SatoshiDICE.com - The World's Most Popular Bitcoin Game

[evoorhees]

Lost 18.6 btc last night lol

Then you're SURE to win the next roll!! 

[Stephen Gornick]

The hashing that decides the outcome is:
  hmac_sha512(hidden, trxhash)

The client determines the transaction hash so that is not random.

The first four bytes of the result of the hmac_sha512() are used instead of a random number generator, and will have values between 0 and 65,535.

Has the use of SHA in this manner been proven to result with a random number (i.e., results over time indistinguishable from a specifically purposed RNG)?  Is it possible that some transaction hashes will always produce a positive number, for instance, as is one type of attack described here?
 - http://crypto.stackexchange.com/questions/767/how-to-fairly-select-a-random-number-for-a-game-without-trusting-a-third-party#comment1303_770

Would a simple  hmac_sha512(sha256(hidden), sha256(trxhash)) or something like that make the outcome less vulnerable to an attack?

This isn't risk to players, but because this is wagering against the house, I wonder if there is some aspect of cryptography being misused here where an attacker could exploit this.  Which would mean that the pot doesn't grow, and the maximum bets which are still too low won't grow.

The thread probably most appropriate to further this discussion:
 - http://bitcointalk.org/index.php?topic=77206.0

[ccliu]

Lost 18.6 btc last night lol

Then you're SURE to win the next roll!! 

It was 1% chance of getting 6 losses I think in a row and it happened.

[runlinux]

ok, im hooked

[runlinux]

Sent 0.01 x 7 to 1dice9wcMu5hLF4g81u8nioL5mmSHTApw at 1:09 PM EDT, only 2 ever showed up on the site... I was hoping to win big!  

EDIT:

Heh, its just slow at times...

[ccliu]

Somehow the system manages to make me lose 8 times in a row, over and over

[evoorhees]

Somehow the system manages to make me lose 8 times in a row, over and over

Natural variance can be cruel to those at the tail

[runlinux]

Man, I'm done....

You rightfully won 0.23 btc. There goes my mining income for half a day...

Be back tomorrow!

[Stephen Gornick]

So that there is an independent copy of the hash.keys file, I put a copy up on github.
 - https://github.com/sgornick/satoshidice/blob/master/hash.keys
 - https://s3-eu-west-1.amazonaws.com/satoshidice/hash.keys

I did a sha1 sum on the file as well:
$ sha1sum hash.keys
d5f03cad9c3e0e2fac85fd0ba110d41fb31a20e2  hash.keys

[evoorhees]

So that there is an independent copy of the hash.keys file, I put a copy up on github.
 - https://github.com/sgornick/satoshidice/blob/master/hash.keys
 - https://s3-eu-west-1.amazonaws.com/satoshidice/hash.keys

I did a sha1 sum on the file as well:
$ sha1sum hash.keys
d5f03cad9c3e0e2fac85fd0ba110d41fb31a20e2  hash.keys

Thank you, good sir.

[Stephen Gornick]

Thank you, good sir.

You're welcome.

I see from spot-checking the Recent transactions that there is a wager on average between one and two per minute (i.e., the 200 row table holds six hours or less of wagers).  That's a run-rate per-day of about 800 wagers per day, with my guesstimate that the 1,000 wagers in a single day threshold has already been reached.

At current levels, that represents 10% of all transactions on the Bitcoin blockchain.

Let me repeat that.

At current levels, the two-week-old SatoshiDice.com is already responsible for 10% of all transactions occurring each day on the Bitcoin blockchain.

Since the site was a tiny fraction of this level about a week ago, if this growth rate continues, we'll probably learn quickly whether or not bitcoin will scale.  I know one thing will likely happen ... there'll be kvetching about how much disk space the block storage is taking up.

[mem]

Thank you, good sir.

At current levels, that represents 10% of all transactions on the Bitcoin blockchain.

Let me repeat that.

At current levels, the two-week-old SatoshiDice.com is already responsible for 10% of all transactions occurring each day on the Bitcoin blockchain.

Yikes, great effort with the site. very impressive.

[SkRRJyTC]

I played this game many times last night with great success!

However this morning I am still waiting on my first bet to show up.  Been waiting about 45 minutes now.  
Just wondering what is the longest response time anyone as seen from this site?

Edit The bet eventually came through... and I won too!

[SkRRJyTC]

I wanna roll the dice just one more time  

[testconpastas2]

same here ! too many transactions or admin run with my 0.01 bet

[Ashkelon]

However this morning I am still waiting on my first bet to show up.  Been waiting about 45 minutes now.  
Just wondering what is the longest response time anyone as seen from this site?

+1

EDIT I moaned and 10 minutes later, recieved sweet sweet btc
   however it came on an address i'd set up for bitconduit. What gives? Can someone explain how the payout addresses are determined?

[kokojie]

yeah seems the system is not working?

EDIT: bet was processed after a 10 minutes, thanks!

[Ashkelon]

Awesome, to the Max!

[evoorhees]

Yeah I think the server is stuck (just tried a test bet of my own)... workin on it.  Sorry for the delay!!

[evoorhees]

EDIT I moaned and 10 minutes later, recieved sweet sweet btc
   however it came on an address i'd set up for bitconduit. What gives? Can someone explain how the payout addresses are determined?

It pays to the address from which it receives the bet. For this reason, some ewallets are not compatible, because they don't send funds from the user's receiving address. If you're not sure if your specific wallet is compatible, just try with .001 as a test.