bustabit.com -- The Social Gambling Game

[fiscorcle]

Understood. I have my own script that managed to stay +EV for about a week before it ran out of RAM and crashed, but I don't really play all that much. I'm more interested in the theory side of it

I assume that you've been running your bot in the strategy editor and–since it's running out of RAM–keeps track of historical data in some way. For what it's worth you could connect your bot directly to bustabit. This method has the advantage of being able to persist data to disk in between gambling sessions among others.

BaBot, a stripped-down version of Shiba, provides an easy example of how to accomplish that.

Nah, the strategy tester doesn't track bonuses. No point in using that if I want to test if my script can be +EV...

[1714799270]

1714799270

[1714799270]

1714799270

[1714799270]

1714799270

[1714799270]

1714799270

[The_prodigy]

This was my first gambling website where i learn on how to gambling without reading mechanics i just follow the graph and test with the faucet (way back time) i always earn bits from faucet where i use that to bet and win bits now im deposit atleast 0.1 per week and making my bankrole doubled then withdrew i love to see my net profit positive.

[devans]

Nah, the strategy tester doesn't track bonuses. No point in using that if I want to test if my script can be +EV...

Perhaps we've misunderstood each other. I wasn't talking about a strategy tester, but suggesting that you run your script locally and connect it directly to bustabit instead of through the browser.

[fiscorcle]

Nah, the strategy tester doesn't track bonuses. No point in using that if I want to test if my script can be +EV...

Perhaps we've misunderstood each other. I wasn't talking about a strategy tester, but suggesting that you run your script locally and connect it directly to bustabit instead of through the browser.

Oh I see what you mean. Yeah, that's the next step. Don't have time with work right now unfortunately. Plus I actually want to run some analysis on the game results, but it's impossible to get a download of the game files

[RHavar]

So due to popular demand, I'm releasing the current dev branch of bustabit v2:

http://dev.bustabit.com

It uses bitcoin testcoins instead of real bitcoins, and comes with a faucet (you can also google "bitcoin testcoin faucet") and deposit them in your account.

You can find the source here:
https://github.com/bustabit/bustabit-client

It's completely full of bugs, and half-done stuff. The best thing is to make issues on that repository -- make note of your bustabit username, so you can be rewarded with some vanity stuff in the indeterminate future.

Also pull requests accepted  




Edit: in an unrelated note, here's a game dump of the last 10M plays: https://ln.sync.com/dl/0c20bf440#78hvmigq-xjiymwmj-vqcdt46n-is8vhdnc

[DarkStar_]

New version of bustabit looks pretty cool, and the "valor" seems interesting. Why are investor profits realized as silver and need to be fused with valor to get BTC though? Doesn't make a whole lot of sense to me.

You also noted that the database will be reset frequently, that means testnet coin balances will be lost as well as accounts, right?


Here's an small issue I found (I don't like using github, so I'll post it here)

https://i.imgur.com/i9dFPEz.gifv

Gives an error when you try to bet a multiplier of 1.1x, but still makes the bet. Same happens with 1.01x, haven't tested others yet.

I also tried placing a bet with 1k tBTC as well as 100tBTC at 1.01x, but it would say "Placing Bet" and then do nothing.


Edit: Does instant withdraw do anything? I added the 10 bits, but my withdraw was still in the queue and I had to pay another 100 bits.

[fiscorcle]

So due to popular demand, I'm releasing the current dev branch of bustabit v2:

http://dev.bustabit.com

It uses bitcoin testcoins instead of real bitcoins, and comes with a faucet (you can also google "bitcoin testcoin faucet") and deposit them in your account.

You can find the source here:
https://github.com/bustabit/bustabit-client

It's completely full of bugs, and half-done stuff. The best thing is to make issues on that repository -- make note of your bustabit username, so you can be rewarded with some vanity stuff in the indeterminate future.

Also pull requests accepted 




Edit: in an unrelated note, here's a game dump of the last 10M plays: https://ln.sync.com/dl/0c20bf440#78hvmigq-xjiymwmj-vqcdt46n-is8vhdnc

Much appreciated!

[RHavar]

Edit: I mixed up my terminology. Valor is what gamblers get. Silver is what investors get.

New version of bustabit looks pretty cool, and the "valor" seems interesting.

Why are investor profits realized as silver and need to be fused with valor to get BTC though? Doesn't make a whole lot of sense to me.

I'm trying to avoid repeating the mistakes I made in moneypot, so I guess here's the backstory: in MP I could've provided the bankroll myself (and/or shared it with BaB), but I intentionally didn't as I thought it would be a community-building feature type thing. I also assumed it'd be less stressful if I wasn't a big investor and didn't need to worry about variance.

However what ended up happening is:  a) Risking other peoples money is actually more stressful than risking your own.  b) Almost all substantial investor logged in only ~twice. Once to invest, once to divest.

