Topic: Chain Archaeology - Answers from the early blockchain (Read 9155 times)

Hi,

Very big improvement today. WOAW !

PUR3 is in fact.. Dustin Trammel !

Bloc 956 > https://mempool.space/fr/block/00000000677e3263de6e6972545f43c9eec825a1d39acbfc5820a94519f1684a
> reward sent to https://mempool.space/fr/address/04946fe4f658ad101126dbe9ca1abbdf9af65caade82862704f8c6e449fe1fb67a24243913502cde52863fe97ffbbc13d3e22eee1a12d4ab6888fa8338c9322e1f
> sent to https://mempool.space/fr/address/12DLpSS6fEHqKmYeRcW4iqwAMjPx53U8mL
> sent to https://mempool.space/fr/address/16R6YGi83cS8nv3ECnV2SLA7AJS4eq6kM9
> this latest address is clearly linked to Trammel from this tx: https://mempool.space/fr/tx/4f2b1766198ff7d4902b7da87652682ffeb098c3f92cc69f4b736f03a0caed93
> and the 12higDjoCCNXSA95xZMWUdPvXNmkAduhWv address is known to be Dustin's one (if he didn't lied about the emails): https://www.bitcoin.com/satoshi-archive/emails/dustin-trammel/7/

Bloc 913 > https://mempool.space/fr/block/00000000eec9714483685a04a58d487d758eadf0b0b15d9db38dd632ccb106f8
> rewards sent to https://mempool.space/fr/address/04e53e82cea4edf62e6d5482c6a7922b97f40af780c86072cf7fb6db8a05e6e38e5328a6fd3c28cfcba9f8d00e9982d531fdcc07cd00fb9058db8b51a275b256f5
> sent to https://mempool.space/fr/address/1PUR3qV3zKcGYXDhmT8o7ekCJP7hnSGzwj
> this latest address is clearly linked to Trammel from this tx: https://mempool.space/fr/tx/6db71e70bd0e5b9f2cc0678ff2bdb47215e2adc68dc72bdac973d93c76a54b7b
> and the 12higDjoCCNXSA95xZMWUdPvXNmkAduhWv address is known to be Dustin's one (if he didn't lied about the emails): https://www.bitcoin.com/satoshi-archive/emails/dustin-trammel/7/

I got blocs 913 and 956 linked to Trammel. I can conclude PUR3 is Trammel (yeah !)

Hi guys,

I'm really enjoying searching deep into the begining of Bitcoin's timechain. I just wanna thank you for all your work (which i've checked myself); and I can add that block 752 had been mined by Hal: https://mempool.space/fr/tx/52b9d1bd1bda273869d0442d41b3244d45ef7f0a386b313dd56debf9485cb1fa

I gonna update my post aswell if I find anything else

Thanks again for your work, Taras !


EDITS:
- block 286 had been mined by Satoshi (cf the 25 BTC tx to D. Trammel)

Very interesting thread.

With a difficulty floor of 1, how likely is it that Satoshi or others had to use more CPU power, that is 'work harder' than they anticipated, to prevent block times being delayed too much? Was mining more experimental in 2009, such that long gaps between blocks were of little consequence, or was it vital to maintain a certain level of hashing power to keep the block times as consistent as possible? My impression is that mining was much more of a curiosity for computer geeks back then, but the early days of bitcoin are so murky, it's hard to tell. I didn't start mining until 2010, and by then, block times were important and bitcoin was slightly more than a curiosity.

So what you are suggesting is that whatever processors were used to hash in the first three months were more powerful by a factor of two than those used in the next six months - on average?

If the first three months produced 10,000+ blocks at 50 coins in each block, then that is half a million coins or approximately half of what has been attributed to Satoshi?

Now that time is a variable, we can see that these four blocks which confused me previously have a large delay in the middle.
For the two unspent blocks on the left to have survived, they must have been mined by the person who mined the two on the right as well, because nobody else would have them written down unless they both stopped simultaneously.
Considering the last two blocks follow the same eN-line as satoshi's confirmed one, we can finally confirm these four blocks were mined by satoshi.

The hashrate wasn't good enough for ten minutes a block at difficulty 1; And difficulty never goes lower than 1.
So, the difficulty won't change at all until the hashrate rises enough so that blocks are solved faster than every ten minutes.

I must still be missing something about the patterns of bitcoin generation in that first year. Why would the rate of mining success fall to roughly half of the first three months rate during the next six months?

If we look at an example of 10-20 minute intervals required to generate a block (there are 1440 minutes in a day), then it seems possible that 72 to 144 blocks could be generated in a day.

Thus if we use 90 days as a rough calculation of the first three months, then somewhere between 6480 to 12960 blocks would be generated in that first three months.

We know from Martti's first block sample in April that the block interval of 10 minutes was pretty much on track for the first three months - yet also indicates that there was at least someone running all the time.

If however, the rate of block generation over the next six months is more like 20 minutes on average (ie half the rate of the first three months), what is that due to?

- there wasn't at least one node working all the time
- the size of the block changed the solution rate
- something else?

Do we have any other examples like Martti's six month long sample that are identifiable?

Here are some interesting quotes to answer that question.


Essentially it was (coincidentally) around one year after release that bitcoin's difficulty rose for the first time. Before then, blocks averaged at more than 10 minutes apart and the difficulty could not ever go below 1.00, even if the network does 1khash/s.

I would like to know the answer about this, too. The difficulty was 1.0 through 2009, see graph. the choice of date seems deliberate though, or it is a random coincidence.

‏@marttimalmi posted  on his twitter feed a few months ago the following message:

"Found the first known bitcoin to USD transaction from my email backups. I sold 5,050 BTC for $5,02 on 2009-10-12. https:/[Suspicious link removed]/8XcBmzJljf "

I was confused by some aspects of this transaction.

I didn't think that BTC valuations were that high that early. The pizza story in 2010 gets all the press, but I have seen New Liberty Standard references for $1 = 1,309.03 BTC.

In any case, Martti appears to have begun mining in early April 2009 and continued into October just before he spent the 5050 BTC. Martti's earliest block number was a little over 10,000 - indicating that there was a lot of mining from January into April.

