I can't think of any way that it would be possible for anyone to prove either way whether the profits are real or fake.  The owners of the site can place winning or losing bets of their own at any time to skew the profit/loss graph as they wish, and any such bets are indistinguishable from 'real' bets.

So yeah, all we can say is "it's weird".

More importantly, how do your odds compare with those at https://bitzino.com/ ?

Are you able to compete with their quite attractive set of rules?

I'm thinking some people only bought bitcoins to invest with Pirate.  So if they get their coins back they'll likely want their "real money" (as they see it) back.

This looks like the major contributor to the most recent SD profits - another martingale disaster:

A more honest way of speaking the same truth would be to say that us humans are very good at spotting patterns.  Even in random sequences.

There are only a few which have been stuck for more than 24 hours:

c9cca3a9e5a1494cee2c87c950c1283f83d091920cb6e6749b3e5930d8d69ac2   1 187983 09:29:03 Jul-07-12         0.01000000 1dice1e6 1VayNert3x1KzbpzMGt2qdqrAThiRovi8
6071357b4e24964f398de274e55c85cadaf33e1f54acaa284a7dfc23933feff3   0 191269 17:47:08 Jul-28-12         1.00000000 1dicegEA 14q94SKrm8W3yBzhi2BZ8xm8VJvPWcXAMG
18455d8b2020cef49579fa2f024375a43e242f4bb0c303cba526327944c65e46   0 191269 17:47:08 Jul-28-12         1.00000000 1dicegEA 14q94SKrm8W3yBzhi2BZ8xm8VJvPWcXAMG
accf0d01f0ee50ff12fc661345ea1c2608fc497aa3a7baf004f70f8210fae654   1 192798 17:28:18 Aug-07-12         0.40000000 1dicegEA 1DvHQKGENEjqKBj86JdhTjr5iK18hVr9HF
e636f45b62073dab7d1ec1533a02a707e1ec09064d38a583b567f68fa5077498   2 193937 17:44:47 Aug-14-12         0.10000000 1dice6Yg 1PDo5MEZ4uFcB6pCN21VXq4mQpkXnjGnUD

Notice the VayNert one has been stuck for over a month!

I don't.  I think that if it's a scam then pirate will have been selling excess bitcoins all along rather than dumping them all at the end.  His lenders didn't have that option.

Yeah, the timestamp has nothing to do with the magic number.  All that is used to make the magic number is the daily secret (which changes only once per day) and the transaction id of your bet (which doesn't contain a timestamp).  So it's just coincidence.

Although it is true that your lucky number will keep going up until it goes down.  Unless it stays the same.

It's a bug.  It was introduced accidentally when code was added to avoid sub-cent change from being created, since any sub-cent output automatically incurs a transaction fee.  The code makes sure the change is always at least 1 bitcent if possible, but forgot about the case where no change is necessary.  It works if the entire payment can be made using a single output, but not if more than one output is required.

It's a bug in the satoshi client.

In an attempt to avoid sending change less than 0.01 BTC, it sends change even when it isn't necessary.

I submitted a fix for the bug a long time ago, but it never got included in the master branch.  See https://github.com/bitcoin/bitcoin/pull/898

Here's a good example.  MtGox is trying to consolidate 10 50k outputs into a single 500k output, but an 11th 50k gets pulled in and spat out as change due to the same bug:

http://blockchain.info/tx-index/2048819/1

A nice 2k BTC profit for SD in the last 24 hours.

Looks like most of it came from a single address.  Here are the biggest losers over the period:

<32000 31c4f5db22f2a432ea3637b1beb2536e78965d47806452848ff6da6bd1877909 1K7LiKpARZFGZwbHhL6GENTw8D9w7LfFug BET  53 LOSE
<32000 a9aad57cf63995cc37bd2b4390d9955a77e61795c65b77df4b98297e1ab757cc 1K7LiKpARZFGZwbHhL6GENTw8D9w7LfFug BET 134 LOSE
<32000 470e362ad4e3675d817e5e15486d30a369a653da8639d9a7e7f3b0fc1f96d142 1NZfuJLeq7jktGB1asESXTZqh8MBEE6hHb BET 151 LOSE
<32000 4883910410301a11bf9e069744f89821b5a10a4431f17d0b84896169311d8636 14wQQyNTFhyV1hauFMFXpeWhQ6PRFC2LzM BET 183 LOSE
<32000 d3ba562936a8e9231a2d35e536d0822ee034caf6e817503169ee79923a50e1b4 1Gx6kKam9Zx9h3zr1zs2Zn1exf44D5shxh BET 223 LOSE
<32000 c20ac7dab6727a0309099ff1d5e9d72d2a8d0d26a9773af469bf2d2fb0b51d70 1K7LiKpARZFGZwbHhL6GENTw8D9w7LfFug BET 234 LOSE
<32000 b6a1718fa559b7697eefd5b9e1df387a97608433770b9052c6f8954bd1621a7e 1FQdAgchLJcQbVbdKVGy2ALhySWya5ycp7 BET 249 LOSE
<32000 45f51a5388da94c53409ccf56e9f7618cd5807e0feedd7d2d874c526836b26c6 1K7LiKpARZFGZwbHhL6GENTw8D9w7LfFug BET 250 LOSE
<32000 d56ddbfa757a1e0e01a04669074dfd8d33bca7ed5d7922eca61cc9cb9df4dabf 1K7LiKpARZFGZwbHhL6GENTw8D9w7LfFug BET 250 LOSE

They always pay a fee when returning your winnings to you (win or lose, they return something to you, and always pay a fee).

I don't understand your "where is the limit?" question, but maybe this answers it:

http://www.satoshidice.com/bits.php says:


i.e. you pay only the first 0.0005 BTC of the fee, and they pay the rest (if any).

Seems to me this isn't really even money (in dollars), in that if Pirate pays out, there will be a flood of people selling their proceeds, pushing the price of BTC down.  So dust either wins a lot or loses a little?

bet = 0.01
multiplier = 1.004
winnings = 1.004 * 0.01
fee = 0.0005
payout = 1.004 * 0.01 - 0.0005 = 0.00954

You got a little more than you should.  Probably the true multipler is a little more than 1.004.

There's nothing in the protocol to stop you sending two outputs to the same address.  What you say about only one of the two being spendable is completely wrong, since when you spend an output you refer to the transaction ID and the output number.  That uniquely identifies the output you're spending.

The satoshi client doesn't allow you to send 2 outputs to the same address, but presumably other clients do, since we see such bets all the time being sent to satoshidice.

You can't bet "above 50%" at satoshidice.  All bets are "lessthan X".

Here's an example transaction:
http://blockexplorer.com/t/3RmgrQ7WPK

Notice how all 4 outputs have since been spent.

And here it is at satoshidice:
http://www.satoshidice.com/lookup.php?tx=400bea64815ce2035310f7caebdf5c3ce0ddd36bff11aece85ab20baa8b47496

Only 2 of the 3 bets were accepted, since the 3rd would have put it over the max bet allowed (0.05 BTC).  So there's (now) no benefit to sending multiple bets to the same address.  It used to be a way to avoid the maximum bet limit, but that has since been fixed, as shown in the previous link.

No.  It was made very clear that as soon as your coins are received by payb.tc they are earning interest, whether he actually passed them through or not.  Since they're earning interest, you can't expect to get them back in the event of a default.  payb.tc was reducing his own investment by however much you sent until he transferred your coins to BS&T to make up the difference.

You can't go changing the rules now.

Is that what Clipse was doing all along to make his pool pay?

After all his rudeness on his thread about how we just didn't understand his business plan he was investing his miners' funds in a probable ponzi?