Just-Dice.com : Invest in 1% House Edge Dice Game

[trout]

  We currently can't match SDICE's 500 BTC max bet on 2x, so I'd say we're not too big yet.  Mind you, they apparently risk a little over 100% of their bankroll per dice roll, so it's easier for them to offer big bets.  It seems to be the "let's hope nobody gets lucky" technique of bankroll manangement.

probably meant "a little over 1%."
The optimal for SD to risk on each bet is 1.9% - their house edge.
Probably they know about it and do exactly that.

[wolverine.ks]

What about splitting the cold wallet up into smaller amounts and giving them to other established members of the bitcoin community? I trust that dooglus is up to his eyes in backups, but this would also remove some of the incentive to take the money and run. The passwords for those accounts could use samir secret sharing so dooglus wouldnt have to trust these other people as well.

[Bugpowder]

  We currently can't match SDICE's 500 BTC max bet on 2x, so I'd say we're not too big yet.  Mind you, they apparently risk a little over 100% of their bankroll per dice roll, so it's easier for them to offer big bets.  It seems to be the "let's hope nobody gets lucky" technique of bankroll manangement.

probably meant "a little over 1%."
The optimal for SD to risk on each bet is 1.9% - their house edge.
Probably they know about it and do exactly that.

Doubly wrong. RTFF.

[Deprived]

What about splitting the cold wallet up into smaller amounts and giving them to other established members of the bitcoin community? I trust that dooglus is up to his eyes in backups, but this would also remove some of the incentive to take the money and run. The passwords for those accounts could use samir secret sharing so dooglus wouldnt have to trust these other people as well.

How can you possibly have it so dooglus doesn't have to trust them? 

Either has access to the accounts - in which case he doesn't have to trust them but all we've done is added new risk.

OR

He doesn't have access to the accounts - in which case he DOES have to trust them.

In general this kind of approach ends up adding risk not removing it.  And, more importantly, it makes it hard for investors to assess risk - as we can't tell in advance who our capital is going to be exposed to.

[drrussellshane]

What about splitting the cold wallet up into smaller amounts and giving them to other established members of the bitcoin community? I trust that dooglus is up to his eyes in backups, but this would also remove some of the incentive to take the money and run. The passwords for those accounts could use samir secret sharing so dooglus wouldnt have to trust these other people as well.

How can you possibly have it so dooglus doesn't have to trust them? 

Either has access to the accounts - in which case he doesn't have to trust them but all we've done is added new risk.

OR

He doesn't have access to the accounts - in which case he DOES have to trust them.

In general this kind of approach ends up adding risk not removing it.  And, more importantly, it makes it hard for investors to assess risk - as we can't tell in advance who our capital is going to be exposed to.

Exactly this.

[wolverine.ks]

it would add some risk but remove other risk at the same time.

there is an incentive for dooglus to take the money and run, everyone knows this. but if he uses shamir secret sharing, then several people across the community would all have to be in on it in order to do that. and even then, they would then likely split the money, reducing the incentive to run even further.

http://en.wikipedia.org/wiki/Shamir%27s_Secret_Sharing

the names/aliases of the secret sharers could be made public so the investors would know who has a portion of the bankroll at any given time, and the addresses would be made public so they could be monitored at all times as well.

yes, if all the people died at the same time, or all the people lost the passwords at the same time, then you lose the money. but that is very unlikely, and there would be several accounts, so you wouldnt lose all the money at the same time even if it did.

additionally, since there would be several accounts, then it is more likely that in the event that a cold wallet needs to be accessed, at least one of the wallets would have enough people available at any time to access it.

[Deprived]

it would add some risk but remove other risk at the same time.

there is an incentive for dooglus to take the money and run, everyone knows this. but if he uses shamir secret sharing, then several people across the community would all have to be in on it in order to do that. and even then, they would then likely split the money, reducing the incentive to run even further.

http://en.wikipedia.org/wiki/Shamir%27s_Secret_Sharing

the names/aliases of the secret sharers could be made public so the investors would know who has a portion of the bankroll at any given time, and the addresses would be made public so they could be monitored at all times as well.

yes, if all the people died at the same time, or all the people lost the passwords at the same time, then you lose the money. but that is very unlikely, and there would be several accounts, so you wouldnt lose all the money at the same time even if it did.

additionally, since there would be several accounts, then it is more likely that in the event that a cold wallet needs to be accessed, at least one of the wallets would have enough people available at any time to access it.