Yet, by mid October 2009, Martti's spend was only at block 24835.

The difficulty didn't change from 1 until the end of the year, Is that because both Satoshi and various early adopters stopped mining or am I missing something about the gating of the hashing rate?

Edit: This is the blockchain link https://blockchain.info/tx/7dff938918f07619abd38e4510890396b1cef4fbeca154fb7aafba8843295ea2

A bitcoiner sent me a very very interesting article today.

Screenshot of Hal's first transactions included.

Bgo is Hal.
Hal mined block 235, which he was previously unknown to have found.

Edit 15:47. I checked all other transactions visible in the screenshot; they coincide perfectly with what I've concluded previously. Feeling really good about myself right now.

Edit 16:32. Here's a list of spent blocks we've linked so far:

0009 Satoshi
0078 Hal
0235 Hal
0268 Vaga*
0309 Druid
0320 Druid
0329 Druid
0357 Druid
0360 Bbz*
0361 Hal
0372 Hal
0394 Druid
0407 Druid
0413 Hal
0417 Vaga*
0419 Hal
0431 Vaga*
0433 Druid
0439 Druid
0442 Vaga*
0450 Vaga*
0461 Druid
0463 PUR3*
0465 Druid
0473 Druid
0490 Hal
0493 Druid
0501 Zzz*
0506 Zzz*
0509 Druid
0512 PUR3*
0521 Druid
0528 Hal
0541 Druid
0562 Druid
0563 PUR3*
0567 Hal
0575 PUR3*
0591 Druid
0596 Cxak*
0598 PUR3*
0607 Druid
0614 PUR3*
0624 Druid
0651 Hal
0658 PUR3*
0666 Druid
0685 Hal
0687 PUR3*
0699 PUR3*
0702 PUR3*
0707 Hal
0720 Hal
0726 PUR3*
0728 Hal
0730 Druid
0739 Cxak*
0748 Zqcym*
0757 Druid
0767 Druid
0772 Cxak*
0773 Druid
0777 Hal
0786 Druid
0803 Hal
0809 Druid
0813 PUR3*
0814 Druid
0819 Hal
0821 Esewdm*
0824 Druid
0828 Druid
0842 Hal
0850 Kwgdyop*
0869 Hal
0885 Zqcym*
0905 PUR3*
0913 PUR3*
0923 Kwgdyop*
0927 Esewdm*
0935 Druid
0940 Druid
0945 Druid
0949 Druid
0955 Narwhal*
0956 PUR3*
0958 Hal
0959 Druid
0964 Narwhal*
0966 Hal
0979 Narwhal*
0984 Kwgdyop*
0986 Esewdm*
0992 Druid
0994 Narwhal*
0996 Narwhal*
0998 Kwgdyop*
0999 Druid
*pseudonyms invented by me

Edit April 5 23:09. I'll now propose a system of nomenclature for early mining patterns that remain unspent.
Rather than getting a pseudonym based on characters in the address, they will be represented by USM followed by the block number where their pattern was first seen. So, the first person other than satoshi to use bitcoin would be USM-12. (No, not Hal)

Edit 23:15. That stands for Unspent Miner.

Edit 23:23:

Here's the next 150 blocks.

Edit 23:24:

Satoshi omitted. See the lines? I sure do. I think I'll get to the end of this iteration before identifying them.

It's possible satoshi was only running one computer at the time, and turned it off while he slept? Since we don't really know if he lives in Japan, we can't assume what day and night for satoshi is yet. Maybe if we can get more data.

Odd timings. Why would Satoshi drop out at 3PM and not restart for an entire day, then next drop out at 6AM and revive at 2PM? If 3PM was in his local time at night and something caused his computer to stop (a bug, perhaps, or network problem; has happened to me many times with overnight runs of something), then what caused a stop halfway around the clock at 7AM, which would likely be during the day?

Here's a fullrender of the first thousand blocks.
It's time to begin identifying our new miners from 660 to 1000.
Looking at this graph makes me proud. The discovery, it beckons...

Edit 3/22 15:48. I'm going to start using my local time, because it's a pain to convert.
I'm going to take a break from the higher blocks and start looking back on the foundation.
Below would be blocks 0 to 25 plotted not by height, but by timestamp:

It seems that both satoshi and the first early adopter (who picked up block 12) have stopped mining shortly after Jan/09/2009 04:33:09.
Bitcoin was essentially dead for 24 hours, until satoshi put his rig back on (and in doing so, reset his extraNonce and begin the next iteration) and found block 15 on Jan/10/09 04:45:46. (He then found block 16, an impressive 12 seconds later (a record at the time.))

I'm not certain why satoshi mined block 0 on the 3rd, and then waited for the 9th to release the client.
There's probably a good reason, but we'll see. Maybe.

Edit 15:59. Here's all the action that took place on January 9:

I'll be honest, I thought we'd see a constant slope. I'm no expert, but I can officially conclude I have no idea what extraNonce is even for, or how it behaves. Just that we can see some significantly applicable patterns.

Edit 16:14. From the data I've got so far, January 10 is cut in half. Bitcoin "died" again at around 7 AM, to be revieved at 3:30 PM, 8 hours later.
Here's the first half:

The two blocks on the bottom right *might* have been satoshi's. I'll get some more data, and see if we can link them to Mr. Nakamoto.

Edit 16:19. Here's the revival of the bitcoin network, up to the end of the iteration.

Those two blocks on the bottom left are not the same as the ones on the bottom right of the previous chart.
Those four have always confused me. Satoshi mined at least two of them... Which ones, I don't know.

Edit 18:34. Whoever mined 955 has an extremely impressive amount of coins.

Edit 18:49. There are so many addresses that instead of choosing a pseudonym from characters in it, I'll just call this miner Narwhal. Get it? It's a whale? Well, if you have a better idea, let me know.

Edit 18:53. Whoever mined 956 was more likely than not PUR3.

Edit 18:57. PUR3 is Druid?

Edit 18:59. If PUR3 is Druid, then he had more than one miner in the triple-digit blocks.

