I had a chat with dutu about the question that GB2 members may owe GB1 members... here's the entire chat.
Basically the answer is the timing that the pools paid out fell into the territory of time where GB2 machines had joined, but the income was accrued when only GB1 machines were active. Read below. Please comment.
Wednesday, February 19, 2014
[3:29:16 AM] agibby5: anyway, i wanted to ask what the whole GB2 owes GB1 calculation is about?
[3:29:18 AM] dutu: i think i missed that
[3:29:20 AM] agibby5: i don't follow that at all
[3:29:37 AM] agibby5: and i don't see how you can assume that expected earnings are == actual earnings
[3:29:38 AM] dutu: i think it's easy to explain
[3:29:55 AM] agibby5: i mean, do you have proof that soniq was mining to other addresses?
[3:30:02 AM] agibby5: that's the only way i can see that happening
[3:30:06 AM] dutu: no i think he wasn't
[3:30:17 AM] dutu: what happened is:
[3:30:22 AM] agibby5: otherwise, the payouts are the payouts and you can see them in teh addresses via recieved btc
[3:30:29 AM] dutu: GB1 started mining before GB2
[3:30:55 AM] dutu: GB! should have been paid for all this mining income which was when only GB1 was mining
[3:30:57 AM] dutu: right?
[3:31:56 AM] dutu: the fact is GB1 has not been paid for all this income, as pard of this income has arrived at soniq wallet later
[3:32:05 AM] dutu: and then it was split between GB1 and GB2
[3:32:21 AM] dutu: when it should only been sent to GB1 users
[3:32:26 AM] dutu: does it make sense?
[3:33:09 AM] agibby5: yea. i follow that
[3:33:22 AM] agibby5: but when i look at your spreadsheet, it indicates alot of "expected" income
[3:33:29 AM] agibby5: but expected results are not equal to actual ones
[3:33:49 AM] dutu: yes, that is the estimated income
[3:34:14 AM] dutu: estimated theoretical income based on the Jupiter hash power
[3:34:35 AM] dutu: and this is in line whit what we have received in soniq's wallet
[3:34:52 AM] agibby5: but GB2 members can't makeup for any variance to actual vs expected income recieved by GB1 members
[3:34:53 AM] dutu: this is why I am saying i believe he hes not mined anywhere else
[3:36:00 AM] dutu: can you repeat the question please?
[3:36:23 AM] agibby5: you cannot take mining income from GB2+ members to pay for any variance related to actual results for any GB1 members
[3:36:26 AM] agibby5: example:
[3:36:37 AM] agibby5: say GB1 was expected to earn 100 btc, but they earned 85 instead
[3:36:48 AM] agibby5: and now GB2 gets on board, and they earn 200 now combined
[3:36:49 AM] dutu: ok, one sec
[3:36:56 AM] agibby5: you cannot take the 15 btc out of the new 200
[3:37:08 AM] agibby5: because GB1 didn't really earn that in teh first place
[3:37:22 AM] agibby5: btw, i'm in GB1 so i would love a bump
[3:37:29 AM] dutu: let us go one step back
[3:37:30 AM] agibby5: please convice me
[3:38:15 AM] dutu: column G "expected amount" is the theoretical mining income based on Jupiter has rate, ok?
[3:38:27 AM] dutu: now look at line 38
[3:38:55 AM] dutu: it shows that the theoretical mining income is equal with the actual mining income
[3:39:12 AM] dutu: good, then we have concluded that
[3:39:31 AM] dutu: 1) theoretical mining income is a good estimate
[3:39:47 AM] dutu: 2) soniq has not mined somewhere else
[3:39:53 AM] dutu: next:
[3:41:19 AM] dutu: line 26 shows that received amount is much hi=hers the expected amount to be mined during that period
[3:41:22 AM] dutu: why?
[3:41:51 AM] dutu: because the coins received were coins mined during thet period + coins mined during previous period
[3:42:34 AM] dutu: late transfer of coins is also shown by negative values in colum g on lines 24 and 25
[3:43:09 AM] dutu: taht is coind mined during periods in line 24 and 25 were only received during the period on line 26
[3:43:11 AM] dutu: got it?
[3:44:38 AM] dutu: btw, we can chat on the group chat if you want i don't mind
[3:45:25 AM] agibby5: i do, but it seems like you can't keep balancing between GB1 and GB2 members. for example, how do you fix the lines 29 and 34?
[3:45:55 AM] agibby5: i guess my question is, at what point do you assume that GB1 and GB2 shares payout equally during that dividend payout?
[3:47:31 AM] dutu: i'm not sure if I understand the question
[3:48:09 AM] agibby5: at what point should soniq assume that payout of shares for GB1 and GB2 are the same?
[3:48:16 AM] agibby5: is it line 25?
