DiceBitco.in - New Thread to Discuss

[xilatleo]

Makes me wonder if same kind of thing happened at his other site diceliteco.in

max win of 444 LTC  Site profit = -17,174.3479 LTC

I wonder if it has same owner as dicebitco.in?

[1715044077]

1715044077

[1715044077]

1715044077

[1715044077]

1715044077

[1715044077]

1715044077

[1715044077]

1715044077

[nicolaennio]

About how impossible mateo's winning were.. I made a few mistakes when computing the prob and corrected it. Sorry ! Since it was discussed here and here, I thought I'd come back.

The approximation actually gives a probability of winning +600 btc like mateo did of about 1 in 1 to 10 millions... (for extreme values like here final result is highly influenced by the computers numerical methods and their accuracy).
See what a hundred thousand simulations of normal 60k bets looks like (histogram=simulations, curve=model), the max profit was 424. Play with the numbers...

I do not think this is the correct model for computing the probability of mateo. It seems you assume that the player arrives and blindfolded bets for 60K times and at the end he just leaves the website with what is left in the pocket. But this is not like that, the players bets until there is money in the bankroll and for fun (I guess winning the bankroll is a lot of fun). If he just wins the bankroll in the middle of playing he just goes away, without extra rollings.

Also I cannot reproduce the result the 1 to 10 millions estimate, I run a simple script simulating a player betting 60K times at 0.5% of the bankroll (half-kelly as in the website) and on 10000 repetitions I get 4 times a final bankroll smaller than 15%.

I'm sorry, but you're using Dubious math.  Kelly Betting is actually the amount the House "bets" which will increase their bankroll the quickest.  The Player does not enjoy kelly bets as they are betting in a -EV situation.  Kelly Betting is only for +EV (which in the case only pertains the House).  Thus the formula you are using to determine the chances of bankroll going to a certain % is not correct.  The correct % will be much much lower than 2%.

I do not follow your reasoning (maybe you can be clearer?) but I am using a very standard tool. If you run some simulations you can check by yourself that the formula you find in the link is correct.

[Joca97]

Makes me wonder if same kind of thing happened at his other site diceliteco.in

max win of 444 LTC  Site profit = -17,174.3479 LTC

I wonder if it has same owner as dicebitco.in?

it probobly is

you should probobly avoid that site as well

[pawel7777]

I wonder if it has same owner as dicebitco.in?

Yes, it's the same owner(s). He mentioned it in the official DB thread (now locked). He also said that they're considering launching a DOGE site, but I sense that's no longer the case

[chalidore]

Is there any info who lost the most money in this scam?

[Joecker]

For the sake of doing "good" math here ;P ..
As pointed out by a1choi, Kelly's formula doesn't not apply. And to take into account the probability of ruin of the casino here (like nicolaennio wish), one as to compute the conditional probability P[S > 600 | S< initialBankRoll]... which leaves the probability unchanged if initBR=1100...

[a1choi]

About how impossible mateo's winning were.. I made a few mistakes when computing the prob and corrected it. Sorry ! Since it was discussed here and here, I thought I'd come back.

The approximation actually gives a probability of winning +600 btc like mateo did of about 1 in 1 to 10 millions... (for extreme values like here final result is highly influenced by the computers numerical methods and their accuracy).
See what a hundred thousand simulations of normal 60k bets looks like (histogram=simulations, curve=model), the max profit was 424. Play with the numbers...

I do not think this is the correct model for computing the probability of mateo. It seems you assume that the player arrives and blindfolded bets for 60K times and at the end he just leaves the website with what is left in the pocket. But this is not like that, the players bets until there is money in the bankroll and for fun (I guess winning the bankroll is a lot of fun). If he just wins the bankroll in the middle of playing he just goes away, without extra rollings.

Also I cannot reproduce the result the 1 to 10 millions estimate, I run a simple script simulating a player betting 60K times at 0.5% of the bankroll (half-kelly as in the website) and on 10000 repetitions I get 4 times a final bankroll smaller than 15%.

Actually, isn't that how you should model someones gambling session?  Blindly gambling until they stop when they've run out of money, or when they've hit a certain number of bets?  (there is 0% chance in which a player can walk away with 100% of the total casino bankroll)  Which is what Joecker's doing in his model.

I'm sorry, but you're using Dubious math.  Kelly Betting is actually the amount the House "bets" which will increase their bankroll the quickest.  The Player does not enjoy kelly bets as they are betting in a -EV situation.  Kelly Betting is only for +EV (which in the case only pertains the House).  Thus the formula you are using to determine the chances of bankroll going to a certain % is not correct.  The correct % will be much much lower than 2%.

I do not follow your reasoning (maybe you can be clearer?) but I am using a very standard tool. If you run some simulations you can check by yourself that the formula you find in the link is correct.

What i'm saying is that your premise is incorrect, Because you are using the Kelly Formula to predict risk of ruin.  But the formula requires the betting situation to be +EV (positive).  Dice is -EV (1% house edge remember).

What joecker has done is the correct model.. let me explain in more detail.  here is a simple model of how this dice session should go.  The Total wagered amount by mateo should be around 50-60K, his profit around 500-600.  The Standard Deviation of dice at 49.5% is 1 (other %'s will provide very different standard deviations).  His total bet # is 60K. So average bet size is 1BTC.  We can then find the probability of this happening by doing a statistical test.  I have reposted Joecker's analysis below.

(...) By the way, for fun can someone explain the math of how monstrously improbable Matteo's run of "luck" (fraud) was?

Let X_i be the random variable modeling the profit of each bet.

Assumptions:
1) profits went from 260 to -330, that is manlteo's profit = (+/-) 600btc
2) N=60k bets of 1btc each @ 2x payout
3) math bullshit (iid random variables), q=0.495, P[X=+1] = q, P[X=-1] = 1-q.

