December 9th, 2010 at 11:54:07 PM
permalink

You've lost on average $10... how do you get to +$300 over 50% of the time?

Simply enough... put together a method, and 1K can be yours.

Simply enough... put together a method, and 1K can be yours.

"Then you can admire the real gambler, who has neither eaten, slept, thought nor lived, he has so smarted under the scourge of his martingale, so suffered on the rack of his desire for a coup at trente-et-quarante" - Honore de Balzac, 1829

December 9th, 2010 at 11:54:56 PM
permalink

Quote:HKrandomPlaying $1 a hand would have an expected loss of $10 over 999 hands of Baccarat since the house edge is around 1% and 1% of 1000 is 10. You can stay up after over 1000 hands by betting $1 per hand.

There is no casino I know of (remember, this is a live challenge) where you can bet $1.00 per hand. The actual house edge is 1.06% so even at only $5.00 per hand you'd expect to lose about $75 after 999 hands. Setting that small point aside, you still haven't adressed my point about being down on your first big bet MORE than 50% of the time and being up on that first bet LESS than 50% of the time. I agree that once you're up $500 or so early on that staying above the $300 threshold shouldn't be too tough but seeing as you're going to be ahead LESS than 50% of the time, how do you expect to beat the challenge with that strategy?

Happiness is underrated

December 10th, 2010 at 2:37:35 AM
permalink

Quote:TheNightflyThere is no casino I know of (remember, this is a live challenge) where you can bet $1.00 per hand. The actual house edge is 1.06% so even at only $5.00 per hand you'd expect to lose about $75 after 999 hands. Setting that small point aside, you still haven't adressed my point about being down on your first big bet MORE than 50% of the time and being up on that first bet LESS than 50% of the time. I agree that once you're up $500 or so early on that staying above the $300 threshold shouldn't be too tough but seeing as you're going to be ahead LESS than 50% of the time, how do you expect to beat the challenge with that strategy?

You don't have to be up $500, since the EV of the 900+ hands is -75 you only have to be up $375 to have a reasonable chance to win. Betting $250 would leave you with around 2/3 (less the house edge) chance of winning $250 and about 1/3 chance of losing $500. After that the chances of winning $125 before losing $750 get rather high and you also use up many hands trying

The chance of losing a banker bet is also less than the chance of winning it, so betting everything on the first hand would indeed give you more than 50% chance of winning since ties result in a push, and winning $475 would give you very big chances of being up after 999 bets of $5.

December 10th, 2010 at 7:43:43 PM
permalink

The computer sim is too long and the live challenge too short to get meaningful results so...

Maybe the simplest solution is to combine the two challenges.

If you can beat the computer simulation, then you must further prove your system works in a live casino.

I would think if someone did beat the computer sim, then there would be the wish to see it work in reality. And this would prohibit someone from creating a system that exploits the challenge but still cannot beat the casino over time.

And Mr. Bluejay would not have to spend an entire month in a casino for a measly grand.

Maybe the simplest solution is to combine the two challenges.

If you can beat the computer simulation, then you must further prove your system works in a live casino.

I would think if someone did beat the computer sim, then there would be the wish to see it work in reality. And this would prohibit someone from creating a system that exploits the challenge but still cannot beat the casino over time.

And Mr. Bluejay would not have to spend an entire month in a casino for a measly grand.

December 10th, 2010 at 8:55:18 PM
permalink

Quote:HKrandomYou don't have to be up $500, since the EV of the 900+ hands is -75 you only have to be up $375 to have a reasonable chance to win. Betting $250 would leave you with around 2/3 (less the house edge) chance of winning $250 and about 1/3 chance of losing $500. After that the chances of winning $125 before losing $750 get rather high and you also use up many hands trying

The chance of losing a banker bet is also less than the chance of winning it, so betting everything on the first hand would indeed give you more than 50% chance of winning since ties result in a push, and winning $475 would give you very big chances of being up after 999 bets of $5.

It looks like you've got a winner there HK. Ignoring ties I calculate the odds of winning on the Banker side are about 2.7% better than on the Player side. So, you play $400 on the first hand and if you win, you'll receive $380 and ride out the remaining hands to hold on to at least a $300 profit. With a slightly better than 50% chance of winning the first hand you'll win MB's challenge more than 50% of the time. Of course if you lose the first hand you're in dire straits but this will happen les than 50% of the time AND you've still got $100 left to at least have a shot at making the money back.

