occur one time in 100.

If that's for one block, it will require total work of 4.60517 times the network difficulty. If it's for two blocks, it will require total work of 6.638352 times the network difficulty.

In the first case we would expect this to happen once in one hundred blocks; in the second case once in two hundred blocks.

That's not even close to infinite.

Just thought I would point something out.  This may change some minds or not.

Shifts
===
Work is measured in proofs of work. Each proof of work is given a score equal to one divided by the difficulty at the time it was submitted. Proofs of work are grouped in shifts. A shift is completed when its total score passes 0.4. Whenever new coins are minted, they are divided between contributing members proportionally to the total score each member has within the last 10 completed shifts. If you are logged in you can see your score in each shift.
===

So based on the last blocks found

315364   2014-08-13 04:33   3d 5h 51m   96,596,781,016   19,729,645,940   99.3%   1,480.3   BTC   tyson76   confirmed
314792   2014-08-09 22:40   3d 2h 48m   98,678,474,846   19,729,645,940   99.3%   1,573.6   BTC   Koi   confirmed
314272   2014-08-06 19:50   14h 44m   19,290,357,334   18,736,441,558   64.3%   1,560.7   BTC   Megatoast   confirmed

The last shifts that I can see on the site

2014-08-13 22:19   7h 16m   19,729,645,940   1,295.1   -   -   -
2014-08-13 15:02   6h 55m   19,729,645,940   1,361.3   Block was found on this shift
2014-08-13 08:06   6h 57m   19,729,645,940   1,352.4   2 previous 10 shifts share it
2014-08-13 01:07   6h 58m   19,729,645,940   1,350.1   3
2014-08-12 18:08   6h 40m   19,729,645,940   1,411.0   4   -   -
2014-08-12 11:26   6h 39m   19,729,645,940   1,413.1   5   -   -
2014-08-12 04:46   6h 20m   19,729,645,940   1,486.7   6   -   -
2014-08-11 22:25   6h 17m   19,729,645,940   1,498.3   7   -   -
2014-08-11 16:07   6h 08m   19,729,645,940   1,532.1   8   -   -
2014-08-11 09:57   6h 10m   19,729,645,940   1,523.6   9   -   -
2014-08-11 03:46   6h 14m   19,729,645,940   1,512.6   10   -   -
2014-08-10 21:31   6h 12m   19,729,645,940   1,519.4   so nobody got anything for this shift 6 hours   -   -
2014-08-10 15:18   6h 03m   19,729,645,940   1,556.0   or this 6 hours
2014-08-10 09:14   6h 02m   19,729,645,940   1,561.9   or this 6 hours
2014-08-10 03:11   6h 06m   19,729,645,940   1,540.4   Block was found in this shift
2014-08-09 21:04   6h 05m   19,729,645,940   1,547.5   2
2014-08-09 14:58   5h 58m   19,729,645,940   1,579.9   3
2014-08-09 08:59   5h 56m   19,729,645,940   1,584.6   4
2014-08-09 03:02   6h 02m   19,729,645,940   1,562.4   5
2014-08-08 20:59   6h 01m   19,729,645,940   1,565.0   6

So if the pool is taking too long for it to find a block like it is currently then users are not possibly able to get full value for their hashing as the pool simply does not pay further than 10 shifts back.  If the pool passes 10 shifts without a block then you are not making what you could at a larger hashing pool.

I thought I would be rewarded for my all my previous efforts but my efforts were for not.  I'm will be shifting my stuff to faster pools until bitminter can hash a block in less than 10 shifts.

I'm not trying to move people from here just pointing out something I learned.  I'm also assuming based on the time it took to find the previous block and not being able to see the previous shifts that there would have been lost hash time before this.

I'll give it another 12 hours and see what happens.  Why would we take the chance of hashing any time without a possible payout?