Edit 19:03. 964, Narwhal.

Edit 19:11. 979, Narwhal.

Edit 19:13. 994, Narwhal.

Edit 19:14. 996, Narwhal.

I've labeled the blocks by their miners, identified and pseudonymous.
We've got a few lines here.
Also, I believe we're going to see Hal again soon. Any of these pseudonyms could be him.

Edit 9:28 PM UTC. I've linked Hal to the former Yogbaf. What the hell.

Edit 9:30 PM UTC. YUYUX is Hal Finney. It's highly likely that Jouat is as well.

Edit 9:33 PM UTC. I feel like I should just jump to conclusions and make connections based solely on extraNonce slope.

Edit 9:36 PM UTC. I guess it's what the data is there for.

Edit 9:37 PM UTC. This will be the first time I've used a bulleted list on this forum.

• Jouat is Hal.
• YVig is Druid.
• Pxum is Druid.
• CLnut is PUR3.
• RhhRY is PUR3.
• PxTAX is PUR3.

Edit 10:17 AM UTC. Here's a list.
009 Satoshi
078 Hal
235 Bgo*
268 Vaga*
309 Druid
317 (Moved to single addy)
320 Druid
329 Druid
357 Druid
360 BBz*
361 Hal
372 Hal
394 Druid
407 Druid
413 Hal
417 Vaga*
419 Hal
431 Vaga*
433 Druid
439 Druid
442 Vaga*
450 Vaga*
461 Druid
463 PUR3*
465 Druid
473 Druid
490 Hal
493 Druid
501 ZZz*
506 ZZz*
509 Druid
512 PUR3*
521 Druid
528 Hal
541 Druid
562 Druid
563 PUR3*
567 Hal
575 PUR3*
591 Druid
596 CXAK*
598 PUR3*
607 Druid
614 PUR3*
624 Druid
651 Hal
658 PUR3*
*Pseudonyms. Names invented by me, based on the conjuncted coin's addresses. Subject to change when identity is discovered or pseudonymously conjoined.

Edit 8:59 PM UTC 20 Mar 2014. Just as a recap, those extremely fine slope lines don't just happen. I had to jump to conclusions, and I guarantee that the corresponding addresses belonged to the same miners.

We're down to a few big ones: Satoshi, Hal, Druid and PUR3, who has not yet been identified.
Bgo, Vaga, BBz, ZZz and CXAK will probably never be identified. But we'll have to wait and see.

Edit. I'm going to stop including the time and date in edits. Just for now. It's march 21st, ten days before the first anniversary of me finding bitcoin.
Identifying PUR3 is going to be fun. By that I mean painstakingly difficult.
Maybe we never will. That's also a possibility.
Anyways, we will assume blocks on a slope (assuming it's damn perfect) were mined by the same computer. That includes satoshi.
Sergio's tip (that satoshi only ever spent block 9) will prove to be very helpful assuming we don't disprove it.
I'll start up a website about early miners and their blocks, including a list of blocks known to be mined by Satoshi Nakamoto.
Oh, and bitcoin was premined.

Edit. Keep in mind premined coins make up less than 1% of satoshi's alleged stash.
Blocks 0 to 10 were mined before the qt was released.
Does block 0 count? Well, the private key to 1A exists. (Unless satoshi destroyed it.)
"But Taras, I read somewhere that the first fifty bitcoins can't be spent!"
Mostly true, humble reader.
It's the first transaction, not the first address, that causes this.
The transaction exists in the blockchain but is not recognized by the client.
So, when there was only one client, satoshi could have spent it. Once there was two, the only way to spend it would be to either
a) wait for there to be only one client
b) make the transaction spendable
Option B would require a hardfork, and that may or may not be worth the 50 BTC behind it.
Satoshi's got rooms full of coins though. I think those fifty will be ignored for a long, long time.
This also means that other outputs sent to 1A are spendable. So, donating to 1A isn't verifiably destroying coins; it's giving them to the person who deserves them most. _{Not that he needs them.}

Edit. Fullrender to block 659.


Edit.

Figuring satoshi's pattern from this is going to be harder. Look at blocks 666 to 670. Which extraNonce route was taken by satoshi?
Also, block 666 technically puts 666 on all our hard drives. So we have the mark of the beast to use the devil's currency and marketplace!

Edit. PIUK HAS HEARD ME

We can now see if a block has been spent or not in one click. That click being navigating to the next block, because we know already.
This is making my life so much better you don't even know.

Edit.

Out of all those blocks, only 666 is mined.
This is all the evidence we need that satoshi took the lower eN route.
That was lucky... I was thinking we'd never know.

Double-take three times edit:
Crap man. Another eN fork in the path at 672-675. I didn't even notice it before.
Here I am all choked up with glory and there's a--- wait wait wait WAIT
Block 673's nonce is slightly higher than 675's. Wow I'm retarded nevermind.
672 674 675 676 all satoshi's.

Edit.

Lookin' GOOD.
As for block 666, it was mined by Druid.
^{INSERT SHITTY CONSPIRACY THEORY HERE}

Edit. Moving on.
Blocks 685, 687, 699, 702, and 707 are spent.
Satoshi's pattern ends on 711.

Edit.

I'll now refer to satoshi's pattern areas as iterations.
An iteration ends when satoshi's extraNonce resets.
Iteration 0 was just block 0, by the way. We're on #9 at this point.
I'll plot graphs now based on these iterations rather than friendly numbers.
Above are blocks 660 to 711. 666 was Druid, we know that already. I'll start tracing the coins from the other five blocks.

Edit. Block 685. Hal.

Edit. Block 687. PUR3.

Edit. Block 669. PUR3.

Edit. Block 702. PUR3.

Edit. Block 707. Hal.

Edit. This post is getting significantly (and annoyingly) long but I don't want to bump until I've made significant progress. Maybe a fullrender up to 1,000 would be good for the next post.

Edit. I'm parsing iteration 10 and starting by sorting out satoshi's blocks.
The extra nonce is going through 2,300 now. It's the biggest iteration of all time (at it's time).