Let S= X_1+X_2+ ... + X_60000. We should expect E(S)= 60000 x ( 1x0.495 + (-1)x0.505 ) = -600 (a loss).

We want to know P[S >= 600]. Central limit theorem states that  [S - E(S)] / stdev(X)*\sqrt(N) =: Z is normally can be approximated by a normal distribution N(0,1).

var(X) = E(X^2) - E(X)^2 = q+1-q - (2q-1)^2 = 4q(1-q) = 0.9999, so stdev ~ 1

So
P[S >= 600] ~= P[Z >= [600 -(-600)] / (1*\sqrt(N)) ] = P[Z>= 1200/244] = 1-P[Z<=4.89922] = 4.8x10^-7.

To be honest, the average betsize might be smaller than 1BTC since after the bankroll dropped smaller than 200btc, the maxbet was smaller than 1btc.  This would make the analysis above on the conservative side.  With a Z-score approaching almost 5, the evidence points more and more to something virtually impossible happening.  It could happen, but at the probability of 4.8x10^-7.

Joecker, I don't understand why you need to model the stdev as approaching 1.  The standard deviation of the winnings of dice played at 49.5% is = 1.  you're probably using a different method to calculate that?

i just used a dice simulator over millions of rolls to determine the STDEV of the dice game at different %s, since different % really do provide a different standard deviation.

[Cluster2k]

With a Z-score approaching almost 5, the evidence points more and more to something virtually impossible happening.  It could happen, but at the probability of 4.8x10^-7.

[jjc326]

Hopefully this will teach people not to flock to one person and trust anyone, even someone like dooglus, with their BTC until they do their own research and a site establishes a track record.

[a1choi]

With a Z-score approaching almost 5, the evidence points more and more to something virtually impossible happening.  It could happen, but at the probability of 4.8x10^-7.

Hehehe.  Thats EXACTLY what i'm telling you.  :p

[wayshegoes]

What's the probability of completely anonymous developers stealing money that people basically gave to them?

1

[Joecker]

mmmkay, if we assume that initial bankroll of casino (u) and of mateonl (v) are both finite, with the above notations and assumptions, the probability we're actually looking for (what is the probability of winning 600btc or more with betting 60k times 1btc @49.5%?) is P[S>600 | v<X1+..+Xj<u for all j]. There's no non-empirical methods that I know of to compute that, but given the number we're talking about it's clearly of no influence here. The probability of ruin of mateo P[X1+..+Xj<-v for some j] and the probability of ruin of casino P[X1+..+Xj>u for some j] are not directly related to the above quantity...

Btw for this type of calculation, due to the nature of normal law, usual tools precision is only about 7-8 decimals. Indeed, when I plugged in the more precise numbers (for stdev etc), Mathematica returns P = 0.

(...) http://www.themerrymonk.com/wp-content/uploads/2013/08/theres-a-chance.jpeg

Of course there's a chance ! .. but you have bigger chance of dying on your way to buy an Euromillion ticket (.. than to win the jackpot). Rip to all those bastards who died their way to the shop!.. that's why I feel much "safer" with online gambling ;P


LOL.. yea not funny for the victims :// But I am amazed by the accuracy of the calculations ;P

What's the probability of completely anonymous developers stealing money that people basically gave to them?

1

[dooglus]

Currently DB is reporting:

Bankroll:   522
Profit:   -320
Invested:   1131

Don't all invested funds get added to the bankroll?  Why is the bankroll less than half of what's currently invested?

"Invested" doesn't change when players win or lose. Bankroll does; that's one big difference.

[dooglus]

Makes me wonder if same kind of thing happened at his other site diceliteco.in

max win of 444 LTC  Site profit = -17,174.3479 LTC

I think that's a player profit of -17k, so a site profit of +17k.

DB used to show it the same way until a few of us pointed out that the negative number could be offputting.

[wayshegoes]

There is also the possibility that mateonl was freerolling, waging numbers on a screen vs. real BTC on a provably rigged game.

When is manl going to publish mateo's seed?

That's right, never, because manl is in another country now doing lots of cocaine and hookers. He can afford it.




Bankroll:   522
Profit:   -320
Invested:   1131

522 + -322 = 844

844 = 1131

Profit = profit??

Maths is truly magic.

[Joca97]

Makes me wonder if same kind of thing happened at his other site diceliteco.in

max win of 444 LTC  Site profit = -17,174.3479 LTC

I think that's a player profit of -17k, so a site profit of +17k.

DB used to show it the same way until a few of us pointed out that the negative number could be offputting.

hmm thats really wierd

but if dicebitco.in got rigged like that

probobly dicelitecoin got rigged as well thats how i look into it

[Cluster2k]

Magic internet net money accounting

[chalidore]

So what's the total damage here, players got some money refunded, investors took their money out. There is that 500-600BTC loss but no one complaining on forum that they got screwed. Did any investor whale complained on the forum about this?

[Cluster2k]

So what's the total damage here, players got some money refunded, investors took their money out. There is that 500-600BTC loss but no one complaining on forum that they got screwed. Did any investor whale complained on the forum about this?

Chances are the people with the highest investments were the ones watching the closest and pulled out before mateo arrived.  Whether someone had 100 btc invested or 1, questions have been asked and no answers have been provided from DB.

[a1choi]

So what's the total damage here, players got some money refunded, investors took their money out. There is that 500-600BTC loss but no one complaining on forum that they got screwed. Did any investor whale complained on the forum about this?

the people that got hit have already complained.  They lost approximately 85% of their investment.  But because it was lost to a gambler, they don't complain as hard as someone might if they had their bankroll stolen outright.