There's problems with that sort of approach.

Assume there's X keys and Y are needed to unlock.

As Y gets closer to X it becomes increasingly difficult to get enough people to gain access when funds need moving.
As Y becomes further away from X the likelihood of theft increases.

It's additionally made risky because, depending on implementation, it could be impossible to determine who had conducted a theft - i.e. you'd know at least Y of the X holders were involved but not which ones (as any Y of X could unlock).  Any solution which involves the possibility of anonymous theft MUST be discounted - as it removes one of the (maybe THE) prime disincentives to stealing in the first place.

Where you run into a problem with that is that in order to remove anonymity, all signings/signatures have to be public - which then begins to add an entirely different set of problems/risks.  For a one-time use scenario it could be fine - it doesn't work so well when withdrawals could occur frequently.

Now imagine that somehow it's all set up and working.  Then someone posts a message saying "I'm a key-holder - please contact XXX@YYY.COM if you're also one so we can cooperate."

What do you do?  Is it real?  Is it a hoax?  Is it a prank?  Has that increased your perception of risk?  What if they then post saying "we only need 1 more keyholder to get rich?"  Would you feel more comfortable?  Or would you feel as though you were in far more danger than if dooglus had just looked after it all in the first place?  I mean you'd know that logically it was probably a hoax - but would you not be tempted to withdraw funds just in case?  Theoretically it could be a keyholder who was NOT intending to steal - but was either (stupidly) trying to place a trap for dishonest ones OR was also an investor and was trying to drive other investors out to increase their profits.  There's a whole raft of reason why people would WANT to spread FUD - and the nature of things is that they'd be likely to succeed because of the way in which such a scheme HAS to work.

Honestly it's a nightmare - both in terms of logistics and in terms of actually delivering improved security and confidence.

[wolverine.ks]

why would one secret sharer not know who the other secret sharers are? is there a reason to keep then in the dark about each others identities? if the identities were secret then what stops dooglus from being all the identities or not sharing the secret in the first place? if all identities were know, PGP signatures prevents hoaxes, right?

[usagi]

What about splitting the cold wallet up into smaller amounts and giving them to other established members of the bitcoin community? I trust that dooglus is up to his eyes in backups, but this would also remove some of the incentive to take the money and run. The passwords for those accounts could use samir secret sharing so dooglus wouldnt have to trust these other people as well.

There are extreme risks in this.

I hasten to remind everyone that this is precisely what destroyed CPA. We split up our funds and gave them to five established members of the community. Ultimately we got less than 40% of our money back. Remember, these were the best people in the community at the time. What happened later only proved the original assumption to be a mistake.

The other lesson learned was about chaining counterparty risk. Like Hashking -- many people invested with HashKing simply because he swore up and down he wasn't in pirate. How can dooglus stop people from investing back into just-dice? Counterparty risk magnifies your risk. If dooglus gave me 5,000 BTC to hold on to, and I invested it into just-dice, a situation where he needs his money back is precisely the situation where I cannot repay it.

I was thinking about this days ago, approaching Dooglus with such an offer. But given the choice I think I might prefer not to take his money and retain the ability to invest into just-dice.

[wolverine.ks]

how would you invest it in anything if the address is known and the key is a shared secret?

[Deprived]

why would one secret sharer not know who the other secret sharers are? is there a reason to keep then in the dark about each others identities? if the identities were secret then what stops dooglus from being all the identities or not sharing the secret in the first place? if all identities were know, PGP signatures prevents hoaxes, right?

I'm not at all clear what it is you think secret-sharers would be signing/decoding.  There's no function that I'm aware of in Bitcoin for a wallet to have n of m signing (transactions can - but that's not relevant).  So what is it you think they control?  The whole process becomes very unwieldy rapidly with new secrets having to be generated after EVERY movement of funds (because there's no way to apply an ongoing n of m signature requirement to a wallet).

And there's practical difficulties unrealted to technical aspects :

Who are the trustees?  If you looked back a year or so and picked the most trusted individuals you'd fund a lot have now scammed/defaulted.

How do they determine a withdrawal request from dooglus is legitimate?  Do they have contact details for all account holders?  Do they have to publish a request for anyone else claiming the amount and wait a few days before allowing large withdrawals (in case it's a dooglus sock-puppet pretending to own someone else's large deposit)?  After how much of the BR has been paid out do they stop authorising withdrawals without a full audit?  