During the ~6 (?) months before I sold the project, I believe those investors made something like 30-40% return in that time period. The whole thing left me thinking that if I were to do it again, I would structure it radically differently.

--

So taking from that, my original idea was that only active gamblers could invest in BaB, or there would be certain wagering requirements. Then after generalizing a bit, I came up with the idea: for every 1 bit of investing profit you make, you need to have wagered 100 bits. But if you're an investor, and not a gambler -- you can buy that "wagering requirement" from actual gamblers (and thus subsidize their play).

So as you can probably tell by now, "wagering requirement" is known as "valor". Gamblers automatically get 1 valor for every 100 bits they wager. And "investing profit" is known as "silver". And they are freely tradable between players, with the ultimate goal of community building. To also further this goal, investors profits will be automatically divested each day. This way investors need to login to every day and interact with the community if they want to experience the magic of compounding returns (i.e. investors profits are divested in the form of valor, and they need to either trade or gamble to get silver in order to convert it to bits, to reinvest).

From gamblers point of view, the new setup is pretty great. You effectively will get a ~50% off the house edge, selling your valor to investors

 

You also noted that the database will be reset frequently, that means testnet coin balances will be lost as well as accounts, right?

Yeah, although if you need/want your testcoins back, just shoot me an email and your address and I'll happily send

Here's an small issue I found (I don't like using github, so I'll post it here)

Thanks, I'll take a look soon. Something came up in my personal life that's taken me out of action for the last couple of days and will stop me working on it for at least a few more. However, I'm still checking in on things  and making sure all the prod support tickets are handled and everything is running smoothly. But there probably isn't going to be any new commits this week.

Edit: Does instant withdraw do anything?

I should probably create a different thread for the dev site, so people don't think there's a production problem. But the withdrawals there don't seem to be working at all, i think I broke something pretty badly. I'll be fixing that soon =)

[DarkDays]

Gamblers automatically get 1 silver for every 100 bits they wager.

When this is implemented on the main game, will it be retroactive?  For example, if I have 50,000,000 bits wagered over the lifetime of my account, will I automatically get 500,000 silver?

[kolloh]

That is a pretty interesting way of handling investing and fusing investors and gamblers together a bit more. I'm wondering how much of an impact this would have on the bankroll and max wins but curious to see how it it goes. Definitely seems like a nice added bonus for gamblers though!

[dooglus]

their profits are divested in the form of valor, and they need to either trade or gamble to get silver to convert it to bits, to reinvest

So 1 silver combines with 1 valor to make 1 bit, causing the silver and valor to vanish?

From gamblers point of view, the new setup is pretty great. You effectively will get a ~50% off the house edge, selling your silver to investors

That 50% number would depend on the market price of silver wouldn't it? Since valor is earned at a 1% rate (by definition), and silver is earned at a 0.5% rate (assuming an average 0.5% house edge) there will be roughly twice as much valor as silver in the world. That would lead to a glut of valor in the world, driving its price ever downwards.

Maybe the average house edge is being increased to 1% so the supply or silver matches that of valor?

Anyway it's an interesting concept.

[RHavar]

When this is implemented on the main game, will it be retroactive?  For example, if I have 50,000,000 bits wagered over the lifetime of my account, will I automatically get 500,000 silver?

It will only be on the new game, otherwise there will be an insane silver-glut  (and not to mention, bustabit current supports 0% house edge bets so it's often pretty easy to inflate your wagered amounts if you're careful)

Maybe the average house edge is being increased to 1% so the supply or silver matches that of valor?

Well derived  My current thinking is that the most straight-forward is just a flat 1% house edge on all bets. That's what's on the dev site right now (although I haven't yet provided methodology, so you can't really prove it). My current plans  are to remove bonuses completely from the main game (and put that in a separate, much more pvp game with ~10% bonuses). I really hate to remove the bonuses and variable-house edge, but with the vast majority of bustabit players playing "bonus oblivious" I can't help but feel I'm doing them a disservice.

With the new system, the effective house edge will be ~0.5% no matter how you play. When players do well and win money, there will be a glut of valor on the market (and no new silver) so I expect the house edge will be rather closer to 1%. And when players lose money, there will be a glut of silver on the market (investors wanting to realize their profits) so the silver be worth more and the house edge will be closer to 0%.

Smart traders I imagine will want to take advantage of this and "time arbitrage" the value of silver and valor to ~0.5 bits.

My hope is it'll be a fun little ecosystem

[Coingiver]

You'll be paying a dilution fee of "10%" when investing. why is it so high ryan ?  Is this used for the Faucet ? 10%

[RHavar]

You'll be paying a dilution fee of "10%" why is it so high ryan ?  Is this used for the Faucet ? 10%

It's actually a little lower than 10% (but 10% is the theoretical max), I made an issue for showing it properly.

