Bitcoin Forum

My original post here: http://www.reddit.com/r/Bitcoin/comments/2mkmoq/til_where_the_bitcoins_at_mysteries_of_the/

TIL: Where the bitcoins at… Mysteries of the Blockchain

I love both the transparency and the pseudonymity that bitcoin provides, so I thought I’d take a gander at the blockchain and see what there was to see.

There are a couple sites that list the “100 richest addresses”, but I’m a curious guy and a bit of a data junky and wanted to see what else the blockchain could reveal.

A little PHP, a current bitcoin node with the full transaction index, a MySQL database, 5 days(!) of processing time on a server and I now have every single unspent output in the blockchain in my DB.

An unspent output is essentially a transaction to a public addresses that not been spent, i.e. has a balance.

I started with the genesis block (#1) and followed the coins all the way up to today. The data is not perfect, there were some transactions I simply could not parse that were either null or non-standard.

Here is some of what I learned as of block 330,433 (Monday November 17th, 2014):

There are currently 14,104,525 unspent outputs in 3,609,649 unique address.

Zero value transactions:

There are 3,393 unspent outputs that contain 0.00000000, this is curious because obviously you cannot spend a 0 balance. In looking at the transactions many of the early ones include a high fee, and then sent 0 BTC to many addresses, talk about blockchain spam and bloat… The first occurrence was in block 82,627 and the latest was in block 329,151.

Here is an example of an early one that simply looks like some kind of spam attack on the network, these stopped occurring so I assume someone fixed it:
https://blockchain.info/tx/9173744691ac25f3cd94f35d4fc0e0a2b9d1ab17b4fe562acc07660552f95518

Here is one of the more recent ones, it only sends 0 BTC to a single address, but its still an unspent output that is worthless:
https://blockchain.info/tx/83277716e29cfaef9525fdd732fa39e008afd7a50302218eaf5073e87a082b24

Perhaps someone with more knowledge then I can shed some light on these…

1 Satoshi Multisig:

There are 16,848 multisig addresses with a 0.00000001 balance. Why? who knows…
example: https://blockchain.info/tx/045727591618d2154239956332f307c754a52c0e7896856a05848ec7065db1c3

Total of 1 Satoshi addresses:

There are 732,959 addresses with a 1 Satoshi balance totaling 0.00732959 BTC

Untouched early mining income:

In the early days of mining using the bitcoin client each 50 BTC mining reward (when 0 transaction fees were included) would go to a unique address, there are still 39,228 transactions with an unspent 50 BTC balance, thats 1,961,400 Bitcoin. Some of these are not newly generated coins, simply a transaction containing 50 BTC, however the vast majority are old mining income. Many of these probably belong to Satoshi…
example: https://blockchain.info/tx/b0a478fc79ecca1173cefe5975dd4a9874593cceb44b66ea4ea14eb40d587a14

Broken addresses:

There are 326 addresses with a balance where the address is “non standard”, a non standard address is one that cannot be decoded by the bitcoin core, and most likely can never be spent again.

Total non standard unspent outputs = 2,614.256208 BTC
example: https://blockchain.info/tx/ddddf9f04b4c1d4e1185cacf5cf302f3d11dee5d74f71721d741fbb507062e9e

Many 1’s:

There are 531 transactions going to a very unique address: 1111111111111111111114oLvT2 (many of them with advertisement like notes)
https://blockchain.info/address/1111111111111111111114oLvT2

While this is a valid address it is very likely that no individual has the private key for it, and the 2.98804915 BTC in it can never be spent



A selection of other unique addresses that are valid and contain a balance, but most likely no one has the key for:

• 1BitcoinKicksAss1111111111114BbAUx
• 11111111111111111111BZbvjr
• 12TisTrueWithoutALie22222221wT3qjn
• 1CertainAndMostTrue2222222225YPnJF
• 12ThatWhich1sBe1ow1sAs222221y3G7mv
• 12ToPerformTheMirac1es222221zShqDE
• 12ofThe1Thing222222222222221wv1hge
• 1AndAsA11ThingsWereFromThe1xxdJJ2U
• 1ByMeansofTheMeditation22221xJd9GS
• 12ofThe1xxxxxxxxxxxxxxxxxxxy1aqzJW
• 12ThusA11ThingsWereBornFrom2vX3xAk
• 191tsFather1sTheSun2222222225iM4gY
• 191tsMother1sTheMoon22222221z3BtXD
• 12TheWindCarried1t1n1tsBe11y31r5NB
• 191tsNurse1sTheEarth22222221zkkvAq
• 12TheFatherofTheWho1eWor1d2249xs5g
• 191sHereXoXoXoXoXoXoXoXoXoXo72uqJv
• 12YouWi11SeparateTheEarth2223tzWxE
• 1FromTheFire11111111111111113MzD3d
• 13Sweet1yWithGreatSki11x3332wyNKnf
• 12oXoXoXofTheWho1eWor1dXoXoXhjQBMS
• 1F1eeFromYou21111111111111111t6TRM
• 1Phi1osophyofTheWho1eWor1d5547QMgV
• 191t1sFinishedWhat1HaveSaid539kgve

The nitty gritty:

• 424,925 transactions with a single output over 1 BTC
• 122,093 transactions with a single output over 10 BTC
• 11,403 transactions with a single output over 100 BTC
• 1,102 transactions with a single output over 1,000 BTC
• 53 transactions with a single output over 10,000 BTC
• 0 transactions with a single output over 100,000 BTC

Some more...

