bustabit.com -- The Social Gambling Game

[dooglus]

We were talking on-site about how to fairly generate the 'crash points'.

Here's what I came up with:

the probability of turning i into 1.0 is p = (99 + i) * i/100

>>> i=0.5; (99 + i) * i/100
0.4975                                  (probability of 2x)
>>> i=0.2; (99 + i) * i/100
0.1984                                  (probability of 5x)

We want to be able to generate multipliers with this distribution.
Suppose we randomly pick p in range 0 to 1
We want to use p to create a multiplier m such that:

p = (99 + i) * i/100       (1)
m = 1/i                    (2)

So we need to solve (1) for i.
It's a quadratic:

p = (i + 99) * i/100
100*p = i*i + 99*i
i*i + 99*i - 100*p = 0

We solve like we were taught at school:
axx + bx + c = 0 has roots x = [ -b +- sqrt(bb-4ac) ] / 2a

In our case:

a=1; b=99; c=-100p

So:

i = (-b +- sqrt(bb - 4ac)) / 2a
  = (-99 +- sqrt(99*99 + 400p)) / 2

and m = 1/i
      = 2 / (sqrt(99*99 + 400p) - 99)

Checking, we find m for p=0.4975 and p=0.1984, and see that it's right:

>>> p = 0.4975; 2 / (math.sqrt(99*99 + 400*p) - 99)
2.0
>>> p = 0.1984; 2 / (math.sqrt(99*99 + 400*p) - 99)
5.0

Note that p=0 gives a divide by zero, so maybe was use a rand() that gives 0<=p<1 and use 1-p in the equation

[1714179567]

1714179567

[1714179567]

1714179567

[1714179567]

1714179567

[dooglus]

I was given an equation my Eric (site owner) that he claims generates the multipliers fairly, and plotted it against the equation I just derived above:



You can only see one line, because they're so close together.

But zooming in at the m=2 zone you see that his formula is being a little stingy, slightly ripping players off:



(click images to enlarge)

I'm surprised the two formulae give such similar curves.

[espringe]

Kick ass work dooglus.

However, I think your last post taken out of context is quite misleading.

I was given an equation my Eric (site owner) that he claims generates the multipliers fairly, and plotted it against the equation I just derived above

I gave you two ways to derive the game crash -- and the you plotted is was the one I said was simpler, but that I *didn't* use

But zooming in at the m=2 zone you see that his formula is being a little stingy, slightly ripping players off:

So harsh. To keep it in perspective, when doubling your money, that's it's a difference between a house edge of 0.5% and 0.5025%

I don't think anyone could possibly have a problem with that. However if I switch to such a formula for my provably fair version, (which is nice, as it has such a simple inverse:`(99/100x -1)`) -- I will update the calculator to reflect the imperceptible difference

[Mitchell]

I have been playing this game for three days now and I really like how it works. espringe is doing a great job

[kolloh]

This is a great game. You should try it!

[dooglus]

I have been playing this game for three days now and I really like how it works. espringe is doing a great job

I agree. It's a fun game to play. The small element of skill really makes a difference - though of course it's mostly just luck:



(click to enlarge)

[Mitchell]

I agree. It's a fun game to play. The small element of skill really makes a difference - though of course it's mostly just luck:

[IMAGE]

(click to enlarge)

Well, you have a shitload of luck/skill as it seems. I keep losing everything, haha. Anyway, it really is addictive and the owner is a great guy if you ask me.

[Ruzka]

Game is definitely very fun, even though I always lose my money

[dooglus]

Well, you have a shitload of luck/skill as it seems. I keep losing everything, haha. Anyway, it really is addictive and the owner is a great guy if you ask me.

Mostly I'm just martingale betting, and starting small enough that I can withstand a decent number of losses in a row.

I just hit 400k profit for the first time, on bet number 1337!



Edit: I got backed-off:

<nathaliesicard> doog~ stop betting
<sadTulipan> NO
<sadTulipan> PUTO
<sadTulipan> WTF!
<sadTulipan> wtf
<dooglus> hey nathalie
<dooglus> if I stop now, I keep my winnings
<dooglus> is that what you want?
<nathaliesicard> yes.
<nathaliesicard> i don't want you to win any more.
<dooglus> but the odds aren't in my favour
<dooglus> if I keep playing, I will probably end up losing it all back
<dooglus> oh - I forgot I was auto-betting!
<dooglus> I'll stop
<nathaliesicard> But you keep winning.
<nathaliesicard> Thank you. Muy amable.
<dooglus> No hay problema.

Edit2: Eric came back and told me I could keep playing.

[dooglus]

Here's a weird one:



As you can see (last but one bet) I bet 50k bits at a multiplier of 60k times, winning 3k BTC, but they didn't show up in my balance...

But seriously, I had accidentally clicked to bet while a previous game was running. Just as the game was ending I clicked 'cancel', and saw it cancel in the GUI, and started editing the bet amount (taking out a zero) but the bet was placed even though I clicked cancel, and hitting backspace in an attempt to delete a zero seemed to do a 'back' in the browser, taking me to the homepage of the site.

When I clicked 'play' to get back into the game, my history showed as in the screenshot, with all the multipliers wrong.


Also, the 'www.' in the domain name is mandatory if you're on the https site.

https://moneypot.com/ gives an error.

It would be better to get rid of the www., and redirect www.moneypot.com to just moneypot.com.

[espringe]

Here's a weird one:

Damn, you hit quite a bug combo there. Working on it.

For what it's worth, when you hit the back button or leave the page or disconnect -- the server will treat that as a cash out. The crazy history view is weird, I have no idea how that happened -- but the server side history looks perfect.

Thanks for continuing to play despite the real alphaness of it all =)

[dooglus]

Here's a weird one:

Damn, you hit quite a bug combo there. Working on it.

For what it's worth, when you hit the back button or leave the page or disconnect -- the server will treat that as a cash out. The crazy history view is weird, I have no idea how that happened -- but the server side history looks perfect.

Thanks for continuing to play despite the real alphaness of it all =)

It's still fun even with all the little bugs.

I do think you should put a warning / disclaimer on the site somewhere obvious though, that the game is still 'alpha' and that errors won't necessarily be made right - or whatever you decide to do about losses resulting from errors.

Since most of the issues are client-side it's probably impossible for you to know when people have lost coins due to bugs and when they're just trying to cheat you...

[kolloh]

Yeah, I've run into that display bug in the history window before as well. Not sure what causes it, but it is just a client side display issue it seems.

[mrccasals]

Cool game, I played for a while and I had a great time, recomended.

[espringe]

Edit: Challenge removed

[espringe]

Wow, pretty incredible! We've just reach 250,000 plays of the game -- which resulted in over 180 BTC been wagered by 1750 users. Can't tell you all how excited I am, and working my hardest on improving the game.

Thanks everyone! Love you guys

[apepoof]

Really nice game... I think the concept is refreshing after so many dice games.

[busterroni]

cool game just deposited some btc

[espringe]

Nice Spodermen, good to see you (And dooglus) in the green again.  Nice work guys! Now just just need to beat:

https://www.moneypot.com/user/alltheluck

[beaknuke]

hey gonna give this game a go, just curious of what programming tools you used and if it took you a long time?