If the house takes large losses does J-D have to close down for a while whilst the trustees collectively check it isn't dooglus manipulating results to syphon off cash?

The moment you appoint people specifically to prevent him stealing they MUST then always assume he's acting in bad faith until proof of good faith is provided - or they become toothless watchdogs and he can still steal most of the funds by jumping through a few hoops making the whole exercise pointless other than for PR.

Who pays for their time and effort - it in no way benefits dooglus (if he intends to steal obviously it doesn't, if he doesn't intend to steal then he's accepting a non-zero risk for no benefit).

In theory dooglus COULD act to satisfy your concerns without risk to him.  Way to do it would be simply to allow maximum risk to be up to 100%.  Then you could all micro-manage your exposure via API or via some third-party you decided to trust.

So if you accepted that with trustees dooglus should only have immediate access to 10% of funds you could approximate that by only depositing 10% of your own funds with a 10% risk on them - you'd then be risking and gaining the same as if he held all your funds at 1% (assuming all bets were spread evenly across all risked capital).

And if you didn't want to micro-manage it yourself you could indulge your own personal level of paranoia to any extent you chose right through to:

Forming an investment group with other like-minded individuals
Hiring a lawyer who employed staff on your behalf to manage the funds
Instructing them only to send topups if N of M of your group provided crytographic proof along with a photo of them holding the current day's newspaper plus their passport and with a shoe on their head.

Then he could only steal 10% of your money.  But no doubt the complaints would still continue that he could steal other people's and that those not bothering with these precautions were making more profit.  And there's the root of it - I believe the current risk (and the reduction of risk if a solution was put in place) doesn't justify the cost (and the new risk).  You and some others apparently do.  So come up with a proposal where you bear the cost and get the benefits you believe exist (and the risks I believe exist) and we have something to discuss.

An alternative to the above would be for dooglus to allow accounts that had a notional balance higher than their actual balance - with the account-holder having to top up before they got more action if their actual sent funds were exhausted.  In practice that works very much like allowing increased risk - it just accounts for it differently.  It does fulfil the same (to me) major objective - that the effort and cost of maintaining partial exposure is placed firmly on those demanding it.

[Deprived]

I hasten to remind everyone that this is precisely what destroyed CPA. We split up our funds and gave them to five established members of the community. Ultimately we got less than 40% of our money back. Remember, these were the best people in the community at the time. What happened later only proved the original assumption to be a mistake.

I was actually thinking of that when I made my own post at the same time.  The people who stole CPA's funds are exactly the sort of people who would likely have been chosen as trustees back then.

And it turned out they were all loaning one another money and cooperating/colluding in covering up their mutual exposure to Pirate.  You can pretty safely assume that if it were possible without being detected they'd have sent J-Ds money to Pirate as well.

[dooglus]

probably meant "a little over 1%."
The optimal for SD to risk on each bet is 1.9% - their house edge.
Probably they know about it and do exactly that.

You would think so wouldn't you?

But they offer a 64000x payout with a max bet of 0.1 BTC.  That's a 6400 BTC payout.  Erik claims that the SDICE betting pool is made up entirely of a loan of "about 6100 BTC" from himself.  That means they're risking over 100% of the pool.

[dooglus]

But given the choice I think I might prefer not to take his money and retain the ability to invest into just-dice.

That's a strange way of putting it.  I don't see it as "my money" and it strike me as odd that you do.

[Deprived]

probably meant "a little over 1%."
The optimal for SD to risk on each bet is 1.9% - their house edge.
Probably they know about it and do exactly that.

You would think so wouldn't you?

But they offer a 64000x payout with a max bet of 0.1 BTC.  That's a 6400 BTC payout.  Erik claims that the SDICE betting pool is made up entirely of a loan of "about 6100 BTC" from himself.  That means they're risking over 100% of the pool.

Has he actually ever said that 6100 BTC was ALL of the pool?  I'd assumed (maybe incorrectly) that was only part of it based on my recollection of some long-distant post in which he discussed their BR management.

[usagi]

I'm not at all clear what it is you think secret-sharers would be signing/decoding.  There's no function that I'm aware of in Bitcoin for a wallet to have n of m signing (transactions can - but that's not relevant).

I believe Armory can do this and at least one other client I can't at this time recall.

[usagi]

But given the choice I think I might prefer not to take his money and retain the ability to invest into just-dice.

That's a strange way of putting it.  I don't see it as "my money" and it strike me as odd that you do.

