I'm actually extremely confident that the win was 100% legitimate. Thanks to the unique provably fair system, I can independently verify the games, and get the result:


Where realized refers to players that have revealed their server seed (if they don't, there's no way for me to verify the games). But users are required to reveal their server seeds to withdraw/tip or invest.

I've also run numerical simulations, and am extremely confident the house edge is between 1/101 and 1/99. [Edit: Just got a private message from blockage, who said he's analytically verified the house edge is exactly 1%]

So I think that gives us these possibilities:

a) The player has independently hacked both me and bustadice, and faked the wins
b) Daniel has hacked me, faked being a player and the wins
c) I have hacked Daniel, faked being a player and the wins
d) The wins are totally legitimate


I'm going to say a) and b) are pretty unlikely; especially considering I have 300+ bitcoin sitting on my computer that no one touched. I know I didn't do c) -- so I'm very confident that the wins are very legitimate  

As I understand it, it's intentionally designed to be somewhat of a black box and depends on things like your ip address. The high-level purpose of the faucet is really just to give people a pain-free way of trying out the game and mechanics, to know if they want to gamble for real 

I actually implemented a "mining faucet" for bustabitv2 where you explicitly turn on a miner and then can collect the value of what you mine into your BaB account. However, the economics really suck. When I sandboxed the script, on my macbook pro it was only able to make about ~3 satoshis per hour. Unsandboxed and using all my cores, it would only get as high as 7 satoshis per hour. Considering the min bet on bustabit is 100 satoshis, I can't imagine many people are going to want to mine for 12+ hours to make a single bet.

I can't imagine it really makes any sense unless you've got millions of visitors on your site

from the owner of bustabit, there's no affiliation with moneypot here

But thanks for the mention!

Just 3 hours to go, and the bankroll is looking pretty healthy!

So at launch it looks like the max-profit is going to be >21.645 bitcoin ($93k USD at current prices). As far as I can tell, this is the largest max-profit[1] of any bitcoin casino at the moment. Or anyone aware of something bigger? Hopefully the whales take note 


[1] Sites like bustabit have a bigger max profit, but they constrain the max-bet. So I'm looking at it from the perspective someone trying to do something like a 2x bet

Weird, where are you located? There's some countries I believe that block bustabit on the DNS level, or could be some intermittent problem. But I don't believe there's been any widespread issue 

I'm going to stop taking over Devan's thread, but one last thing:



One future extension I'd like to make, is introducing *another* party: the arbitrator. Currently the game-server and audit-server communicate through a (validated-on-both-sides) tls connection. So instead of communicating directly, you could proxy the connection through an arbitrator (who is running on the same datacenter to avoid latency) and acts as a dumb proxy with no means to decrypt or modify the connection. However, the arbitrator would run tcpdump on the traffic for logs.

Now if the audit-server and the game-server ever got into a dispute (either because 1 party was malicious, or one party was undetectably hacked and had wrong data). The tls keys could be revealed, and the arbitrator sees who was right. This would work especially well if the arbitrator was the 3rd (2-of-3 cold storage key holder), so in a hypothetical dispute would know who to side with.

But this is a bit overkill for now 

The neutral third party is quite well known in the bitcoin community, but has asked to not be named to avoid their participation looking like an endorsement. In the case of any dispute or problems between me and Daniel, the third key holder will not arbitrate nor in the general case has enough information to sanely do so. So their primary role is simply help out in the case of Daniel or me has an issue that would prevent us signing (e.g. gets hit by a bus). You can pretty much think of it was 2-of-2 cold-storage, with a backup in case someone dies.  (This could have also been achieved using time-locked transactions, but it's incredibly tedious)

Edit: In the interest of transparency, we are still working on testing/verifying the cold-wallet solution. So there is currently no money held in it, and we've told interested investors to hold off for a day while we make sure everyone is comfortable with it.