Edit. Wow, it's finally over. Block 1295 is the top of the iteration with extranonce 2,614. Time to find which blocks from 713 to 1295 are spent.

Edit. 720, Hal.

Edit. 726, PUR3.

Edit. 728, Hal.

Edit. 730, Druid.

Edit. 739, CXAK.

Edit. 748, new kid in town. 18L8V7DaFBjAbzF5rzmZqCym31KAbjDQXC. Designation ZqCym.

Edit. 757, Druid.

Edit. 767, Druid.

Edit. 772, CXAK. This is the last time we'll ever see that one.

Edit. 773, Druid.

Edit. 782, PUR3.

Edit. 777, Hal. Lucky.

Edit. 786. I THINK it's Druid. I'll investigate later.

Edit. 803, Hal.

Edit. 809, Druid.

Edit. 813, PUR3. I'm starting to get hungry. But I need to know more.

Edit. 814, Druid.

Edit. 819, Hal.

Edit. 821, another short-lived early adopter. Sent 150 BTC to 1EQkvhEsewdmDuvNTSgvYafu79F8Q4B3CW and left. Designation Esewdm.

Edit. 824, Druid. It's funny; I know the addresses by heart. This one just said 'Pxum' instantly in my head.

Edit. 828, Druid.

Edit. 842, Hal.

Edit. 850. I think this is another new miner, 1PxeCXMZBuXHt4CqWWEQ7Kwgdyob9P955L. I'll check to make sure.
Also, there's a clear mining pattern of unspent blocks in this area. Shame.

Edit. This IS a newcomer. I checked all the addresses they have. Welcome, Kwgdyop. They left later on, probably before this iteration ended.
I can't take it any longer. I need beefaroni, now.

Edit. All the noodles came out of the can, without me needing to scrape any out. This is turning out to be a great day.

Edit. I heard a popping noise coming from the microwave's direction. Good thing I covered the pasta.

Edit. 15 seconds left woo

Edit. I have successfully eaten a bowl of beefaroni.

Edit. 869, Hal.

Edit. 885, ZqCym. This would be that miner's last block.

Edit. 905, PUR3.

Edit. 913, PUR3.

Edit. 923, Kwgdyop.

Edit. 927, Esewdm.

Edit. 935, Druid. This confirms 786 was also Druid.

Edit. 940, Druid.

Edit. 945, Druid. 3x combo!

Edit. 949, Druid. 4x combo!

Edit. 955. Newcomer? Possibly Hal or Druid. Spent to 1DLyRUi4ibz26TAxsWsXyRDFVtgpJkYWPW.

Edit. 956. I have no idea who the hell this is.

Edit. 958, Hal.

Edit. 959, Druid.

Edit. 964, will investigate later.

Edit. 966, Hal.

Edit. 979. This is getting confusing.

Edit. 984, Kwgdyop.

Edit. 986, Esewdm. That one's gone forever now.

Edit. 992, Druid. Wow there are so many non-satoshi blocks to go before 1000.

Edit. 994, somebody.

Edit. 996, I'll figure this out. Just let me get to 1000, I mean damn.

Edit. 998, Kwgdyop.

Edit. 999, Druid.

Satoshi got 1000. Congrats, nakamoto.

Edit. Now I'm going to do a fullrender in a new post.

Edit. Looking at the fullrender and I screwed up somewhere.

Edit. God damn how did this happen.


Edit. Here's where I went wrong.


Edit. 752, Hal.

Edit. 760, tragically, is unspent. At least now our fullrender is done properly. I'll be more careful to prevent errors like this in the future.

I will in the future

Will you post a list of those blocks?

Sorry about all the noise but I've got to start a new post.

I'll start paging through the next batch of blocks now.

Edit. 8:28 PM UTC. Block 509 is spent and obviously mined by Druid.

Edit 8:30 PM UTC. Block 512 is spent to 1CA2W1u66UtA1XoqhXmfCLnutSA1Szk8F7. It's a new one if I'm not mistaken. Designation will be CLnut.

Edit 8:32 PM UTC. Block 521. Pxum.

Edit 8:33 PM UTC. Block 528. Jouat.

Edit 8:33 PM UTC. Block 541. Druid.

Edit 8:34 PM UTC. Block 562. Pxum.

Edit 8:35 PM UTC. Block 563 was mined and the coins were sent to 16qPL5UJzsKCaQ2AAFiRQrYY9KfiXRhhRY. Another pseudonym in line. How about RhhRY.

Edit 8:37 PM UTC. Block 567. YUYUX.

Edit 8:38 PM UTC. Block 575. PUR3.

Edit 8:40 PM UTC. Block 591. YVig.

Edit 8:41 PM UTC. Block 596. We've got another lost miner like Vaga; the coins went to 1C16qVq3kZpfjCXAKvrrppqxNmRcnfbFdu along with two other generations. We'll call this forgotten fellow CXAK. I promised myself I'd use five letter pseudonyms from now on, but we're only going to see this one two more times.

Edit 8:43 PM UTC. Block 598. RhhRY.

Edit 8:44 PM UTC. Block 604 is the first unspent block that isn't satoshi's that I've seen since I've started looking through this batch. Unbelievable.

Edit 8:46 PM UTC. Block 607. Druid.

Edit 8:47 PM UTC. Block 614 mined and sent to 6aZtmtjCvjT9rEZTz9CV98KJeQvVPxTAX. Another new one, we'll designate PxTAX.

Edit 8:49 PM UTC. Block 624. Pxum. extraNonce is 1.

Edit 8:51 PM UTC. Block 651. Jouat.

Edit 8:52 PM UTC. Block 658. PUR3.
Block 659 is the peak of satoshi's "extraNonce pattern" (we have to make up a name for that.)
It's at this point that I will plot the blocks.

Edit 8:56 PM UTC:

Here's our chart, including the newly parsed blocks.
The four unspent ones between 600 and 650 immediately appear to have been the same computer.

Edit 8:58 PM UTC:

I've omitted satoshi's blocks for readability.