So, MB, has HKrandom found a flaw? It looks like it to me and is almost too obvious... is something missing here?

Happiness is underrated

December 10th, 2010 at 10:27:38 PM
permalink

I ran a simulation using this algorithm for bet sizing in Baccarat :

if(credits < 400 ) { betSize = credits} // (bet everything is almost out!)

if(credits <=500 && credits >= 400){ betSize = 400} // (if you have 400 credits, bet big)

if(credits > 500){ betSize = 1; } // (if your above 500, bet 1 at a time)

You could easily make this more complex. Over 10,000 attempts, 50.93% of the attempts ended up above $800. The average loss was -$6.68.

This attack is similar to the first one weaselman made... shit or bust early, then grind out (his was sit out, but the difference is minimal).

If you try :

if(credits <=500 ){ betSize = credits}

if(credits > 500){ betSize = 5; }

I only got 40.5% winners.

if(credits < 400 ) { betSize = credits} // (bet everything is almost out!)

if(credits <=500 && credits >= 400){ betSize = 400} // (if you have 400 credits, bet big)

if(credits > 500){ betSize = 1; } // (if your above 500, bet 1 at a time)

You could easily make this more complex. Over 10,000 attempts, 50.93% of the attempts ended up above $800. The average loss was -$6.68.

This attack is similar to the first one weaselman made... shit or bust early, then grind out (his was sit out, but the difference is minimal).

If you try :

if(credits <=500 ){ betSize = credits}

if(credits > 500){ betSize = 5; }

I only got 40.5% winners.

"Then you can admire the real gambler, who has neither eaten, slept, thought nor lived, he has so smarted under the scourge of his martingale, so suffered on the rack of his desire for a coup at trente-et-quarante" - Honore de Balzac, 1829

December 11th, 2010 at 12:44:50 AM
permalink

Quote:thecesspitI ran a simulation using this algorithm for bet sizing in Baccarat :

if(credits < 400 ) { betSize = credits} // (bet everything is almost out!)

if(credits <=500 && credits >= 400){ betSize = 400} // (if you have 400 credits, bet big)

if(credits > 500){ betSize = 1; } // (if your above 500, bet 1 at a time)

You could easily make this more complex. Over 10,000 attempts, 50.93% of the attempts ended up above $800. The average loss was -$6.68.

This attack is similar to the first one weaselman made... shit or bust early, then grind out (his was sit out, but the difference is minimal).

If you try :

if(credits <=500 ){ betSize = credits}

if(credits > 500){ betSize = 5; }

I only got 40.5% winners.

Betting all your bankroll at the beginning is not a good idea since you are taking too much risk for very little additional reward; if being up 60% by the end is enough to win the challenge then the second system is not appropriate. You could also increase your chances of winning with the first system by betting big every time you are down in the last few rounds (e.g. if there you have a bit over $700 and there are 3 rounds left the optimal strategy would be to bet $100 and martingale until you reach $800 or bust).

Requiring to double a bankroll rather than getting a 60% return would make these systems much less likely to win.

December 11th, 2010 at 12:47:41 AM
permalink

The biggest difference is once you get into maintenance mode... if you can bet $1, your okay, as soon as it got to $5, not so successful.

The second was just a rough test to see the effect of a simple approach. There's plenty of different changes I had a try at but they all came in around 40-45% if the grind bet had to be $5.

The second was just a rough test to see the effect of a simple approach. There's plenty of different changes I had a try at but they all came in around 40-45% if the grind bet had to be $5.

"Then you can admire the real gambler, who has neither eaten, slept, thought nor lived, he has so smarted under the scourge of his martingale, so suffered on the rack of his desire for a coup at trente-et-quarante" - Honore de Balzac, 1829

December 11th, 2010 at 1:10:58 AM
permalink

What if you only try 1,000 rounds like proposed in the initial live challenge? 10,000 round would mean an initial bankroll of $5,000, which would be much easier to preserve with $5 bets.

December 11th, 2010 at 1:16:18 AM
permalink

I am running 1000 hands. Just repeating the test 10000 times to estimate a win/lose ratio."Then you can admire the real gambler, who has neither eaten, slept, thought nor lived, he has so smarted under the scourge of his martingale, so suffered on the rack of his desire for a coup at trente-et-quarante" - Honore de Balzac, 1829