can this way beat baccarat?

it is in Macao.

going to the vip room. you buy the dead chip: 100: 125, i am sure, i checked yesterday. it is 1.25%.(by the way, in every different.casino, even the same company, they have vip rooms in different casinos, their ratio is different.)

as we know, the standard advantage of betting banker from casino is 1.06%.

if so, when we buy dead chip, the EV is 1.25-1.06=0.19%. so every bet we make is making 0.19%?

but for dead chip, if when lose, we don't get cash chip. only after we win, we get cash chip. does this mean EV is 1.25%/2 or not? i guess not. but i want to get sure answer from some pros here.

also, i just want to ask this, i will not try this. because i did one test, by 50,000 hands, betting banker can be 1.32% house edge from casino. the 1.06% is only from millions of millions of hands.

That seems complicated. And kind of scary(the last part).

Quote: Dean

That seems complicated. And kind of scary(the last part).

it can be like this..thanks. i guess i will do another post in blackjack on house edge of short run.

maybe i need to reput my question:

i buy dead chip: 10000 dollor to get 12500 dollar dead chips. to play baccarrat.

for dead chip, if we lose, we don't get cash chip. only after we win, we get cash chip. does this mean house edge is cut by 1.25% or 1.25%/2?

for baccarat, as we all know, betting banker all the time, house edge is 1.06%.

can i win money if i keep betting banker?

dude, 1.25% is 10,125 for 10,000, not 12500
get your primary school cert before you go to casino.

and no, you can't beat the casino with that. you are not the first one to have think of that.

it's half, assuming that you keep the dead chips and your win is paid in negotiable chips.

if you were the teacher there, i wouldn't go.

it is just a careless mistake.

to prove the asker is knowing sth, i actually had the answer, i just want to be sure when some one also thinks like this.

in theory, it can.

every betting, you lose 1.06% by betting banker because of house edge, but because of dead chip, you make 0.19%.

it happened some one beat casiso by this.

but it is a hard job. because the house edge can vary even in 50,000 hands.

i even had written formula in excel,(if u have college degree, it should be easy to learn it.).

this is 10 times, each time is 50,000 hands.always betting banker;
-0.018409
-0.010453
-0.015406
-0.003628
-0.005266
-0.015679
-0.012832
-0.015679
-0.01174
-0.004759
by average
-0.0113851

and so it is still risky and long time job.

thanks for your reply. i will think about it.

Firstly I am assuming the punter makes $1 bets on either the Player or Banker. A win will result in a payout, which is kept and the [free] bet stays in play. Thus, rather than 102500 hands, the calculation should be when there have been 102500 losses (i.e. Player when betting Banker or vice versa) and see how much was returned.

The following table is based on ten sims of 2784 shoes and (since the numbers weren't exactly 102500) factoring the numbers accordingly. Thus in the first row a bet on the Player would have waited for 102500 Banker's (i.e. we have lost all the free chips), so the calculated return is 100399*(102500/102573). Similarly the return on betting Banker = 102573*(102500/100399)*95c .

You can see the variance, but at the end of 10 sims the figures still aren't correct, you lose six out of ten playing Bank. I agree that the theoretical numbers should be 99.74% and 100.07% and you can see running a million shoes gets closer to this. While in theory you might be able to make money, in practice you have to wait 2784 shoes to make $7, you'd be better off putting the money in the bank and earning interest.

Hands	Player	Banker	Tie	Player	Banker
				return	return
224 402	100 399	102 573	21 430	100 328	99 484
224 391	100 125	103 215	21 051	99 431	100 380
224 360	100 123	102 787	21 450	99 843	99 966
224 358	99 885	103 263	21 210	99 147	100 668
224 364	100 530	102 594	21 240	100 438	99 374
224 324	100 444	102 397	21 483	100 545	99 268
224 349	100 417	102 571	21 361	100 347	99 464
224 481	100 103	103 170	21 208	99 453	100 358
224 360	100 405	102 632	21 323	100 276	99 535
224 436	100 225	103 017	21 194	99 722	100 088
Totals
2243 825	1002 656	1028 219	212 950	999 531	998 585
One million shoes
80 592 603	35 969 216	36 951 133	7 672 254	99 776	100 033

thanks for your time and effort.

can u help run this simulation to 1 million?

it is after 3P, bet a banker.