Edit 8:59 PM UTC. Let's have a look at what we've observed.
009 Satoshi
078 Hal
235 Bgo*
268 Vaga*
309 Druid
317 (Moved to single addy)
320 Druid
329 Druid
357 Druid
360 BBz*
361 YUYUX*
372 YUYUX*
394 YVig*
407 Druid
413 YUYUX*
417 Vaga*
419 Jouat*
431 Vaga*
433 YVig*
439 Pxum*
442 Vaga*
450 Vaga*
461 Druid
463 PUR3*
465 Pxum*
473 Pxum*
490 YUYUX*
493 Pxum*
501 ZZz*
506 ZZz*
509 Druid
512 CLnut*
521 Pxum*
528 Jouat*
541 Druid
562 Pxum*
563 RhhRY*
567 YUYUX*
575 PUR3*
591 YVig*
596 CXAK*
598 RhhRY*
607 Druid
614 PxTAX*
624 Pxum*
651 Jouat*
658 PUR3*
*Pseudonyms. Names invented by me, based on the conjuncted coin's addresses. Subject to change when identity is discovered or pseudonymously conjoined.

I've circled blocks we know Druid mined. Look at the 400-500 range. See some potential lines of incremental extraNonce? Me too. Maybe those who mined those blocks are also Druid? It looks like the same machine got them. We'll see.

Edit:

These are xLyn blocks. It looks like again we'll have the opportunity to merge some pseudonyms.

Edit:

Circled are Vaga's blocks. Vaga also mined block 268 or something, and when you draw a straight line from these four to that one, you get a pretty straight line.

Edit:

This line is probably going to merge everyone on it into Druid. It's probable but we'll wait and see.

Edit:

Looks like Pxum and YVig are the same miner. More charts to follow.

Edit:

Here's all the other blocks. Next edits will be me making connections.

Edit.
First thing's first.

Vaga and PUR3 are on the same slope. Vaga put their coins into a single addy at block 586 though, and PUR3 has eons more blocks further in the chain. It's possible that they're the same person, but it looks like this is a coincidence.

Edit:

Yogbaf, xLyn and Jouat seem to form a single slope.
Yogbaf and xLyn's change addresses spend their outputs to 1LseHYUYUXkL4Jdd4YTTXvWNqDceTdQXrt, YUYUX's address.
This means that Yogbaf, xLyn and YUYUX are the same person, hereon referred to collectively as simply YUYUX.
Come to think of it, YUYUX's block 490 also fits on the slope that Jouat's and the former Yogbaf/xLyn's blocks are in.
Jouat at this point appears to obviously be YUYUX as well, but their coins are a bit harder to connect. I'll do my best.

Edit. I can't make a connection between 1LseHYUYUXkL4Jdd4YTTXvWNqDceTdQXrt and 17afxUJouat3fkaaQ9tZrDThxdkXGL4WrM. If anyone else can, please let me know but at this point I can't connect YUYUX to Jouat.

Edit:

I've labelled the blocks on Druid's slope.
It looks like Pxum and YVig are Druid, but I'm not sure yet.
I'll try and connect the coins.
There's an unspent block on the slope too, but that could be coincidence. It's a tiny bit off.

Edit. I can't connect Pxum OR YVig to Druid. It HAS to be, but I cannot see where the coins conjoin with Druid's address.
It's worth noting that YVig's address still has 1 BTC in it.

Edit. I figured why not ask Druid himself?
He hasn't been online in a while but we'll see if he has any answers for us.

Edit. Here's what we've learned so far:
009 Satoshi
078 Hal
235 Bgo*
268 Vaga*
309 Druid
317 (Moved to single addy)
320 Druid
329 Druid
357 Druid
360 BBz*
361 YUYUX*
372 YUYUX*
394 YVig*
407 Druid
413 YUYUX*
417 Vaga*
419 Jouat*
431 Vaga*
433 YVig*
439 Pxum*
442 Vaga*
450 Vaga*
461 Druid
463 PUR3*
465 Pxum*
473 Pxum*
490 YUYUX*
493 Pxum*
501 ZZz*
506 ZZz*
*Pseudonyms. Names invented by me, based on the conjuncted coin's addresses. Subject to change when identity is discovered or pseudonymously conjoined.

009 Satoshi
078 Hal
235 Bgo*
268 Vaga*
309 Druid
317 (Moved to single addy)
320 Druid
329 Druid
357 Druid
360 BBz*
361 Yogbaf*
372 xLyn*
394 YVig*
407 Druid
413 xLyn*
417 Vaga*
419 Jouat*
431 Vaga*
433 YVig*
439 Pxum*
442 Vaga*
450 Vaga*
461 Druid
463 PUR3*
465 Pxum*
473 Pxum*
490 YUYUX*
493 Pxum*
501 ZZz*
506 ZZz*
*Pseudonyms. Names invented by me, based on the conjuncted coin's addresses. Subject to change when identity is discovered or pseudonymously conjoined.

Block 372 is spent with a bunch of other generations to 1CEq15CvdwHZmkfpWxse1UDGBxLynQepTe for 2,000 BTC total in 2011. The circumstances are similar to Yogbaf's coins, but there are no connecting points yet. Pseudonym will be xLyn.
Block 394 is spent along with 7 other generations to 1FBgZPF9ceV7JFDgvWzwJdPjdn6VFJqc5s for 400 BTC on the 20th of 1/09. The coins then move to 1LNYVigk4tKYEduYnA9XDNJCFrm8wo4XAp, summing to 2,000 BTC. This is also in 2011, with 2,000 BTC and change, like xLyn and Yogbaf. I'll call this one YVig for now.

Edit.
Block 407 spent to 12higDjoCCNXSA95xZMWUdPvXNmkAduhWv. Druid.
Block 413 spent to 1CEq15CvdwHZmkfpWxse1UDGBxLynQepTe. xLyn.
Block 419 spent to 17afxUJouat3fkaaQ9tZrDThxdkXGL4WrM, 1,000 coins total, January 31. Designation of pseudonymity will be Jouat.

Edit.
Block 433 spent to 1FBgZPF9ceV7JFDgvWzwJdPjdn6VFJqc5s. YVig.
Block 439 spent to 15oUEZFKAC8E8BTLt1s1jx4fPxumwB3ecr, 500 coins total, January 19. This is a new one. Designated Pxum.
Block 461 spent to 12higDjoCCNXSA95xZMWUdPvXNmkAduhWv. Druid.

