Depends on the game, and depends on what you mean by "maximize betting volume". Dice for instance has a fixed edge (typically 1%) so your expected loss is always going to be 1% of you wager. In other words, your expected volume is going to be 100x your bankroll. So really, the only thing you can control for is the volatility. Make a lot of bets, and you're volume is going to be extremely close to that. Make a small amount of large bets, and you'll be able to do either a lot more, or a lot less than that.

Other game (like poker or bustabit) you should be able to have infinite expected maximal volume if you only play when the game conditions suit you

only if you verify it. They for instance could selectively be ripping people off. If you want to prove it was fair, you need to verify the games for yourself.

On bustabit, it would be: https://www.bustabit.com/game/2318863

Turning a 3rd-of-a-faucet claim into ~$40 =)

Provably fair simply means that you can check if you lost your money fairly or were cheated. Take notes of the hashes before the game, and after the game you can verify it. Might take a bit of research to figure out how it all works, but it's a great feeling. You will now have the power to know 100% if you were unlucky or cheated.

Try: https://github.com/kungfuant/webserver/blob/master/client_new/scripts/game-logic/script-controller.js




We've just passed 300k BTC wagered! At todays price, that's almost 125M USD in wagers!  

Awesome work Dexon! Cool to see this come to fruition. Framing minesweeper as a series of independent provably fair bets is pretty cool!

I'm going to put in another vote for PrimeDice. While it's probably not my favorite site, but it's the only site I'd personally have no worries with depositing, playing and withdrawing large amounts, which really is the most important thing in a commodity like dice.

Very interesting, and I do think doing something like this is better than trading on havelock for instance (external single point of failure, and what not).

But if you have an altcoin, how are you planning on doing distributions? Let's say you make 1 BTC you want to distribute, how will you do so? Perhaps buy 1 BTC worth of BKCoin from the market and destroy it?

Have you by any chance looked into colored coins? It seems it would simplify a lot of this stuff, and be a lot less work/maintance, and you could get a fair bit of free publicity for being the first real business to actually use it.


(Although if I was doing it personally, I'd just use the account system, and store equity % in the database. There's not much advantage to being decentralized when people need to trust you 100% anyway)

+1, yeah. It's getting ridiculous.


Yeah, I actually sponsored the development of some patches that landed in bitcoin 0.12 to support this. Assuming you put a lock on bitcoin (so you don't send in parallel), you can now go:

bitcoin-cli settxfee `bitcoin-cli estimatefee $n`
bitcoin-cli sendtoaddress $address $amount


My intention was that I'd default to $n of like 3, but users could pay extra to switch to 2 or 1. The big problem is, it's pretty easy to run out of utxo's if you're not careful.



There's also a privacy cost to this, but it does have a lot of benefits. I suspect this would work especially well for a site like PrimeDice, which could probably save a lot on its tiny withdrawals. At least my site the main cost is just sourcing the transactions, so it wouldn't help that much (but segwit would)

This one?

https://thebitcoinstrip.com/bustabit-stats.html

Glad to hear. Yeah, the current fees are starting to become pretty painful. For perspective, the last few withdrawals we processed had average of ~615 bits of fees (as determined by bitcoin-core's default dynamic fees), or about 21 cents. It's probably higher than most users due to our wallet receiving a lot more transactions than sending (but I suspect quite typical in any merchant-type situation). For now we'll continue to subsidize withdrawal fees but hopefully things rapidly improve. I think today I had over 10 support emails from people who's deposits aren't confirming.

The whole situation is pretty disappointing on multiple levels, I really hope bitcoin can get its act together.

Awesome, I've been thinking someone should do something for a long time.

What I think would be more fun though, is if you were in a "room" against someone. So you can rapidly play games in succession against a person, and see their history, and try spot and predict their patterns etc.


Anyway, when your site is fully ready send me a PM or get me on the bustabit chat, I can probably help you seed your site with some players =)

Yup, exactly this.  The real litmus test of a bustabit script is if it even tries to go for the bonuses. I've seen quite a few scripts shared that don't even make an effort, and then claim they're EV+.  And of all the people I know who actually built EV+ strategies (although they only work depending on game conditions), I don't think I've seen them ever share/sell the scripts

I guess it's probably the same in other fields, like sports-betting or insider trading 

Because what allows players to be +EV is the player-vs-player aspect, so just because someone is playing +EV it doesn't mean the house is losing money. It's actually much the same as how a poker house works in fact. There's a pretty technical FAQ the covers all the stuff and the nitty gritty maths and what not

On a side note, I'm not trying to promote BaB. I'm probably the only casino in existence who's trying to do the opposite. I've got 320 players online right now, and the server is bursting at the seems (aided by the syn floods going on), I'm probably going to remove the faucet next week to lower the amount of players

What makes you call it a "-EV gambling site"? Bustabit is one of the few games that has strategy, and people have been known to play with expected profits

While martingaler's are statistically profitable for the house, they're the opposite of fast profits. People who flat-bet make for much better customers

On a side note, what's the high-level motivation for wanting to use a timestamp? The client adds their bit of randomness with the client-seed, and the server already picks the server-seed and server-secret, so why do they also need a timestamp?

Even though many people hate MP's provably fair system, I think it makes a lot of sense as it's designed to be machine-auditable, and it's the absolute simplest way to do that. But since it seems rubies is user-facing, I'd strongly suggest to just copy JD's scheme, it's definitely the most user-friendly and human auditable.

Also seems my BB account had 1200 rubies in it, which like a true gambler I promptly lost. It's funny, because I did the same thing with clams when they first launched. I should stick to running casinos  

Please be careful as there is a bit more lag/latency than usual, and it's important that you always set a sane "auto cash out" (which happens server side). bustabit is being hit by an aggressive DDoS campaign, and some of the counter-measures required are causing a few problems

Hopefully they'll get bored soon, and I can step them down =)

It was changed on Feb 06, how time flies

Because we get a lot more deposits than withdrawals, our withdrawals tend to need to source a lot of inputs, which makes them big and expensive. Our average cost for a withdrawal is well over 100 bits (and I've seen a couple over 8000 bits). Once segwit is out, I can move to using segwit addresses, which will make spending from them considerably expensive.

One alternative would be to to add a "Deposit" fee of like 20 bits, and ~40 bits for a withdrawal. It would best reflect our costs, and probably the most fair; but I don't really want to field support from hundreds of users complaining about their missing 20 bits from deposits =)

Awesome news! Great to see something happening!

I have two questions:


How can you offer that? If the casino uses investor money to increase the site bankroll, you would increase the casinos-risk, while lowering it's profit. For instance, imagine if someone invested 1k BTC. It doesn't make sense to increase the max-bets, because you can't lose that 1k BTC (unlike a normal investment)




What does that mean? If someone places a 0.1% house-edge bet, what is the investor commissions? Also are investors bets still kelly-compliant when taking into consideration the new investor commission?