More just a shower thought, rather than something I imagine will ever get implemented. But one of the things I dislike about my bitcoin wallets, is the long history they have accumulated over the years. At a certain point it just becomes a liability, and yet moving to a new wallet is rather annoying (not least because of the large consolidation that would take place).


To generate the ith address, an HD wallet does something along the lines of hash(seed + i)   -- while what I'd like is something along the lines of  `seed = hash(seed)` i times.


This would have the property of being able to fast-forward the seed a certain amount of slots, in order to forget the previous history. For instance in one of my wallets, I could safely fast foward the seed 130 slots without losing anything of value (and in fact, only losing information I'd rather lose).

I still have exclusive and sole access to the audit-server used for bustadice -- along with being a keysigner for it's cold storage (in a 2-of-3 scheme). In this time, the cold wallet has never changed and Daniel still uses the same GPG key used to communicate with me as he did when first creating bustadice. I think however the strongest evidence to suggest that ownership has not changed would be that bustabit uses the same hashchain as originally created by Daniel. I can not imagine any sane person would operate bustabit knowing that another party knows all future games.

I think your point about .com is right, and it's a bit silly to be using anything but it (unless you also have the .com secured). But I also just don't think domains are that important in general, as long as you can find something moderately brandable. Also for the love of god, domains that are just some keywords (e.g. bitcoincasino.com ) are pretty awful.

I'm going to go out on a limb here, and guess that Daniel probably doesn't care. The people that are really important, and that he really needs to please are gamblers, which are his actual customers. Bankroll investors are just a means to that end (i.e. make sure the limits are high enough, and the site can smoothly operate through huge profit swings, etc). And according to the stats, Daniel has already paid  investors  3875 BTC  for their services (i.e. averaged out to 176 BTC a month?).

Since I have no shame, and don't mind making an ass of my self speculating -- I'm going to guess he looked at that number, and realized since the size of the bankroll is already bigger than it needs to be -- and decided he'd rather just keep some of that money for himself. He probably realized he couldn't easily raise the %-of-wagered commission, because that would lower the max-profit by a lot, and put too much variance on investors who would be tempted to divest, which could create a vicious feedback cycle.  So switching to % of profit, makes a lot of sense. It means that max-profit will go *up* unless more than 1/3rd of the investors divest -- which seems highly unlikely (afaict, actually there's been 1 more btc invested than divested since he announced the change).

---


From a personal perspective, I'm a bit of two minds about it. It's sad to watch these great investing opportunities diminish (or even go away, like yolodice recently) but it's also nice to see these businesses mature too. The vast majority of crypto projects that ask for "investments" are so focused on getting investor money, that they lose sight of creating a viable business.

On bustadice's "sister" site you announced:


is this going to come to bustadice too, or do you plan on having different commission structures on the different sites?

Hmm, I don't think that's the way investors look at it -- or at least not how I do.

When I look at a bankroll investment, I try calculate if I invest X ... what is my expected return? What is the expected variance? And how would I (roughly) quantify the counter-party risk (i.e. malicious or screw up sort of problems).

---

So trying to do rough calculations against BaB, I'd eye-ball the average volume as 400 BTC/day, which would give investors (after commission) an expected 730 BTC per year. 1 BTC would buy you ~0.015279% of the bankroll, so an investment has a rough expected return of 11.15% a year?



So it's considerably worse than yesterday (which was double that) and even less than when i originally invested (I was estimating 30-40% p.a. returns)  but counter-party risk is the real thing that scares me (i.e. is the real reason I have only invested a small fraction of my bitcoin in the bankroll and not the whole thing).

I also think all the smart investors appreciate the importance of diversification (especially in something ultra-high-risk like bitcoin-casino-bankrolling) so I imagine it's going to more just be an issue of changing how much they invest, as opposed to investing or not investing.

Honestly, I think this your fault. You should've just waited.