 
You are charged the dilution fee on the difference of your (onsite+offsite). So if you add 1 bitcoin to the onsite, and subtract the same from the offsite (at the same time) your net position is the same and you are not charged any fees.  I think the whole UI needs a bit of improvement, but it should be more obvious when it tells you the expected dilution fee, so you can easily see how it works.



Yup! Nice job on that. It looks like pretty much the exact same thing. The only difference I see is that instead of a weekly audit, we do it in real-time. We require users to reveal their hashes before using their money


I am just confused to not see the audit hash anywhere and not mentioned in the verifier either.


Yeah, but it's abstracted away. The "Server seed" is actually the concatenation of the two seeds. But the user just needs to copy and paste the whole thing into the verifier script. So from their perspective it's exactly like verifying any other game. We'll come up with a more technical specification, but this is the high level overview:
https://gist.github.com/RHavar/285c295f7906e03369cd66580a1b5f45

But you can see it in the verifier:
view-source:https://bustadice.github.io/verifier/

Thanks! Daniel and I spent a bit of timing going back and forth about how the commission structure should work, but I think it's a pretty reasonable compromise. Only charging a commission on profits is definitely better for investors (even though over a long period of a time it's the same) but I think it makes sense to offload as much variance to investors as possible: after all, that's what they're +EV (and +Expected bankroll growth) for!

That said, investing is inherently very risky and the commission structure does exacerbate that, so it is pretty sane to risk less than you otherwise would.

I am the exclusive owner of bustabit, and a partial owner of bustadice. Bustadice is it's own project separate from bustabit. I would describe it as "a bustabit inspired dice site".

Daniel is the guy who will be running it on a day-to-day basis (and in fact, for security I do not and will never have access to the servers). I helped a bit with the development, and going forward my primary role is just running the "seed server" and auditing the game results and co-signed of the cold-storage.


The site charges investors 25% of the house edge on each bet (regardless if win's or not). Other than that, everything goes directly to the bankroll (and thus investors).

Yeah, it's distributed exactly according to your investment. E.g. if you owned 10% of the bankroll at that time, you'd get 10% of someone else's dilution fee. (But then they'd be diluting you, so you'd later own less of the bankroll but own more money)

Pretty much[1]. But in exchange when people invest after you, you get their dilution fee. So all in all it should more or less balance out.

One of the main goals of the dilution fee is to provide stability to the bankroll, because it supports pretty extreme leveraging -- you don't want investors who leverage hard (to dominate the % ownership) and then divest as soon as a whale or big player comes (and actually needs that extra amount).

The other nice side effect is as you get diluted over time (assuming there are more and more bankrollers), you at least get compensated for your increasingly diminishing stake. In sites without a dilution fee, as an investor you often don't want new investors -- because you know they'll just dilute you



[1] It's actually a little different to prevent abuse. If you want to invest 10 BTC, what happens if first 10% (1 BTC) is deducted and then you invest the remainder (9 BTC). And then after that, the 1 BTC is given to all investors (which includes your 9 BTC). So the fee is actually always a little less than 10% as you're always getting a little back.  The reason it's done that way, is to avoid people fragmenting their investments into tiny pieces to save on fees

Congratulations on the launch!

I'm obviously pretty biased (as a partial owner), but I am really excited about this site and it's future. I know dicesites are a dime-a-dozen, but I think bustadice goes a few steps beyond. Things I think are really special:

*  Insane coin-selection optimizations, for both privacy and fees. We'll need real data to know exactly how well it works, but in something like 50-85% of the time (in simulations) it can avoid creating change (which is both expensive and extensively to cluster bitcoin transactions).   And users will enjoy *by far* the lowest fees of any bitcoin service. This alone was 3 months of solid work.


* An advanced "offsite" investor system, that allows investors to not just leverage but say exactly how much they want to invest without having to risk it all to a counter-party. Most investment schemes don't even offer leverage, but when they do they tend to use something like multipliers (e.g. 5x) which means when you lose, you're under utilizing your money and when you're up you're over-leveraging. The offsite system keeps it perfect all the time.