Edit.
Block 463 spent to 1PUR3qV3zKcGYXDhmT8o7ekCJP7hnSGzwj, 45,000 coins total, 2011. The expenditure goes into 2012 that I can see. This one could be connectable to a bitcointalk member but for now I'll call them PUR3.
Block 465 spent to 15oUEZFKAC8E8BTLt1s1jx4fPxumwB3ecr. Pxum.
Block 473 spent to 15oUEZFKAC8E8BTLt1s1jx4fPxumwB3ecr. Also Pxum.
Block 490 spent to 1LseHYUYUXkL4Jdd4YTTXvWNqDceTdQXrt, along with an ass mcTon of other generations for exactly 2,000 coins no change. This is a new one as of yet, we'll call them YUYUX.

Edit.
Block 493 spent to 15oUEZFKAC8E8BTLt1s1jx4fPxumwB3ecr. Pxum.
Block 501, although not yet plotted on our chart, is spent with 506's coins to 133fZZzNNbHL5VSuCWrUkLW2oL9ZPJELbY. They're still there today, this miner, like Vaga, disappeared. We'll call them ZZz.

Edit. Time to draw some lines!

There were probably a LOT of people who simply deleted those keys out of disinterest at some point.
See this chart. Blocks from 150 to 250 are pretty much exclusively unspent.
Meanwhile blocks from 300 to 500 (not including satoshi's coins) are almost all spent.
Coincidence? Maybe. It's still pretty much all ciherpunks at this point, but there might be a new crowd already on day three. I'm not sure what to think, seeing all these blocks going from being all unspent to being all spent.
Maybe it's because it's a smart few (draw the lines), rather than the 150-250 area which is what looks like a dozen people.

Makes me think most early adopters probably lost their privkeys(aka deleted their wallets) long before they realized the true worth of Bitcoins.

Starting a new post. It's getting crowded up there.
I have fancy-ified the chart, and I've done my best to separate satoshi's pattern.
Below are blocks up to 400, after going through and more thoroughly finding spent blocks.

Edit. Fullrender up to block 500:

Edit... That was incomplete data. Whoops.

Edit. Blocks 407, 413 and 417 are spent.

Edit. Block 419 is spent.

Edit. Blocks 433, 439, 461, 463, 465, 473, 490, 493 and 501 are spent.

Edit. Proper fullrender up to 500:


Edit. Shrank points for readability:


Edit. Time to make some connections and identify who mined our spent blocks.
I'll continue updating this section rather than reposting it constantly (unless someone replies to this thread.)

009 Satoshi
078 Hal
235 Bgo*
268 Vaga*
309 Druid
317 (Moved to single addy)
320 Druid
329 Druid
357 Druid
360 BBz*
361 Yogbaf*
372
394
407
413
417 Vaga*
419
431 Vaga*
433
439
442 Vaga*
450 Vaga*
461
463
465
473
490
493
501

Blocks 9 78 268 309 320 357 417 431 442 and 450 identified to this point.
*Pseudonyms. Names invented by me, based on the conjuncted coin's addresses.

Edit.
Block 235's coins were spent in conjunction with the reward from block 8,970 to 1BgoUL3Mf8vDSDU4jjpFARKBJ6zPDpa4RU. Pseudonym will be Bgo for now.
Block 317's coins were moved to a single 50BTC address, 1QKJnFPvRoZ7avBhxqwxhsEv62JWYoB1R9.

Edit.
Block 329's coins were cut in half, some sent to 1LEWwJkDj8xriE87ALzQYcHjTmD8aqDj1f and the rest returned to the gen addy as change. The change is later spent to 12higDjoCCNXSA95xZMWUdPvXNmkAduhWv, meaning they're Druid's coins.
Block 360's coins were sent to 1BBz9Z15YpELQ4QP5sEKb1SwxkcmPb5TMs in conjunction with 11 coins that originated in block 9, sent around the time satoshi sent 10 coins to Hal in block 170. Since this person has a relationship with satoshi, I'll assign the pseudonym BBz.
Block 361's coins were spent to 16ds1vC8yuFkYogbafQjMJZXYsC4xbXdEX in 2011, along with an asston of other generation outputs totalling 2,000 BTC. We might be able to find their owner but it won't be easy to trace the coins. For now the pseudonym Yogbaf from the addy will have to do.

Those three blocks mined by other users were probably just lucky values.
In fact, I'd say they were without any chance of doubt lucky values.


Continuation of research:

I've drawn some sample lines onto the chart.
There ARE some patterns sticking out. Were these blocks found by the same computer?

Also, |)ruid's second block, compared to his first, is sloped like satoshi's extraNonce, but obviously less frequent. I've highlighted them with a yellow line.
Until we see more of his blocks I won't draw any conclusions.
The spent block inbetween |)ruid's blocks, if I remember correctly, was sent to another address and then left there to this day. Can't be |)ruid's.

The purple lines in the bottom right of the chart are very clear. Looks like we've found two early adopters.
The cyan line connects five blocks that were all spent to the same address.

Edit. Looking at the 'jumbled mess of blocks'

There's  not really any conclusions to be drawn from this. Not even leads. Maybe if we have more information... Like another factor...

Edit. I missed block 360 it seems. It's spent.
It is not compliant to satoshi's pattern, but whoever mined it must have known satoshi, for they also owned 11 BTC from block 9.
COULD be Hal. He got another 10 BTC from block 9, and the remaining 29 BTC of block 9 are locked up across a few addresses since 2009.
Not saying it's Hal. I'll investigate.

Edit. It isn't Hal; he'd have spent the coins. They're lodged in 1BBz9Z15YpELQ4QP5sEKb1SwxkcmPb5TMs at the moment.
Our only chance of identifying their owner is through historical conversations.

Edit. March 15. Block 329 is spent.

Edit. Block 357 is spent. Traced to |)ruid.

Edit. Block 361 is spent. Looks like it's Hal's, but not sure yet.