That said, I really don't like how casinos like this handle it. Especially the "You may continue playing while you wait" is a little predatory. I've long suspected some casinos have intentionally done this sort of stuff to encourage people to gamble (i.e. i've seen some casinos where they ask large winners to wait >2 days+ etc).  I think the best way to handle an empty hot wallet is to queue any withdrawal that couldn't be processed (and deduct the amount from the persons balance, so they don't gamble it). And of course, any serious casino should be able to process almost every single withdrawal without delay.

Sorry for your loss though, 

My guess is their very generous affiliate program. And plus, it's not like they screw everyone. I suspect 99% of the time, they operate pretty reasonably. Just they're always on the lookout for stupid stuff to screw customers over, if it benefits them. Like a simple example is if someone wins a lot and tries withdrawals, they generally try do an extensive KYC to hope the person doesn't want to go through the process, or can't go through the process, is in breach of something etc. And then they just pocket the money. It's insane. But if someone does pass and hasn't gambled their winnings in the mean-time then they probably will begrudgingly process the withdrawal.

My favorite example, which is something they no longer do, but I think gives a great insight into their method of operating is this:

https://bitcointalk.org/index.php?topic=1454521.msg14704015#msg14704015


Where the tldr; is their terms-and-conditions had a very predatory "Do not bet over X" clause, but their software never stopped you betting over X. Someone won a lot of money and happened to also bet over X  (although that was totally inconsequential to their win) so they thought it was reasonable and defended seizing the persons entire balance before of an accidental ToS violation that their own software didn't enforce.

So they might not directly be a criminal organization, but they operate like one of those seedy strip clubs whose seemingly only goal is to try rip off everyone as much as possible while having some veil of legitimacy.

Ok. Let's say this happened, why do you think it's reasonable to keep the dude's money?

Sure, by all means, you should ban him (and probably his mom for good measure) from using your services in the future. But just pocketing his money is outrageously unethical. And please don't act like it's a legal requirement, as you guys appear to operate under a curacao license, which is the same (rubber-stamp) process I went through several years ago back when I used to run a casino -- so I know for a fact there's no legal requirement to pocket people's funds in situations like this (just one to deny service).

bitstarz are a bunch of clowns. As far as I can tell their business model is basically offer a huge referral program (40%+?) to get a lot of exposure from reflink aggregators like askgamblers -- and then if players happen to win, try find any possible excuse to discourage or outright not pay winners.

It's a real shame really, because not all casinos are so trashy -- but it really pays to do a significant amount of research on a place before you start playing to avoid a repeat

There's easy-enough ways to make the jackpot a lot closer to provably fair, like using a (verifiable) delay function on the player-list. Actually it wouldn't even need to be cheap to verify, as people don't need to do so in real time.

As you said, even with that, you could also use house-players to increase your chance of winning. But at least that's something that would be a bit more obvious (amount of volume spikes before each jackpot, making something someone expected to be +EV to -EV).

Anyway, the real fix is just being honest and direct that the jackpots are not provably fair, and retracting previous claims. As really they have none of the guarantees expected of provably fair. I think it's a big deal, considering it seems like the vast-majority of your users are playing for the jackpot (all the 1.01x bets) so it's very important they are aware it's not provably fair in any sense (i.e. you could trivially win them all, or make a particular user lose them all in a non-detectible way)

Given the way provably fair works, it should be relatively easy to go through all the old bets and award wins that were improperly graded as a loss (and notify people of the extra balance now in their account). I don't see how you could justify not doing it.

I wouldn't hold my breath though, I think they've always had a bit of a checkered past ... only doing the right thing if their feet were held to the fire.

 
I am pretty sure that I remember to register an account, there was something like  "I have read and agree to the <link>terms of service</link>". I think the popup we all just saw is for users who have agreed to the old ToS to let them know there's now a new one. I can't find an archive of the old ToS, but if my memory is correct the only substantial is the account recycling?

Since everyone is talking about ToS, the one thing I feel is missing is an official policy on how investors are exposed to things like mistakes/hacks. For instance, there was that bug in the API that allowed someone to find out the current-game multiplier and bustabit paid for it (150 btc?) instead of investors. Is that an official policy? Presumedly if the loss was too big, at some point bustabit couldn't/wouldn't be able to pay for it (as after all, the point of investors is to lower/eliminate risk). As an investor, it'd be nice to know if there's an official policy on handling situations like this.