Fahlcor

The mathematical probability of hitting 2 blocks in the 99th cumitive distribution factor is astronomical, roughly the same as hitting 2 at <1, not quite the same as the 99th is infinite since 100% can never be achieved. And away we go again already at 72. It is starting to look like something is amiss (wrong setting somewhere, lost blocks??) I do not know. If a third is reached then there will be a mass exodus from bitminter.

Chief at the controls, check the switches! Something has gone awry

You're a little off on the math (but round numbers are okay normally).  However, you did 3600 x 12, not 3600 x 14.  The difficulty should change every 14 days at neutral (which would be 144 blocks -> 3600 coins).  While difficulty adjustments lately are much faster than that, it doesn't change the fact that the # of coins mined during a difficulty period is 3600 x 14 = 50,400.  If difficulty increased faster, the x14 would go down, but the 3600 would go up, but the product will always be 50,400 coins mined during a given difficulty period (+0.5% or so for tx fees).

Lets say that you have 4,600 GH/s of miners (4.6 TH/s).
So at the start of this difficulty change, you would have seen that difficulty moved to 19,729,645,941.   To get the network hashrate implied by that divide by 140,000,000
You will get the network size as 140.926 PH/s    That is 140,926 TH/s.
So what could you expect on average before the next difficulty change?
4.6 / 140,926 (solves the share of the network you are) X 3600 (bitcoin rewards in a day) X 14 (days)  = 1.65 bitcion is what you should have expected to receive (less bitminter's share) in this period.
I would not divide it by day (but I am sure everyone will) because it is really the period and that is not a constant 14 days....

EDIT:   Corrected for mistake that michael pointed out.

Could you please post the formula to calculate what your expected earnings should be at any given difficulty? I've seen in a bunch of times but have never taken the time to write it down.

Glad I'm not the only one that sees it that way.  I was actually debating trying solo.  And then see if I can hit the 25 btc on my own ;-)

Donation should not be needed as there are fees to cover that.  But then again if the pool is not churning out the blocks then there is no donations either I guess.

I know how you feel.  I just had to move 7 thps to another pool. I  like Bitminter  but money is the key. I have offered to donate money to help the pool grow but Doc was not interested.



Good luck to all, but after 7 days and 2 blocks and the owner of pool seems unconcerned about growth I have to move on. I left 1.8 thps to see what happens in the next day or two.

ok I'll give it until the next difficulty change.  No more.

If I don't average .128btc/day like i have at ghash I'm gone.  I used Bitminter as my first hasher long time ago and I do believe that minter is the better pool.  I'm back in the top 50.

I just hope we can get more hashers to get the blocks coming faster.  I have not been doing BT that long with these new ants but I will take the wisdom and experience of the group at fae value and assume that anyone offering experience has compared the rest also.

I could be wrong about ghash not being free but it was nice to see the btc ticking into my wallet.

C'mon bitminters get up the hash!!!

Fahlcor

I posted this before but I will again.
When difficulty changes, figure out what your expected number of bitcoins should be for that period for the amount you are hashing.   Do this for each pool you have miners pointed at.  Then at the end of the difficulty period see how close you are to your expectation.   Do not try and do this on a daily or weekly basis.   It is far simpler to use an entire period.
I can tell you what you will find (and I have enough money at risk to REALLY care about this answer).   BitMinter and BTCGuild will be the closest to expectation that you will find.   There is nothing free in this world, if you believe someone is giving you something for free you likely are still waiting for Josh at BFL to tell you how to invest your money.

EDIT:  and I should add that my kids' miners are pointed at only ONE pool.  Bitminter    caromei, they still have not solved a block, but it will be at bitminter when they do.  

You can't extrapolate from 25 hours to several weeks. Most miners that have been around for a while check their income weekly to avoid variance-panic, which is what you've been hit with.

Stick with a pool for a week or so, then compare what you've earned in a week to what you should have earned in a week. You'll be surprised how close to expected it will be. You certainly wont have to wait a couple of weeks for your pay.