• 732,959 transactions where exact output = 0.00000001 BTC
• 84,155 transactions where exact output = 1 BTC
• 13,946 transactions where exact output = 5 BTC
• 13,438 transactions where exact output = 10 BTC
• 2,683 transactions where exact output = 25 BTC
• 39,228 transactions where exact output = 50 BTC
• 5,358 transactions where exact output = 100 BTC
• 866 transactions where exact output = 500 BTC
• 485 transactions where exact output = 1,000 BTC
• 27 transactions where exact output = 2,500 BTC
• 50 transactions where exact output = 5,000 BTC
• 30 transactions where exact output = 10,000 BTC
• 7 transactions where exact output = 20,000 BTC
• 1 transactions where exact output = 30,000 BTC
• 1 transactions where exact output = 100,000 BTC

Largest single unspent output: 100,000 BTC
Balance: https://blockchain.info/address/1i7cZdoE9NcHSdAL5eGjmTJbBVqeQDwgw
Transaction: https://blockchain.info/tx/796684ce2c46c73eec4e67a4bd530603c2e0529855080187cde241a6b06c2a5b


10 Richest Addresses:

• 1i7cZdoE9NcHSdAL5eGjmTJbBVqeQDwgw - 144,341.54672136
• 13Df4x5nQo7boLWHxQCbJzobN5gUNT65Hh - 134,173.14791082
• 1JEC8vYP9cEDSu6N6DXkkYd3RaeWAdsCqN - 120,223.31441612
• 1FeexV6bAHb8ybZjqQMjJrcCrHGW9sb6uF - 79,957.10818636
• 1HQ3Go3ggs8pFnXuHVHRytPCq5fGG8Hbhx - 69,471.10124134
• 1PnMfRF2enSZnR6JSexxBHuQnxG8Vo5FVK - 66,452.06330135
• 1AhTjUMztCihiTyA4K6E3QEpobjWLwKhkR - 66,378.80759963
• 1EBHA1ckUWzNKN7BMfDwGTx6GKEbADUozX - 66,233.73540359
• 1DiHDQMPFu4p84rkLn6Majj2LCZZZRQUaa - 66,141.88463485
• 194DnvmLR2HULRvxUsVag8mn2fm7dA3U2B - 66,112.30181684

There is a lot more to be discovered, but that’s it for now.

If there is anything you would like to know about unspent transactions just ask, if your question can be expressed as SQL I’ll do my best to get you an answer…

I’m considering keeping the database updated in real time and publishing some or all of this on my website, http://www.CoinCadence.com would that interest you?

http://www.righto.com/2014/02/ascii-bernanke-wikileaks-photographs.html

This is quite cool. Very nice explanation on that url and nice find by OP

This is awesome!

Have you seen this analysis yet?

https://bitcointalk.org/index.php?topic=675321.0

Very interesting. However I think it is possible that at least some of the above addresses could be addresses created via vanitygen or some other similar address generator. Although I would think they would generally be very "difficult" to generate and a lot of effort would have likely been put into making such addresses, I would not discount the possibility that someone has the private keys to at least some of these addresses.

EDIT: Someone almost certainly controls the private key to 1111111111111111111114oLvT2 as there is a custom tag for that address on blockchain.info and in order to create a custom tag, you must sign a message that blockchain.info asks you to sign

No, they almost certainly don't.

That is the address for a RIPEMD-160 hash of all zeroes.  It is extremely unlikely that anybody managed to find a private key that computes to a bitcoin address that is all zeroes in it's RIPEMD-160 hash form.


That is not true.  There are many tags at blockchain.info that were created without signing a message.

I had not, my data contains most of these coins, however some were certainly discarded in my results. Planning to store any found errors for further investigation the next time through...

Thx everyone for the positive feedback!

Multisig testing?

so do you think Satoshi is stocking up on bitcoins because he knows that these things will be worth a fortune,and he will than cash out?Could this mean that we should invest in bitcoin?

i dont think satoshi would endanger his creation. i think the coins will remain there for many many years to come. Moving them all too fast would make the price unstable. i think satoshi had a vision and it wasnt to get rich, it was to give the world a superior currency system. he/she/they would risk harming it.

That is just a assumption such as mine,but my assumption was based off of human nature,as humans we only do anything for two reasons,money or sex. What you just assumed was just jibber jabber madness.