Yeah, that's my point. They've likely overlooked the error for so long, since it benefits them. (Losing money is always a lot more obvious)  

I tried to play at FortuneJack to verify it, but trying to login gives an account suspended error. (lol?)


Anyway, it looks like an off-by-one error. Pretty sure this is the same bug that crypto-games had for a while, except they did it the other way (an extra possibility to win, so it was a negative house edge on high mults).

Not really sure how this happens to be honest. The maths is pretty insanely simple, something you'd expect most high-school students would figure out. I guess though when the error is in your favour, it's easier to overlook. 

I might be wrong here (I had trouble finding the old ToS), but I'm pretty sure the only new term is actually that about accounts being recycled after 2 years of no use. I don't think BaB has ever allowed US users (and other places where online gambling is illegal) and has always dealt with it by asking them to withdraw their balance and refrain from using it in the future. I think just the new ToS makes it a bit clearer (like right at the top, instead of buried in the middle).

Ok fair. Seems like you're using an established magic value, and I haven't been able to create two different valid signatures from the same message. So I retract my claim about it not being provably fair, it appears to be.

--

I will spend more time testing it, if you can confirm:  If I am able to find a way to create 2 or more valid signatures that would validate using your openssl verification you use (openssl dgst -sha256 -verify $PUB -signature $SIG $MSG) for a given message, will you award the 100 ETH bounty?

Can I claim the eth bounty? This is not provably fair, because step-2 is not verifiably deterministic. You could create as many different "[Seed]" as you wanted with RSASSA-PKCS1-v1_5 signing, that all pass the test 5 by twiddling with thing like the padding and stuff. For this scheme to work, you'd need to use a digital signature algorithm that generated only a single valid signature for a given message.

That would be a matter between rocketpot and the CGF, but I very much doubt it. I think the biggest issue would be them misrepresenting their jackpots as being provably fair, when they're clearly not. And it's not even a small matter, considering most players there seem to be only interested in hitting the jackpots (betting at 1.01x).



To be honest, I don't think incorporation really protects players. If anything it protects the business against players.  Imagine there was a huge fuckup and the site lost everyones money. If it's incorporated, in general the individuals aren't really liable for damages and they can just close down the business as bankrupt.

Except that's not what provably fair means. You can directly control who wins the jackpot, if you have a house-player. And there's absolutely zero way for people to have any idea if you're doing it or not.

That's not bad per se, and doesn't mean you're cheating. But it means it's not provably fair. Full stop. Any attempt to market is as provably fair is dishonest.



Actually you have a (small) point.  In bustabit for instance there's a "forced cashout" when the server will cash people out due to a per-game stop-loss being hit. And you're right, this is not provably fair. But it's also why bustabit tries so hard to make it so hard for this to ever happen. Like such that it it doesn't even accept single bets that could trigger it. I suspect it hasn't even happened at all in the last year, so in reality it's just not a real issue. But you're right on this point.  But there's several orders of magnitude difference.


I think you're smart enough to know this is technobabble. Before the round starts the house has all the information it needs to know who will win the jackpot. It also has the ability to control who joins the rounds (which dictates who wins). It can *very* easily add extra players and modify the bet amounts such that the house player wins. The house can trivially brute force even a million combinations in a fraction of a second to find out which combination will result in the house winning, and then execute that one.


---


Anyway your response to this is extremely disappointing to say the least. There were two obvious ways to correctly handle this situation:

a) Make the jackpot provably fair

There's a few ways to do this, one neat way would be run a verifiable-delay-functions on the game player bet list, where each client also sends a randomized seed with their bet. Then if the game goes over XXX you use that result to calculate the jackpot winner.


b) Just let everyone know "our multipliers are provably fair but our jackpots aren't! it's just a bonus"


but instead you've decided to go down the third pat, and just spout nonsense and misrepresent it as something it's clearly not. I would encourage people to not play here until your marketing matches reality.