i did it in excel, the house edge is lower than betting banker all the time. but i can not run too much data in the excel.

here is the results, i did.

every time is 50000 hands.(i am not doing Tie, because the excel simulation is not so powerful.)

betting banker all the time:
betting banker all the time,(each data is after 50000 hands)
-0.008581
-0.008971
-0.006592
-0.002614
-0.015913
-0.011233
-0.017005
-0.010921
-0.008503
-0.01135
in average:
-0.0101683


betting banker one time after seeing 1 P.
-0.001724551
-0.024842945
-0.010221349
0.003301435
-0.013821656
-0.000545426
-0.016875
-0.016476863
-0.023977575
-0.012380153
in average:
-0.011756408

betting banker one time after seeing 2 P.
-0.006657942
-0.008192363
0.002947883
-0.011227004
-0.007095683
-0.015125776
-0.02342437
-0.004174757
0.003972868
-0.007002915
in average:
-0.007598006

betting banker one time after seeing 3 P.
-0.048732911
-0.000377173
-0.010613682
0.023408864
-0.009320643
0.002029703
-0.007640013
0.003373016
-0.022403462
0.010062686
in average:
-0.006021362
this is what i want u to help test. if PPPP, 1 unit lost, if PPPB, 1 unit won.

thanks in advance.

i could be wrong, because the way i did the simulation in excel is like this: i do random number from 1-10000. i put the number 1-5076 as banker, 5077-10000 as player.

Quote: tomchina123

i could be wrong, because the way i did the simulation in excel is like this: i do random number from 1-10000. i put the number 1-5076 as banker, 5077-10000 as player.

I would use up to about 5068 for the banker. Forget ties. You're not trying to beat ties here.

Quote: tomchina123

....after 3P, bet a banker....house edge is lower than betting banker all the time...

Your test seems to be based on random numbers 1 to 10000, thus previous results technically cannot affect (or should not) the next result. If a real simulation it is possible that three player wins tend to use certain combinations of cards, thus making banker more favourable - I wouldn't know except that Baccarat is usually considered uncountable.

My guess is that you're only using one random number to generate a result and looking for three high ones before making a bet. My feeling is either the random number generator isn't good enough; it may be cycling too quickly so you're seeing repeated result (unlikely) or has an inherent bias. btw https://support.microsoft.com/en-us/kb/828795 hints at limitations in earlier versions of excel - so personally it shouldn't be used for this type of thing.

When I was coding my random number generator and developing the card shuffling routines I initially used simple high-low games as a test. It highlighted a bug in my program which showed up by dealing cards and seeing whether the 1st card was higher than the 2nd card and not getting 50%. Having eventually spotted the bias, one of my later tests was to run Baccarat and confirm it tended to get the correct House Edge.

Finally unless you're running millions of hands (with a good random number generator) results can be put down as luck.

Suppose you're trying a number of simulations to determine the likely (say) House Edge. Mathematically your estimate is the average of all your runs. However the range of your results should be used to determine the how accurate (or close to being correct) your average might be. Typically if they are fairly close, then you can be reasonably confident the average you calculated is near to the answer - if they're not then you can't. Technically you work out the Standard Deviation of your results and use it to create an estimate (I forget how) of the Standard Deviation of your average. So suppose you work out the House Edge for a Blackjack variant at 0.47% and gets result which 95% of the time stay within 0.46-0.48%, then you can be fairly confident. The figures you have seem to vary too much (I haven't worked it out) to prove a bias (i.e. lies outside within 2 or 3SDs).

If you really want to investigate the countability of Baccarat and prove a bias after three Player's, then you'll need to look at Mersenne (or similar), shuffling algorithms and run billions (not millions) of hands - and buy a very fast PC!

Quote: tomchina123

if you were the teacher there, i wouldn't go.

it is just a careless mistake.

to prove the asker is knowing sth, i actually had the answer, i just want to be sure when some one also thinks like this.

in theory, it can.

......

when I say you are not the first one to think of it, you are not the first one to think of it

https://wizardofvegas.com/forum/gambling/tables/20835-baccarat-edge-possible-to-beat-with-rebate/

if it worked, I would be sitting on a truck load of money.

and it is not about how your "simulation" doesn't "fit" the theoretical number. You have got the concept completely wrong.

you have to run the chips twice, so the bonus is in fact only 1/2 of 1.25%, and therefore not enough to cover the house edge.