These are just awesome, Brilliant research OP, I wonder how much time it must have took to create these address and how much pc power must have been used. Best thread I saw today!

It took almost no time at all, and very little pc power.  Creating a valid address is a very easy process.  What is difficult is finding a private key that results in an address that matches a particular pattern.

The vast majority of the addresses listed in your post are so unlikely that it is almost certain that nobody has the private key for them.  There is a small chance that someone might have generated such an address, but without proof, you'd be far more likely to be right if you just assume that the addresses were generated without a private key.

I see, so these were just created as show pieces. I wonder why though.. most of them has 20 k satoshi balance sent by "Twitter - DanMAbraham"

Very interesting research , great job !

This is original work and I think it is very useful, and entertaining.

Satoshi could run now with mass cash from his 1 million+ coins. why wait? Satoshi seemed passionate about this project so i doubt it was a money grab. At 1 million coins $400. whats the difference between $400 million and a few billion? Satoshi couldnt spend either in their lifetime. hasnt happened yet and i doubt it will. what could be the motivation to wait?

Thx guys, released the source for the collector:

http://www.reddit.com/r/Bitcoin/comments/2mkmoq/til_where_the_bitcoins_at_mysteries_of_the/cm5qecn

Before I go any further into this conversation,are you really asking me whats the difference between 400 million and a few billion?

He wouldn't be able to sell a million coins at the market price as the market for bitcoin is not liquid enough to handle a sale of a million coins like that.

@OP - Do you have more information about the one satoshi multisig addresses? I would be interested to know when exactly they were "funded"
Do you think they sold the tag to the address? Most of the "non-signed" blockchain.info tags that I have seen are those of "interest" to the community, like the "found" coins from gox or the stolen DPR coins

in usable terms arent both incredibly wealthy beyond comprehension? For me 10 million would be that point. i just cant see satoshi doing it. If he did, why not now already? what would be the reason to wait? it may sound like a fantasy but some people have life works they believe in that arent motivated by money or fame.

Excellent analysis OP, and quite an interesting read for the Bitcoin enthusiast.

I have a few questions, not just to you but anybody feel free to chime in on answers if you want.


How could this have happened? How could Bitcoin be transferred to an address that it can't decode when Bitcoin has a check-sum that does not allow you to send to addresses that don't exist (i.e. mistakes typing)? I guess what I'm asking- if it could "read" the address the first time to send an output to it, why couldn't it read it again?
I apologize if that sounds stupid; I love the technology in Bitcoin but am not the top expert on it.


Still learning about the generation of public & private keys, but why is it so unlikely that people don't have the private keys for these strange addresses? Is it because it would be relatively quick & possible to process through all possibilities of addresses to find one containing those letters/numbers there in order to be able to find a matching public key to send transactions to it?   But take exponentially longer to both find a matching (to selected letters/numbers) public key and a private key that matched that address?


I hope my questions make sense the way I wrote them out; if anybody is somewhat unclear about what I mean, let me know.

Here is the SQL dump of: SELECT * FROM outputs WHERE address = "** multisig **" AND balance = "0.00000001"

http://pastebin.com/6xh4XhHd

Hmmm you'd like to think it would be easy for anyone to find the private key to a given address... yet bitcoin is irrationally popular in spite of this?

Thx!


This thread explains some of them:

https://bitcointalk.org/index.php?topic=130392.0


The odds of generating one of those addresses from a known private key are 1 in 1,461,501,637,330,902,918,203,684,832,716,283,019,655,932,542,976.

The number is so large it is difficult for humans to comprehend.

To try and generate one of these from a known private key, each human currently living on the planet has to generate 500 million of addresses for each single nano-second (10⁻⁹s) during the entire age of the universe (15 billions of years).

Simply coming up with a valid public address that contains certain characters is much easier, on a cosmic scale...

Not quite true, there was a time when there were book orders on MtGox looking to buy more BTC that will ever exist; you could have sold 21M BTC and there would still have been more orders yet to be fulfilled without the price dropping below $0.20 or so. Instantly executing 1M BTC would not have even cut the price in half.

Interesting to read, I didn't know about all the binaries in the blockchain that the second post's link article describes.


Perhaps see vanitygen software, where it takes days to months to search for an address like in my sig, where just the first eight characters are specified.

Trying to find the private key for an address like where the whole address is user text, would be equivalent in difficulty to cracking and stealing bitcoins from any address. It would be as hard to find the private key for that address as to find the private key for the address that has Silk Road's seized 100,000 BTC.

Very interesting. It looks like the block that confirmed many of the transactions was around 230020 which is from 2012 (and AFAIK was before multisig was discussed seriously).

It would certainly cause the price to decrease significantly. Also your mention of "gox" also makes me somewhat question the validity of your statement (for example if someone tried to place a "market" order for 1 million btc on gox then a good portion of the order book could "disappear" or they would have some other issue)

No. Its a script that uploads/downloads arbitrary data to the blockchain by abusing multisig.
Keep looking "inside" the blockchain to find it.