[3:48:24 AM] agibby5: if so, then you can't go to line 26, and send a "rebate" to GB1
[3:48:47 AM] agibby5: if it's line 26, then you can argue that GB2 also earned a bit less because of line 25s lower income
[3:48:49 AM] dutu: no you don't do it at every line and I agree with you you cant
[3:49:24 AM] dutu: becuase the mined bitcoin is not coming to the wallet in real time
[3:49:57 AM] agibby5: i just take issue what what you're saying a bit here... becuase it seems like you want to take income that's been recieved at the wallet when GB2 was in action and put it into GB1's hands.
[3:50:05 AM] agibby5: and that's hard to explain to people
[3:50:13 AM] agibby5: i think i understand now... but I don't know how it'd be justified
[3:50:26 AM] agibby5: i believe it's due to mining variance, right?
[3:50:31 AM] agibby5: everyone shoulders that risk
[3:50:46 AM] dutu: well, that's a small amount per user anyway
[3:50:54 AM] agibby5: what's the real difference?
[3:51:12 AM] agibby5: do you see my point?
[3:51:12 AM] dutu: but the fact is mining income that should been received for line 24
[3:51:20 AM] agibby5: but it wasn't though
[3:51:28 AM] agibby5: variance showed us that we couldn't get that many coins
[3:51:40 AM] agibby5: but the next time period, the variance worked in our groups favor
[3:51:44 AM] dutu: has only been received during the period on line 26
[3:52:56 AM] dutu: the mistake soniq did was thet he didn not collected at wallet all the bitcoin mined before GB2 started to mine
[3:53:10 AM] agibby5: that's not his mistake
[3:53:15 AM] agibby5: that's the pool paying out at that time
[3:53:19 AM] agibby5: because it found the blocks when it did
[3:53:19 AM] dutu: and then when he received this income later he splited it between GB1 and GB2
[3:53:23 AM] agibby5: that's the definition of variance
[3:54:13 AM] dutu: hm... not sure about that
[3:54:42 AM] dutu: pools would pay per mining share and will not have such a big variance i believe
[3:55:37 AM] agibby5: what explains that income coming in later then?
[3:55:45 AM] agibby5: magic?
[3:57:03 AM] agibby5: despite my humor, that's a serious question
[3:57:11 AM] dutu: when did soniq stared to mine at Bitminter?
[3:57:38 AM] agibby5: offhand, i don't know. he would say that's public information and it could be found... but i don't know where.
[3:57:55 AM] dutu: on the Oct 16 we have the first payouts
[3:58:07 AM] agibby5: how about this?
https://bitcointalk.org/index.php?topic=226319.msg3364200;topicseen#msg3364200[3:58:11 AM] dutu: of a total of 39.78 BTC
[3:58:50 AM] dutu: which means teh miners were mining
[3:58:58 AM] agibby5: that was oct 18th
[3:59:04 AM] dutu: but the payout address was only setup on the 16
[3:59:47 AM] agibby5: yes, first payments look like the 16ht
[3:59:48 AM] dutu: after that you have regular payments of ~5 BTC
[4:00:14 AM] agibby5: yeah, 16th it looks
[4:00:58 AM] dutu: which explains why you ahve a huge amout of coins received for the period on line 26
[4:01:16 AM] dutu: because some coins where mined before
[4:01:33 AM] agibby5: where did you get your figures?
[4:01:44 AM] dutu: what figures
[4:01:52 AM] dutu: teh transactions?
[4:01:54 AM] agibby5: in the spreadsheet
[4:01:58 AM] dutu: from the blockchain
[4:02:16 AM] agibby5: it looks wrong
[4:02:27 AM] agibby5: 10/16+ should be GB2+3 starting
[4:02:30 AM] agibby5: not the 15th
[4:02:37 AM] agibby5: right?
[4:02:52 AM] agibby5: and 10/15 is still part of the original 15 miners
[4:04:28 AM] dutu: as per soniq's information:
[4:04:38 AM] dutu: "Date reached advertised hashing rates"
[4:04:44 AM] dutu: GB2, GB3
[4:04:48 AM] dutu: 10/15/2013
[4:04:55 AM] dutu: this is what I have in the file
[4:05:19 AM] agibby5: i'm still unclear how this happened
[4:05:25 AM] agibby5: if it's not variance, what is it
[4:05:29 AM] dutu: but if something is wrong i can correct
[4:06:01 AM] agibby5: for example, i buy a car that says it can get 25miles per gallon. sometimes it gets 32, sometimes it gets 17
[4:06:27 AM] agibby5: can i call the manufacturer of the car or the builder of the road to complain when I get 17?
[4:06:36 AM] dutu:
[4:07:33 AM] dutu: are you saying one day the jupiter is hasing 100oGHS and another day at 300?
[4:07:42 AM] dutu: i don't get the analogy
[4:07:43 AM] agibby5: no, i'm trying to come up with an analogy
[4:07:57 AM] agibby5: but i wonder this...
[4:08:20 AM] agibby5: i know larry was switching pools a few times and trying to let everything settle during those first few days. people weren't happy with teh output
[4:08:24 AM] agibby5: maybe that's what caused it
[4:08:32 AM] dutu: cause of what?
[4:08:56 AM] agibby5: so yes, on average, i'm guessing perhaps that this would be attributed to teh lower income... hence the lower hash output because the miners weren't hashing 100% of the time due to switching pools?