And why do I have to run the chips twice? thats a question for you to think.

Quote: andysif

...why do I have to run the chips twice?...

My understanding of the original question is you receive 102500 promo chips for $100000 (or similar). You then place each $1 promo chip on Banker and (i) if it is Banker the croupier pays out 95c and leaves the promo chip on the layout (ii) if it is a Tie nothing happens (iii) if it is a Player the chip is removed.

Suppose you then played every permutation of Baccarat hands. There would be 2230518282592250 Player hands, therefore you would need that many promo chips (since you lose one every time there's a Player). In the meantime you would have encountered 2292252566437880 Banker hands.

If you do the maths that means for an outlay of $100000, you get 102500 promo chips and on average encounter 105337 Bankers paying out $100070,06.

Obviously if you change the rules to say you have the play through twice or more then the first iteration of 100070 isn't "cash", is worth less and there's no expected profit.

Quote: charliepatrick

My understanding of the original question is you receive 102500 promo chips for $100000 (or similar). You then place each $1 promo chip on Banker and (i) if it is Banker the croupier pays out 95c and leaves the promo chip on the layout (ii) if it is a Tie nothing happens (iii) if it is a Player the chip is removed.

Suppose you then played every permutation of Baccarat hands. There would be 2230518282592250 Player hands, therefore you would need that many promo chips (since you lose one every time there's a Player). In the meantime you would have encountered 2292252566437880 Banker hands.

If you do the maths that means for an outlay of $100000, you get 102500 promo chips and on average encounter 105337 Bankers paying out $100070,06.

Obviously if you change the rules to say you have the play through twice or more then the first iteration of 100070 isn't "cash", is worth less and there's no expected profit.

that's why i can't stress enough: "get your primary school cert before you go to the casino."

I have pointed out to the OP 1.25% for 10,000 is 10,125, not 12500, and not 10250. And what did he answer me with?

"if you were the teacher there, i wouldn't go. it is just a careless mistake."

And here someone makes the same mistake.

it is a careless mistake.

yes, 10000*1.25%=125. then 100000+125=10125.

anyway, thanks for all answers, i decide to check it by simulation. it is timetaking, but i just want to figure this out.

here it is the bacarrat hands,

https://wizardofodds.com/games/baccarat/simulation/bac-sim-25k-1.txt


Average hands per shoe = 80.884000
Total player wins = 903179, ratio = 0.446654
Total banker wins = 927131, ratio = 0.458499
Total tie wins = 191790, ratio = 0.094847

all hands 2022100
house edge 0.011079843

then i buy 2022100 dollars of chips. the commission is 1.25%:=25276.25 dollar, then i have dead chips of 2047376.25 dollars.

i lose because player wins: 903179. so i have only 1144197 dead chips.
i tie because of tie: i still have 1144197 dead chips
i win because i bet all the banker: then i have cash chip of 927131 bankers*0.95=880774.45 cash chips.

so after 2 million hands,
i invest USD2022100.
i leave: USD880774 cash+ USD 1144197 dead chip=USD 2024971.7

The profit is USD2871.7.

because we have dead chips, we need to do another 2 million, support the hands are the same.i use only dead chip to play.
....
it is too-much time taking, i give it up.

in all, for the 1.25% commission as dead chip, the house edge cutting is neither 1.25% nor 1.25%/2.

it makes money, but it is extremely low profit.

this is by simulation, in reality, it is better than now commission.

Quote: tomchina123

....
it is too-much time taking, i give it up.

in all, for the 1.25% commission as dead chip, the house edge cutting is neither 1.25% nor 1.25%/2.

it makes money, but it is extremely low profit.

this is by simulation, in reality, it is better than now commission.

I can't believe after all these you still think "it makes money" by doing a "simulation".

The other guy had practically done the math for you, except that "careless mistake":

"for an outlay of $101234 (note: corrected by poster = 102500/1.0125), you get 102500 promo chips and on average encounter 105337 Bankers paying out $100070"

for a LOSS of 1164

Let's turn the question round. You need to get at least 2.43% before you show a profit on banker.

Perms	Perms	Results	Payout
Starting fund	102 428.25
Banker	2 292 252 566 437 880	105 263.167	100 000.009
Player	2 230 518 282 592 250	99 669.682	99 669.682