Edit. Block 372 is spent.

Edit. Block 394 is spent.

The nonce is a random value tried to find a hash that < target. So yeah, nonce at this point is random(you may increase it sequentially, but finding blocks with some nonce is still random).

After further investigation,
the nonce factor seems totally redundant at this point.
Back to extraNonce only for now.

I thought it'd be interesting to note that block 167's nonce is 3,735,299 units larger than block 168's. That's the closest consecutive number I've ever seen, but maybe it's not as rare as I think it is.

Oh, and happy block 290,000. That's a lot of blocks.

This is news to me. Satoshi only spent the coins from block 9? I've got to see where they all went for myself.
It's also a good thing for the press. People might think twice about calling ponzi after they hear the one who started it made no money from it.
I could never want to put someone through what Dorian is going through. I'll remember not to make myself.
Chain-archaeology.
I'm going to use this term.

Edit, I forgot to ask you something:
You have data for which blocks are spent, right?
See, currently I'm manually parsing blockexplorer click by click to know.
I mean... If you could share which ones are spent, that'd be golden

I'm trying to connect the normal nonce of a block to blocks we know were not mined by satoshi.
This is probably redundant. But I want to analyze everything.






See blocks 12, 64 and 78.
We know all three were mined by people other than satoshi.
The block's nonces are all less than 500,000,000.

I don't think this is related (it might be but I doubt it), though I'll leave this here for completion.
I'll look at some more blocks we know satoshi didn't mine.

Oh, and it's important to note some of the blocks satoshi DID mine have nonces below 500,000,000. Another reason why I think this is a bit redundant... BUT NO DATA LEFT BEHIND!

I believe Sergio (the author of the post directly above yours) wrote an article with that as a possible thesis.




Sergio, you are an inspiration. I'm going to plot some more factors tonight and look for some patterns. Deathandtaxes's resolution on the qt's extranonce has me wanting to separate as many candidates as possible.
I'll draw some lines. Maybe for that we should plot on time, rather than block height...

Wasn't Satoshi shortly after release mining on 6 different computers parallel?
I thought I read that somewhere.

As I researched some time ago "Satoshi" pattern has a distinctive restricted nonce LSB range, and that's the reason why we see Satoshi pattern to be steeper than the other patterns. But also the pattern is more dense than it is expected, meaning the computer had several threads working on the same ExtraNonce incremental sequence, which suggests something like a special mining rig consisting on a master thread managing 5 child dumb threads, in a quad-core computer.

Regarding having access to the private keys, we know that the generated coins in block 9 (which seems to be part of the pattern) were used to pay for Hal's transaction. I inspected the pattern by eyesight and concluded that never again the coinbase of a block from the Satoshi pattern was spent. For me, this is a clear indication that he wasn't going to use that coins to manipulate the market or he purposely destroyed the keys to access the coins.

After all this media circus we've seen this week regarding his real identity, I suggest we take into account people's privacy. We researchers also make mistakes.
My own research was motivated by the need to understand the market in order to decide if I should invest in Bitcoin or not. And from that day I know that I should have invested more.

Have a nice chain-archeology day...
Sergio.

I enjoyed the pointers on the screenshot

Maybe this will help…


https://bitslog.wordpress.com/2013/04/17/the-well-deserved-fortune-of-satoshi-nakamoto/
https://bitslog.wordpress.com/2013/04/17/more-on-block-mining-history-1st-half-of-2010/
https://bitslog.wordpress.com/2013/04/24/satoshi-s-fortune-a-more-accurate-figure/
https://bitslog.wordpress.com/2013/09/03/new-mystery-about-satoshi/

That is my understanding.  Even w/ all Satoshi's hashpower the entire network didn't have enough hashrate to sustain 10 minutes per block @ difficulty 1.  The average time between blocks in the first year was closer to 14 minutes.


Possibly although the original miner never stored the extranonce to disk so a miner shutting down (reboot, needing computing resources) would reset back to an extra nonce of zero.


IIRC Hal reported his first block was 78 but it is possible he was mistaken and that entire chain is his?  Also the overlap could simply be coincidence.  Hal mined through 78 extra nonces before solving a block.  His block is unrelated to another miner who solved the earlier blocks w/ lower extra nonce value.  I don't know just throwing ideas out there.

One thing to understand is that the early miner was very conservative with use of extra nonce.  Technically on any block change you could reset the extra nonce as the rest of the header has changed but the software didn't.  It simply incremented the extra nonce on every nonce range attempted.

They were
https://blockchain.info/tx/4d6edbeb62735d45ff1565385a8b0045f066055c9425e21540ea7a8060f08bf2
So, satoshi wasn't a special case; He just had a lot more power than everyone else it seems.

If all the early blocks are incremental we may be able to draw a lot of lines, not just for satoshi but for all identifiable early miners.
What still confuses me is the blocks in the 70s range. Hal mined 78. It appears to be at the end of a chain of incremental extra nonces, but Hal didn't mine those. I'll look into it.

Your evidence that Satoshi had a higher hashrate does make sense and support the "blind miner" theory. Satoshi didn't refer to hash rates in the white paper so much as CPU power (See abstract in: https://bitcoin.org/bitcoin.pdf ) so this definitely lines up with him running more than one system. And it would make sense he had a daemon process running somewhere to ensure every 10 minutes a block was found and transactions processed, leaving him free to work on and run other instances on demand as he was developing and testing the software for all we know the "unknown" was also Satoshi experimenting with other approaches.

Interesting work, keep it up.

Regarding the extra nonce the early mining code always incremented the extra nonce until the application is restarted (in hindsight it would have been better to use a random number).  You can find this is the earliest copies of the bitcon software (which still had mining integrated in the client, you were always mining when the client was running). 

So all miners will show an incrementing extra nonce.  However the lower the particular hashrate of a given miner the less steep the slope. For example in the graph immediately above you will see that the red blocks between 400 and 450 were probably all from the same miner.  Even the utterly "random" blocks by unaligned solo miners are using incrementing nonces however if they have a very low hashrate and/or restarted frequently they would never operate long enough (measured in hashes) to find a block with a large extra nonce. 