[4:09:14 AM] dutu: yes, this is understandable
[4:09:15 AM] agibby5: i'm asking you what, other than variance would cause this
[4:09:33 AM] agibby5: and along the way, i'm trying to think of reasons and making up analogies that i can understand
[4:09:35 AM] dutu: OK,
[4:09:56 AM] dutu: as you can see the payout from bitminter at least is done every ~5 btc
[4:10:04 AM] dutu: if you mine 7.5 btc a day
[4:10:10 AM] agibby5: it looks like 5, then it jumps to 7.2 i think for the next few
[4:10:14 AM] dutu: one day you get 5 the second day you get 10
[4:11:03 AM] dutu: but in average and for a long period of time the value should reach close to the theoretical value
[4:11:25 AM] agibby5: yeah, this seems to be a prime example of luck and variance
[4:11:29 AM] dutu: now when mining income is verly low cvolume you'll probably see that variance is almost none
[4:11:37 AM] dutu: it's not luck
[4:11:44 AM] dutu: it's how the pool pays out
[4:11:52 AM] dutu: it doesn't pay out in real time
[4:12:01 AM] dutu: it pays out in blocks of 5 for example
[4:12:06 AM] agibby5: yeah, it waits until it finds a block to satisfy those shares mined
[4:12:17 AM] agibby5: it cant payout without hvaing a block reward already found
[4:12:58 AM] dutu: regardeless, the average is the same
[4:13:38 AM] agibby5: if i had a machine in my house, i'd swallow that variance and i cannot call teh pool and complain
[4:13:43 AM] dutu: and you ahev it here
[4:13:43 AM] dutu:
http://www.bitcoinx.com/profit/[4:14:31 AM] dutu: well, I'm mining with my machines at home for quite a long time already... and have no variance
[4:14:48 AM] dutu: but the payment comes in portions
[4:14:57 AM] dutu: one day 5 the next day 10
[4:15:11 AM] dutu: because they pay every 5
[4:15:19 AM] agibby5: so how do you explain and convince GB2+ members that GB1 members deserve a portion of the payouts they generated?
[4:15:42 AM] agibby5: because i think the expected/theoretical comparison doesn't work
[4:15:50 AM] dutu: GB1 do not deserve portion GB2 has generated
[4:16:44 AM] agibby5: where does the "adjustment" amount come from?
[4:16:51 AM] agibby5: the later earnings of GB1 machines?
[4:17:41 AM] dutu: the adjustement comes from coins mined by GB1 which were not receiving the walled before the 16
[4:17:43 AM] agibby5: if you make a sheet with that info, i think that proves your point.
[4:17:58 AM] agibby5: you'd have to do a sheet of the GB1 income only
[4:18:17 AM] agibby5: the GB2 mix i think confuses things too much
[4:18:48 AM] dutu: cell E24 is amount recived at wallet by GB1
[4:19:02 AM] dutu: this is too low
[4:19:30 AM] dutu: because the coins were received on the 16, that is cell E26 which is too high
[4:19:32 AM] agibby5: if you can show that they earned 5 btc on day 1, and 10 on day X... but day X payouts were distributed incorrectly to GB2 members, then perhaps you have a point
[4:20:06 AM] agibby5: because it shouldnt take almost an hour long convo to hash this out
[4:20:12 AM] dutu: i can but i need the list of transactions from the account at the pool
[4:20:27 AM] agibby5: cant you use the blockchain?
[4:21:04 AM] agibby5: i think GB1 is the one payout address, and GB2+ is the other, right?
[4:21:19 AM] dutu: I ahve used the blockchain to show that for GB1 miners thay neve mot received the complete amount
[4:21:29 AM] dutu: 51 received
[4:21:33 AM] dutu: expected 85
[4:21:48 AM] dutu: the delta was received on the 16
[4:22:22 AM] dutu: but really i won't spend more time on this
[4:23:30 AM] dutu: I have received no feedbak
[4:23:37 AM] agibby5: so maybe i'm stupid, but why does it take me this long to undersatnd this? even with your spreadsheet and almost hour long convo
[4:23:41 AM] dutu: so what's the pojt to spend time and explain more
[4:23:59 AM] agibby5: you've not presented the data in an easy to understand way
[4:24:07 AM] agibby5: and convincing others is going to be tough
[4:24:15 AM] agibby5: you probably didnt get a reply becuase most peopel don't get it
[4:24:46 AM] dutu: that's soniq job to figure out correct payment
[4:24:49 AM] agibby5: i know larry didn't know what you were talking about
[4:25:08 AM] dutu: i'm sure he had zero interest to figure that out
[4:25:28 AM] agibby5: i didn't really follow it until we talked
[4:25:37 AM] dutu: and he will have zero interest to figure out he owns us 76 BTC
[4:25:58 AM] agibby5: so basically the answer is the timing that the pools paid out fell into the territory of time where GB2 had joined, but the income was accrued when only GB1 was active?
[4:26:12 AM] dutu: correct