help needed for math and analysis of a baccarat game system

hi all,

after reading information on baccarat games, some questions come to my mind but i am not mathematical savvy to answer my own questions thus the post. Hope someone can help. Thanks in advance.

With the facts

Banker wins - 0.458597
Player wins - 0.446247
Tie - 0.095156

for banker to win
3 time in a row - 0.06420894
4 time in a row - 0.03254347
5 time in a row - 0.01649424
6 time in a row - 0.00835989

for player to win
3 time in a row - 0.06079105
4 time in a row - 0.02997989
5 time in a row - 0.01478497
6 time in a row - 0.0072914

during the 1st 5 to 10 games, if i look for uneven display of % for banker, player, tie, i assume that the winning % of the bet type will try to balance it out. for e.g. it open B,T,P,B,P,T,P,P,P for 1st 10 games, this will create a ratio (B:P:T) 1:3:1 which is B - 0.2, P - 0.6, T - 0.2

i will bet on B because 0.2 is not anywhere near the winning % of 0.458597. i also understand this is too small of a sample size, but if i keep looking for such off balance trend and bet on the bet type which has less winning %. over 1k or 10k shoes, eventually 0.2 will get nearer to 0.458597 meaning i will win 0.258597 of the time.

my strategy

1. to look for any trend that created either player or banker <0.3 winning percentage during the 1st 5-10 game and bet the lesser winning percentage and monitor the winning percentage. i will stop play after game 20, i feel by next 10-15 games, if the shoe does not balance back. since there will be imbalance shoe after all (like player more than banker) after end of shoe. i shall look for other opportunity
2. i will stop loss at 6 unit since that less likely to happen <0.01, and if it happen i will take it as my loss.
3. i will bet with 1 unit and double on 3rd bet if i win the 1st 2 and continue until i lose a game, after i shall look for other opportunity
4. with line 3 progress winning betting by the time it balance, i will either
a. win 0 unit win 2 and lose 3rd game
b. win 2 unit win 3 and lose 4th game
c. win > 2 unit if catch on a continuous trend until lose.

i like to know
1. if my math assumption on winning 0.258597 of the time if banker is only 0.2 out of 10 games.
2. if my strategy could work?
3. if my strategy could work, what is the variance of this?
4. how to or how much unit as capital to overcome the variance

Quote: tea

i assume that the winning % of the bet type will try to balance it out.

Why would you make that assumption?

I'll try and help.... If you flipped a coin and it came up heads 7 times and tails 3 times, do you think over the next 10 flips "it will try and balance out" with 7 tails and 3 heads? The phrase people use is "the cards have no memory". Meaning each and every baccarat hand is a new event, and each and every hand has a built in edge for the house, either by losing more often than you win (player), or getting paid less when you win (banker).

hi soopoo,

i understand every result is almost 50/50, but looking at the math statistic, unless i read and understand something wrong.

let assume we want to play 100,000 hands and we know that

Banker wins - 0.458597
Player wins - 0.446247
Tie - 0.095156

so when we look at 1st 10 game with the following ratio, even with 100s, or 1000s of hand with this starting winning percentage.

Banker wins - 0.2
Player wins - 0.6
Tie - 0.2

and i keep looking for such off balance stats, as i play more hands to 100s, 1000s, we should be getting stats nearer and near to the 100,000 hands stats isn't it? meaning banker got to win 0.258597 more in order to be able to reach 0.458597.

Quote: tea

and i keep looking for such off balance stats, as i play more hands to 100s, 1000s, we should be getting stats nearer and near to the 100,000 hands stats isn't it? meaning banker got to win 0.258597 more in order to be able to reach 0.458597.

No, that is literally what the Gambler's Fallacy says. The truth is more subtle: if you play for 10 hands, and you have "off balance stats", and then you play for another 990 hands that yield the expected distribution, the overall 1000 hands actually do yield "stats nearer and nearer" to the expected ratios.

Do the math and see for yourself.

The trick is to realize that each hand is a very large factor in a small sample size, but only a tiny factor in a very large sample. If you play exactly two hands, you obviously can't yield the expected ratios. The probability of Red in American roulette is 18/38, but you can't observe that in one or two spins. You might, over 38 spins. But -- and here's the clincher -- it is far more likely to see a ratio of exactly 18/38 after 38 spins than it is after 380 or 3800 spins. You will be *closer* in terms of the ratio after more spins, but you will actually be *farther away* in terms of individual spins. Suppose you spin 3800 times and see 1780 heads. That's 20 spins less than expected and a ratio of 1780/3800. The ratio is pretty close and in context being off by 20 is no big deal. But if you only spun 38 times, being off by 20 would be far more unlikely.

You should expect the raw difference between actual wins and expected wins to *grow*, not shrink, as you play more hands. But the ratio of actual to expected should shrink. Once you understand how that's possible then you'll understand why the Gambler's Fallacy is wrong.

While there may be ways indeed to beat Baccarat... Betting systems is NOT ever going to be one of them.

Why not apply the same principle to roulette? If you see 10 blacks in a row, start betting on red because it's got to come up to "level" the stats out, right?

hi MathExtremist,

i wish i can do the math, i think i need to read yr post a few more time to digest the information. i will read on Gambler's Fallacy.

tyvm.

Quote: tea

unless i read and understand something wrong.

You did. My last try to help you...... Answer these questions honestly

1. If you have flipped a coin 3 times, and it has been heads twice and tails once, what are the odds that the fourth one will be tails?

2. If you have flipped a coin 19 times and 12 heads and 7 tails, what are the odds that the 20th will be tails?

3. If you played baccarat 3 hands and twice has been player, and once banker, what are the odds that the fourth will be player?

4. If you played baccarat 19 times and 12 player and 7 banker, what are the odds that the 20th will be player?

Now, after thinking about the answers for a few minutes, substitute ANY numbers for the "19" or "12" or "7" in questions 2 and 4. Will the answer be anything different?

hi Romes,

While there may be ways indeed to beat Baccarat - care to introduce those that u are aware of?

but so far u have a point comparing player banker to red black of roulette, i see it more clearer since it a ball and and wheel, every spin is a result of its own.

Quote: tea

hi Romes,

While there may be ways indeed to beat Baccarat - care to introduce those that u are aware of?...

If everyone did this then there wouldn't be many ways to beat it anymore =P. Good try though =).

lol, ok no problem, no harm trying. thank you

it all going to be 50/50, a result of its own.

Quote: tea

it all going to be 50/50, a result of its own.

Same with baccarat.