I independently came up with it, but I'm sure it's been proposed many times before. I know a lot of people have worked on coin selection ideas that make change amounts difficult to distinguish from the sending amounts, but in practice it's very difficult to do that well. On the other hand, having multiple change addresses makes it easy to have them completely ambiguous with the actual withdrawal, at the expense of making the transaction a little bigger and occasionally spend-linking a few extra outputs together.

(The other disadvantage being that transactions with 3 or more outputs are relatively uncommon, so I can't use them too much or it will be characteristic of a MP withdrawal)

It started off when I was playing around with blockseer on MP and some other sites, and it was a bit surreal to see that in what seemed like ~99% of the cases you could build up a pretty accurate picture of the hot wallet, purely by tracking transactions and deducing which was the change and not. I'm pretty sure that with enough effort, it would be quite easy for someone (e.g. coinbase or circle?) to build quite an accurate idea of what is a deposit and withdrawal (but not really link them together).

As another benefit, the code now monitors the size of the hot wallet and actually automatically spills it over into cold storage (using bip32 addresses) which avoids the other potentially privacy damaging sweep operation =)

The way the code works is that it generates 1 of 10000 discrete outcomes and lets you pick if you think the result will be [0, 5000) or [5000, 10000). The provably fair shows that they have committed to that result and can't change it.

Even if you assume nothing about the distribution of the result, it is easy to show that if you randomize the pick of low or high range you have a 50% chance of getting it right. But furthermore, there's compelling reason to believe that the outcome is indistinguishable from random as well (otherwise it would be used to players advantages, for one)

Yes, that's why they normally use more than one deck. Otherwise if you got a 9 of hearts, you know the chance of another card also being a 9 was 25% reduced. Actually under a lot of common rule variations if casinos only used a single deck (even reshuffling on each hand) it would be possible to play +EV. But with multiple decks, the information gain you get from seeing each card becomes absolutely tiny.

As a new privacy feature, some (but not all) of your withdrawals will now have multiple change addresses (which may be from the cold, hot or warm wallet).

Here's an example withdrawal: https://blockchain.info/tx/93edee61b741dca549f4fad518caa7a4728b9e3f73b055756feb977fd8d7c7fc

This makes it even more of a PITA for anyone trying to do blockchain analysis on MP withdrawals (as it makes it almost impossible to reliably tell change addresses from withdrawal addresses).  And remember to generate unique addresses for each transfer you receive, and when you need that extra level of privacy use our awesome cold-addresses ( https://blog.moneypot.com/cold-addresses/ ) for deposits.

Spamming really isn't the best way to make friends

Soon there will be an elegant native solution to this after bip 65 gets deployed which will enable you to send money to yourself in such a way that's not spendable until a certain time in the future. Right now the best you can probably do is create (and store) a transaction to yourself that is nLockTime'd  and throw away the original private key. Unless you're an experienced programmer, I'd strongly recommend against doing that though.


Anyway, if I was in your situation I'd probably just create a 2-out-of-3 paper wallet, send all your bitcoins to it, and give 2 out of the 3 pieces of paper to some friend or family you trust, and ask them to not return until some condition has passed.


Also, I'd advise someone with a gambling problem against using the locked investment in MP. If you have a gambling problem you should consider locking away all your funds, and MP (or any site like it) is far too risky to put all your money. And secondly the actual investment is volatile enough that it almost constitutes gambling itself (investor profits literally went from 0 to 50 to -10 and back to 47, all in a few months).

Well this was quite the anti-climax:



Great competition though, best of luck to everyone else

Back to normal!

There's some serious chat-server issues at the moment, due to a widespread service-provider partial-outage. This will likely continue for a few more hours, but the actual games, deposits and withdrawals should be unaffected.

Site's been a bit slow with no huge winners in the last couple days, but at least this went to a pretty cool guy:

You can check with:


So yeah, 10.69 million is the highest it's been so far (almost a year ago)

Seems like a good idea, although:



Perhaps instead of a point being worth 0.001 BTC, if you made their value increase as you accumulated more points it would give players incentives to accumulate instead of cashing them out at the minimum required? So e.g. 50 points might only give you 0.025 BTC, but 2000 points might give you 4 BTC

Just a tip, with a name like "LuckyBit.xyz" you're damning yourself for two reasons:

Firstly is that when ever the name "LuckyBit" is mentioned, people will think luckyb.it and not you (and also they will rank above you in google, for people who search it). And second, running a gambling site is all about building trust, and for good reason no one is going to trust someone who rips off a name (and assets) from another site.

Most (all?) of the sites you list have a provably fair system, which without a doubt disproves the idea that they are cheating you. And the truth is that they don't need to cheat you in order to make money, as they all openly have a house edge.

From the sounds of it, I think you may have a gambling problem which is something serious you should address.



It doesn't help to blame the casinos, because you'll find they are all the same: they're in the business of making money from you. And they don't need to cheat you in order to do this.


That got dark fast.

This is a rather perplexing position to take? A deposit address is quite simply: "Send bitcoins here, so I get them and know it was from you".

One particularly frustrating extension of what you said what a few other altcoin's do. They support bitcoin addresses as well, and users (or developers?) will send $altcoin to one of my deposit addresses. Then they'll go on to demand that I need to credit $altcoin to their account (which I don't support) or send the $altcoin back, and if I don't I'm a scammer. I've never yet sent one of these people their $altcoin back, as doing so would risk my entire HD wallet if a single private key is leaked by incorrect or malicious code, and in general it's just a waste of time.

Or another good example of it doesn't make sense, is this recently happened over at my site:

A user was intending to make a deposit, but still had in their clipboard another users deposit address (to send a tip?) and mistaken sent the funds to an other users deposit address. The facts or timeline wasn't disputed, but the user who was getting the incoming payment believed they were entitled to the money because it went to "their" deposit address.

But I instead manually credited the account of the user who sent the money, rather than the user who's deposit address was mistakenly used. You'll have a really, really hard time convincing me I did the wrong thing or the receiving user has a legal expectation to get that money.

And Dexon has risen to the challenge:


In browser provably fair checker: http://jsfiddle.net/1L1uqcgv/6

Some people have wanted to generate the previous games in their browser to verify the games, so I made this quite quickly:

http://jsfiddle.net/1L1uqcgv/

Note that it logs to the browser console, so if anyone wants to polish up that jsfiddle and a little UI to it it'd be quite nice =)

Well, that's completely unverifiable now.

--

I guess the conclusion of this thread is that the server seed that unibtc was playing with was lost, which means it was either done intentionally to hide the fact some of his rolls were rigged to win back the money ... or done as a silly mistake.


I'm going to assume it was the later, as I know I've made more than my fair share of mistakes and would like people to assume the best.

The issue isn't you changed the server seed, that's not a problem at all. But if you do change the server seed you should do three things:

a) Show the previous server seed  (so past bets can be verified)
b) Prove you've committed to a server seed (i.e. give the hash of it)
c) Allow the user to pick their own client seed.


So from the sounds of it, the lack of a) is the biggest issue here. Is there a way for unibtc to verify his previous rolls?  But changing a server seed while keeping the users client seed the same is another big mistake, but something separate and mitigated by his unpredictable playing.


Well, pretty much all lucky winners look like they're cheating. You might as well assume they the best, as if they're legitimate players you risk driving them away. And if they're cheating, they're really not going to care if a mod is being rude.



Completely understandable. It sounds like that was the right thing to do, but steps a) b) and c) just need to be done as well

From the posts, it sounds like he ended his session in a bit of a losing spree? Is the server seed available from this period of time, as if so it should allow the verification of the whole session and put the matter to rest