EDIT:
Pulled the CSV exports of the staking history of the two public keys listed above (current #3 and #4 on the Rich List -
https://chainz.cryptoid.info/piggy/#!rich), and did some more comparisons:
Rich list #3 (psVEnGqj...)
Baseline analysis (March 13-14) - annualized interest rate of 3.16%
Staking today (3/16) - annualized interest rate of 1.31%
Stakes/hour on 3/16: 9.24 (so far)
Rich list #4 (pjbxSW6y...)
Staking today (3/16) - annualized interest rate of
15.18%!
Stakes/hour on 3/16: 29.22 (so far)
Something is out of whack. It's as if #4 on the rich list is staking at the once-proposed 15% interest rate, and is somehow crowding out everyone else and depressing their interest.
EDIT 2:
From 3/15/2015 21:17 to 3/16/2015 1:56, (pjbxSW6y...) staked 206 times (in 4.64 hours) for 5125 PIGGY, an annualized rate of 53.38%.
Okay, quick answer first ... #4 pjbxSW had more coin-age (had not staked coins for some time) so rewards would "appear" larger. They would be the same if the coins were constantly staked the same as #3 psVEnG. Further, more blocks were awarded to #4 pjbSW because it had more splits (looking in the wallet that 18m was split out to over 1,500 UTXOs) ...
If, for example one were to enter the network with 1x1m coin UTXO that had 1 year coin-age (A) vs another with 10x100k coin UTXO w/same coinage (B) vs a third 1m coin UTXO w/6months coinage, I would expect to see something like the following to occur:
(A) Wins 1 block, staking 1m coins which get split into 2x500k new UTXOs and awarded 30k.
(B) Wins 5 blocks, each staking 100k coins which get split into 10x50k new UTXOs and awarded 3k in each block.
(C) Wins 1 block, staking 1m coins which get split into 2x500k new UTXOs and awarded 15k.
(B) Wins 5 blocks, each staking 100k coins which get split into 10x50k new UTXOs and awarded 3k in each block.
So (A) and (C) only won a block each, whereas (B) won 10 blocks due to splitting. (A) and (B) had a total award of 30k each, which is 3% because they had one year coinage coming into the stake. (C) was awarded 15k as coin-age was only 6mo.
In the above example if you took a sample that included the above vs. those claiming stakes every 10 or so hours they would all appear disproportionate, with (A) and (B) looking like they were earning 100% more than (C) and (C) earning 100s of % more than the rest of the network. One must take coin-age into account.
The real world, re interest differential:
Lets take a look at what appears to be a disproportionately large stake by pjbxSW and coin-age:
Block 426094:
https://chainz.cryptoid.info/piggy/tx.dws?916597.htm Block Date/Time: 2015-03-15 00:14:18
UTXO Block 333745:
https://chainz.cryptoid.info/piggy/tx.dws?722809.htm (In this block you can actually see the UTXO splitting as well.)
Block Date/Time: 2015-01-10 01:17:01
Approx. coin-age-days: 64 days
Interest calculation: 125k * 3% / 365 * 64 days = 657.534 (actual reward 657.08988322 PIGGY) - VALID.
Now psVEnG:
Block 426076:
https://chainz.cryptoid.info/piggy/tx.dws?916561.htm Block Date/Time: 2015-03-15 00:06:57
UTXO Block 425031:
https://chainz.cryptoid.info/piggy/tx.dws?914395.htm Block Date/Time: 2015-03-14 07:04:38
Approx. coin-age-days: 14/24 = ~0.583
Interest calculation: ~63.9k * 3% / 365 * ~0.583 days = 3.062 (actual reward 3.72962237 PIGGY) - VALID.
But if you were to extrapolate based on those two stakes without taking into account coin-age it would appear, as you noticed, that pjbxSW was earning interest orders of magnitude higher than psVEnG. From the above, it is not, you may verify any block manually this way.
Re percentage of network successful stakes:
You will notice now the network is normalizing again, and yes pjbxSW is being used to "test with purpose", "normalize" or secure the network as there is a (pre-existing) issue with stake-splitting (much thanks to greenmo for noticing this behavior TBH). Currently the wallet will split down to a minimum of 1,000 coins any successful stake that has a coin-age of less than 24 hours. In the case of PIGGY we have allot of believers w/ fair holdings mixed with lots of smaller addresses (32.4% are held by the top 10) and actually psVEnG concerns me the most, should it come to be controlled by a bad actor.
Staking is based on UTXOs ... So if I have a low stake network difficulty and I posses enough coins I could theoretically split my coins out to optimize my consecutive stakes. This is effectively what was done with pjbxSW with the view to increase network difficulty but preserve the 1 minute block times. In the example above we saw how 1m coins split out could, given the right conditions, be split to get close to 10 consecutive successful stakes in a row. Given 6 confirmations this is a reasonable attack vector. pjbxSW had been split into over 1,600 UTXOs pre-fork, then held unstaked for a time to insure its height post-fork when it was put to stake again giving the new chain a higher overall difficulty to entice at fork time clients to migrate to the new chain. When removed, the difficulty dropped again due to the splitting down to 1k UTXOs of stakes. A new wallet was compiled with a 256k lower split boundary (and consolidate coins) and pjbxSW was put to stake again raising the difficulty. A release will be made in the next week or so targeting a lower boundary of 48k, by my calculations, which seems to be the sweet spot to balance splits vs difficulty vs blocktime vs attack vector. As the number of coins increase and the percentage of stakers change, this will have to be adjusted, or a dynamic method implemented. Luckily the Team can to a degree influence this, and that was the case when, yes, pjbxSW held 70% of stakes for a while and "adjusted" the network difficulty. All I can say is there is much truth to Sunny Kings comments
https://github.com/ppcoin/ppcoin/blob/master/src/wallet.cpp#L1215Many words, and I hope some explanation ... and here is a signed message I hold pjbxSW6yFSbZAX5rjeN9n1c8EL8oULLwpY (I hope you trust me)
:
Proof of holding. neurocis.
H4jUUR/YfisqpL05JnhyZZQ1j5EhLm2GBWCZ5CzSlNqjiQwMIm5BS3qCcvHrfS7PHe6hnIBujL8q58iOmvkKT1I=
Cheers ... and of course Oink!!!
Poll