But anyway, this is a one-time fee and actually doesn't go to me .. but goes to existing investors (based on their stake). The way it works precisely is that if you invest 1 BTC, you get a stake as if you had invested 0.9 BTC and everyone gets diluted as if you had invested 0.9 BTC. Then the bankroll is increased by 1 BTC.

This means the dilution fee is actually 0-10%. Ideally I would have made it a flat 5% or so, however a flat fee would be abusable (you could fragment your investment to get some of your money back).

Ostensibly the high level purpose of the dilution fee is to to compensate previous investors for the fact they are now being diluted. But the hope is that there are some other benefits, like less fickle investors (i.e. investors that divest when they see a whale playing, as that's when I need investors the most) and some marketing benefits.



BTW I'm still trying to figure out what a good dilution fee is, I'm thinking of anywhere from 5% to 25%. Any thoughts?

--

P.S. It's not spelled out, but the only fee that actually goes to me is a 10% profit cut from investors. So I think it's quite reasonable in that side

[dooglus]

My current thinking is that the most straight-forward is just a flat 1% house edge on all bets.

Innovative!

When players do well and win money, there will be a glut of valor on the market (and no new silver) so I expect the house edge will be rather closer to 1%.

That's only the case if the winning players insist of selling their valor instantly even though the price is low. A smart player (yeah, I know) would wait for others to lose, increasing the supply of silver, and bolstering the demand for valor. In the long run the price should return to 0.5 bits per valor, if not more, and so the house edge for the patient player should be no more than 0.5%.

And when players lose money, there will be a glut of silver on the market (investors wanting to realize their profits) so the silver be worth more and the house edge will be closer to 0%.

Similarly smart investors will be unwilling to part with their silver too cheaply, understanding that demand for silver will pick up when it becomes rarer than valor.

Smart traders I imagine will want to take advantage of this and "time arbitrage" the value of silver and valor to ~0.5 bits.

So it would be +EV to buy any silver or valor that is on sale below 0.4 bits, with a view to selling it at 0.5 bits later. That makes sense.

My hope is it'll be a fun little ecosystem

Will there be an exchange built in to the game site?

[dooglus]

the only fee that actually goes to me is a 10% profit cut from investors

Paid in silver, I presume?

[RHavar]

the only fee that actually goes to me is a 10% profit cut from investors

Paid in silver, I presume?

Yup!

Edit:

Will there be an exchange built in to the game site?

Yeah, there's already a half-assed one. It's pretty crappy as it doesn't have a matching engine (you either need to explicitly hit someones trade, or create one of your own). And to keep it simple, all order are "all or none". Improving it is something I eventually want to do, but something simple (along with a #trade channel) is probably good enough for quite a while

[DarkStar_]

BTW I'm still trying to figure out what a good dilution fee is, I'm thinking of anywhere from 5% to 25%. Any thoughts?

I would go 2% to 5%. It's high enough to prevent day trading (for most people anyway. Losing 0.5 BTC (Assuming you were a 10 BTC investor) every time a whale shows up is quite heavy), but not too high that new investors won't enter, especially large ones. If a 100 BTC investor showed up and saw a 10% dilution fee, I'm not sure if they would still invest because losing 10 BTC instantly really sucks, and the bankroll would probably stay similar to the amount it was before the grace period is over.



You need to add something showing the max bet on the game page, or a notice if it's too high and what the max is. I tried betting 1000 tBTC with auto cashout at 1.01x, which couldn't be placed but didn't return an error.

[kolloh]

BTW I'm still trying to figure out what a good dilution fee is, I'm thinking of anywhere from 5% to 25%. Any thoughts?

I would go 2% to 5%. It's high enough to prevent day trading (for most people anyway. Losing 0.5 BTC (Assuming you were a 10 BTC investor) every time a whale shows up is quite heavy), but not too high that new investors won't enter, especially large ones. If a 100 BTC investor showed up and saw a 10% dilution fee, I'm not sure if they would still invest because losing 10 BTC instantly really sucks, and the bankroll would probably stay similar to the amount it was before the grace period is over.



You need to add something showing the max bet on the game page, or a notice if it's too high and what the max is. I tried betting 1000 tBTC with auto cashout at 1.01x, which couldn't be placed but didn't return an error.

Yeah, I think that if the dilution fee is too high, it probably would turn away some investors after the grace period. You mentioned that investors profits will be automatically divested each day, would you have to pay the dilution fee each time you reinvest those profits?

[RHavar]

You mentioned that investors profits will be automatically divested each day, would you have to pay the dilution fee each time you reinvest those profits?

Yup   But all in all it won't make much difference (you pay dilution fees when you re-invest, but collect them when other people re-invest..)