Point well taken. I'm just saying that the money is under your control. You are in essence a money manager by might or right, and I would expect you to manage it prudently, and to invest the money you have been given in limited ways. I scratch my earlier comment about real-world casinos invetsting their money though, it may in fact be a better deal to just keep the money locked down and increase the max bet  

But if we are discussing splitting up the money and giving it out to people, we should also determine the risk, and appropriate interest rate.

[Deprived]

But given the choice I think I might prefer not to take his money and retain the ability to invest into just-dice.

That's a strange way of putting it.  I don't see it as "my money" and it strike me as odd that you do.

Point well taken. I'm just pointing out that the money is under your control. You are in essence a money manager by might or right, and it is certainly within your scope to invest the money you have been given in limited ways. I assure you there isn't a casino in the world that doesn't -- you're not gaining anything by not doing so. But I understand that there aren't a lot of places to put 30k BTC these days.

He most definitely IS gaining something by not investing it - a lack of CP risk and a lack of risk of loss through bad investments.  Perhaps more importantly he's gaining the trust of investors that the cash isn't going to vanish due to some cause allegedly (or actually) beyond his control.

Even depositing that cash in some exchange or online wallet is a risk that, as an investor, I wouldn't want to accept without some extra financial benefit.

[Dabs]

Would escrow work? There are a few people on these forums who hold large sums of coins. And there is at least one who is offering to hold > 10k to > 100k. The problem again, is trust or incentive to not disappear, or whatever, and then there is the possible drama from all the accusations of fraud, false or not.

I mean, it is dooglus own site, the investors decided to put their money there. That is their decision. If they want to make dooglus have someone else hold their coins, they merely remove financial responsibility from one person to another.

[Bugpowder]

it would add some risk but remove other risk at the same time.

there is an incentive for dooglus to take the money and run, everyone knows this. but if he uses shamir secret sharing, then several people across the community would all have to be in on it in order to do that. and even then, they would then likely split the money, reducing the incentive to run even further.

http://en.wikipedia.org/wiki/Shamir%27s_Secret_Sharing

the names/aliases of the secret sharers could be made public so the investors would know who has a portion of the bankroll at any given time, and the addresses would be made public so they could be monitored at all times as well.

yes, if all the people died at the same time, or all the people lost the passwords at the same time, then you lose the money. but that is very unlikely, and there would be several accounts, so you wouldnt lose all the money at the same time even if it did.

additionally, since there would be several accounts, then it is more likely that in the event that a cold wallet needs to be accessed, at least one of the wallets would have enough people available at any time to access it.

There's problems with that sort of approach.

Assume there's X keys and Y are needed to unlock.

As Y gets closer to X it becomes increasingly difficult to get enough people to gain access when funds need moving.
As Y becomes further away from X the likelihood of theft increases.

It's additionally made risky because, depending on implementation, it could be impossible to determine who had conducted a theft - i.e. you'd know at least Y of the X holders were involved but not which ones (as any Y of X could unlock).  Any solution which involves the possibility of anonymous theft MUST be discounted - as it removes one of the (maybe THE) prime disincentives to stealing in the first place.

Where you run into a problem with that is that in order to remove anonymity, all signings/signatures have to be public - which then begins to add an entirely different set of problems/risks.  For a one-time use scenario it could be fine - it doesn't work so well when withdrawals could occur frequently.

Now imagine that somehow it's all set up and working.  Then someone posts a message saying "I'm a key-holder - please contact XXX@YYY.COM if you're also one so we can cooperate."

What do you do?  Is it real?  Is it a hoax?  Is it a prank?  Has that increased your perception of risk?  What if they then post saying "we only need 1 more keyholder to get rich?"  Would you feel more comfortable?  Or would you feel as though you were in far more danger than if dooglus had just looked after it all in the first place?  I mean you'd know that logically it was probably a hoax - but would you not be tempted to withdraw funds just in case?  Theoretically it could be a keyholder who was NOT intending to steal - but was either (stupidly) trying to place a trap for dishonest ones OR was also an investor and was trying to drive other investors out to increase their profits.  There's a whole raft of reason why people would WANT to spread FUD - and the nature of things is that they'd be likely to succeed because of the way in which such a scheme HAS to work.

Honestly it's a nightmare - both in terms of logistics and in terms of actually delivering improved security and confidence.

I think Mark Twain had some insight on this very topic...

http://www.eastoftheweb.com/short-stories/UBooks/CorrHadl.shtml