This seems to be a bunch of random anons. It's wholly possible satoshi stopped mining during this time. There's no pattern whatsoever.

Edit. Looking at the big picture:

The giant leaning towers of pisa are one blind miner, which we are still speculating to be satoshi.
The other blocks down below were early adopters; solo miners.
The four spent blocks on the far right were all mined by the same entity, which I refer to as Vaga.
The extraNonce is incremental. Maybe it's not a pattern exclusive to satoshi? Or it's blind luck.
We are going to need more information to come to a conclusion.

I'm trying to determine:
1) Who mined early blocks, for historical purposes
2) Just how many bitcoins Satoshi has
3) The behavior of the network in 2009

What do you hope to achieve by this?

The above post is getting a bit crowded. I think I'll start a fresh one.
Below are blocks 80 to 162.

Every single block on this graph follows satoshi's pattern...
Except block 131. Unspent, we won't know who mined it. Until it is spent. Which by the way could never ever happen, ever, but we'll see.

Edit.
Below are blocks 150 to 200.

Uhm... Did satoshi turn off his miner? Look at that jumbled mess of blocks.

From now on I'm going to monologue my research.

Block 12 as was said by the guy in the paper was mined by another client. More likely than not a cipherpunk on the mailing list. Whether or not this was that person's only block remains unknown and will stay unknown until the block is spent, which could be forever.

Satoshi's block 14 is the peak of his first eN pattern. It resets for block 15.

Edit. Next twenty blocks below.

I should assume that blocks 15 to 25 were all satoshi's, since it's his distinct (so far) pattern, but I'm not going to jump to conclusions yet.
It resets again after 25. Now, this is interesting. It looks like someone else mined at least one block between block 27 and 29, but I don't think we'll ever know which. The pattern resets again on block 37.

Edit. If the miner anon who mined a block between 27 and 29 mined block 28, they would have had to mine 29 as well so as not to disrupt satoshi's pattern. They also could have mined 26, meaning satoshi's pattern started at anywhere between 26 and 29. These four blocks are the most confusing so far. We'll probably never get a conclusion on them. Not with this data.

Edit. Next twenty below.

This looks like it was exclusively satoshi's pattern.

Edit. Next twenty below.

Block 64 disrupts the pattern and is unspent. We will never know who mined it; Only that it wasn't satoshi.
Block 73 is where satoshi's pattern resets again. The next blocks are interesting:
Block 78, which we already know was mined by Hal Finney, fits into the pattern from 73 to 78.
Were blocks 73-77 mined by someone else? This makes no sense to me yet.
They weren't mined by Hal. His first block was 78, or "block 70-something" in his words.
COULD be satoshi. It fits his pattern, but it's really short, and why would it reset after Hal's block?
Maybe we will never know.

Edit.

The next 20 were mined with satoshi's pattern starting at 79.
This is nuts. Either satoshi was the only one mining at this time, or his hashrate was significantly higher than anyone else.

Edit.
To put things into perspective, here's the global share of mined blocks from 1 to 100.

Here's what I have so far.

0 - Satoshi - Genesis block
1 - Satoshi - Premining starts
2 - Satoshi
3 - Satoshi
4 - Satoshi
5 - Satoshi
6 - Satoshi
7 - Satoshi
8 - Satoshi
9 - Satoshi - First block to be spent
10 - Satoshi - Premining ends, Bitcoin client released
11 - Satoshi - extraNonce pattern continues
12 - Unknown - Pattern disrupted
13 - Satoshi - Pattern continues
14 - Satoshi - This is the block in the first incremental pattern
78 - Hal - Coins traced to his paper wallet, personally claimed to mine "block 70-something"
163 - Unknown - There is no pattern at all it seems until about 190 (see chart below) Has satoshi stopped mining here?
193 - Unknown - Probably satoshi, incremental extraNonce starts
235 - Unknown - Coins spent, possibly identifiable
268 - Vaga
309 - |)ruid - Coins spent in conjunction with coins from 12hrp
317 - Unknown - Coins moved to another address, but are still there today
320 - |)ruid
417 - Vaga - Vaga's extraNonce pattern starts
431 - Vaga
442 - Vaga
450 - Vaga - Pattern ends, Vaga's coins move to one address, Vaga disappears forever

I'm basing this on that original research. I'm looking for not only a way to identify "satoshi" but as many miners as possible.
I will search for any patterns at all. Not just the extraNonce either, I'm going to look into everything.

OpenOffice Calc

Curious, what is your drawing/graph software/library?

This is not the first time this has been attempted via the extranonce:

http://www.bitcoinrumors.com/2013/04/18/the-well-deserved-fortune-of-satoshi-nakamoto-bitcoin-creator-visionary-and-genius/

The standing theory is that back in the early days Satoshi had a "blind miner" running to keep the network ticking over until enough people joined and stayed connected continuously to keep it running and allow the net-work effect to take hold.

Regardless, it's undeniable we can see Satoshi's footprint in those early blocks but whether or not anyone bothered to keep the coins being generated by the "blind miner" process is another question. After all, they had no value beyond academic interest back then.

Look at these cryptic patterns. I can't wait to separate spent from unspent.

I've done my best to trace down the people who mined the blocks with spent coinbases between height 0 and 450.

More extensive research is imminent.

Maybe this was a message intended by Satoshi

I was going to say the same thing. Get your smut out of here 

You know, with all those blue spheres, they kind of look like a particular male organ when stacked like that, if you know what I mean.

Once I find a better way of telling which blocks were spent and unspent, we will have a better idea of which blocks were satoshi's and which blocks weren't.
If anyone mined a block with a height less than 1,000 please let me know.

Absolutely undeniable stuff:
We know Hal mined block 78.
We know Satoshi mined blocks 0 and 9.

Given the extraNonce is incremented by the bitcoin client, the upward trends above sergio theorized as being single clients. This is sound.
The extraNonce seems to reset time to time, but I have no idea why or at what occasion.
Perhaps only satoshi's client did this? Maybe everyone else's extraNonces are just pseudorandom?
I'm going to look into this some more.