 * A really powerful provably fair system, that gives investors (and everyone) significantly improved guarantees. I tried to create a high level technical overview of how it works: https://gist.github.com/RHavar/285c295f7906e03369cd66580a1b5f45

I guess there's a million ways to structure an ICO, so it's hard to answer that. But I think the "bet on the house" system is fairest for everyone. I'm not really doing anything crazy, so the only thing I need money for is the bankroll. So I think it makes sense to just do that directly, and give people the freedom to enter/exit at will. Also the investor system has a really neat way to lower your counter-party risk by leveraging. So you can say "I want to invest 50 BTC, but I'll only put 10 BTC on site" and it'll risk as if you had 50 BTC (e.g. put it to toward the max-profit) but you get automatically stopped out if you lose 10 BTC. I'm not sure I could possibly reproduce something like that with an ICO.

I guess it comes down to a few things: Offering bigger bets will definitely be nice, it's pretty commonly requested by whales. The current max-bet is 1 bitcoin, which is pretty small compared to some other sites. But even so, largely because of the low amount of games per day (and everyone playing them at the same time) there's a *lot* of variance: (e.g. I think today I made ~50 BTC, and yesterday lost about the same) you can can see here: https://dicesites.com/bustabit


It'll also help in that investors will have to pay 25% of the expected value, regardless of how the site performs -- so that will also lower my variance considerably.

But probably the biggest reason, is I hope that the invest system makes bustabit itself a more interesting/popular/vibrant community. Giving people the option to share in bustabit's success (even if they only invest $50) I think has the potential to cause the the site significantly. I'm also expecting that I will be the vast-majority investor (I have a huge advantage over other investors: I have no counter-party risk) -- so even if it only achieves a modest increase in the site volume -- I think it'll be a net plus.

And finally, it gives me a lot more personal economic freedom. Pretty much >99% of my personal net-worth is tied up in bitcoin. One big reason I've been reluctant to diversify is that I really do need a bit stack of bitcoin for the bankroll (e.g. around this time last year, I lost >400 BTC bitcoin to a very unlucky month). If there's enough investor interest to the point the bankroll is bigger than it needs to be (e.g. players are no longer constrained by the bet limits), I'll feel a lot freer to start taking my own money out.

If anyone can help me find issues with:

dev.bustabit.com

It would be greatly appreciated. Issues can be reported here: https://github.com/bustabit/issues/issues

Or via the dev.bustabit.com chat, or through the dev.bustabit.com support system.


--

(And remember, dev.bustabit.com uses it's own isolated database and testbitcoin instead of real bitcoin. The testnet withdrawal system is currently not functional either)

Hmm, that doesn't sound right.

My back of the envelope calculations is anything >1000BTC.  Each bet you have 1 in 10 million chance of winning the jackpot. To qualify for the jackpot you have to "spend" 0.0001 BTC in EV. So I think you can model it as approximately: (1 / 1e7) * jackpot - 0.0001 = 0 which gives you jackpot = 1000 to be 0EV, and thus everything above that "profitable"

Yeah, that is coming with the v2 release (which you can play around with at dev.bustabit.com )

Stay tuned for a big announcement:

https://www.timeanddate.com/countdown/generic?iso=20170930T14&p0=60&font=cursive&csz=1


I don't want to spoil the surprise, so I won't say anything more than it's not related to bustabitv2 or anything  (which is going pretty well, you can play out with a pre-beta at dev.bustabit.com with some fake bits)

But it's exciting news!

I no longer have anything to do with MP. It's their project now, I try not get involved. I suspect what they mean in those profiles is they are the founders of the "MoneyPot Ltd." company, as opposed to the MP product. If I were them, I'd be using the phrase "co-owner" instead of "co-founder", but I honestly don't really care either way.




On a bit of a side note, I'll be fully winding that down soon (primarily to avoid any confusion) in the coming weeks. All the assets have already been transferred to  "Ryan Holdings" and that will the company I use