In a negative expectation game like Baccarat you don't get paid fairly when you win, therefore in order to make up for this you must win more hands than probability dictates. However, this can't be done in an essentially random game like baccarat where you can't predict what's going to come out next with any reliability. You certainly can't get an edge by looking at patterns and what's just happened because all patterns are equally likely and past hands don't influence future hands. If follows from this this that there are no "opportunities" to be had. A genuine opportunity can only arise if the probabilities of an event change in your favor and represent "value" (meaning that the probability of a win overcomes the unfair payout).

Since baccarat is like a coin flip in might be easier to understand the principle in those terms. Suppose we agree to play a game of "flip the coin". We take it in turns to call and when I win you pay me $1, but when you win I only pay you 95 cents. Since the probability of either H or T is 0.5, you can surely see that eventually you are going to lose money, UNLESS you can find a way to win more often than 50% of the time (in order to overcome the unfair payout).

How much better than 50% does your probability of winning need to be? If the probability of winning is 50% then your expectation looks like this:

probability of winning * win amount - probability of losing * loss amount

which is 0.5*$0.95 - 0.5*$1 = -$1/40 or 2.5 cents loss per game on average.

Now suppose your probability of winning is not 0.5 but "x". Your probability of losing must therefore be "(1 - x)". With a little algebra you can work out what "x" must be in order to do better than break even.

x*$0.95 - (1 - x)*$1 > 0

$0.95*x - $1 + $1*x > 0

$1.95*x > $1

x > $1/$1.95

x > 20/39 = 0.513

So your probability of winning needs to be at least 51.3% in order to overcome the unfair payout.

Can this be achieved? Actually in coin flipping there is good evidence that the side which shows can be controlled to a certain extent by the flipping technique and the "initial conditions" (the side which is up before you flip), but for casino games it's not so easy.

Unless you can find a way to increase your winning percentage you're just gambling, in which case all I can say is : good luck!

That is a very simplistic way in which to view the game. I haven't witnessed anybody play the game of Baccarat flat betting, fair to assume we all are aware a negative (or positive) progression is required.

A 'smart' friend of mine gave me a bet selection (he asked me not to share) which makes mathematical sense, he tested against both Zumma data sets and it ended positive. I sometimes utilize it when playing solo and am in a position of having to bet every hand (flat betting). Doesn't seem to work when the shoes are what I would describe as rubbish, but I guess they are just those clusters bombs.

Anyhow, we should all appreciate testing is of limited value, even against a million shoes, other than to provide the player with a feel if you like, of average/expected "losses in a row" (not withstanding anything can happen in games of independent trials).

We should also be aware that if we had a large enough data set, then no matter what we test, the result is going to be

*"egalite"*, in other words "equal". I'm under no illusion that if ran my current mode of play against 2^70 (1,180,591,620,717,410,000,000 hands), the expected outcome should be 50/50. Yet that doesn't tell the whole story, if my loss strings are averaging 4LIAR, with the odd 7, then discipline, composure and a risk adverse MM approach will pull you through, if you keep your head and appreciate what is happening is a freak occurrence and you should not expect to see the same for the rest of your session.

If you want to read about a proven edge, that ain't going to happen, not now, not ever, as it simply can not exist relating to betting B or P (excludes the side bets), TT's prove this conclusively.

I will provide an outline on this public forum, shouldn't be too difficult to fathom the rest. If a binary sequence has odds of 1/256 of occurring, what are the odds of that 1/256 sequence repeating in single shoe consisting of 70 hands (approx 9 instances of 1/256). Pretty slim I would hazard a guess, yet have already run into a shoe were this happened 3 times within 70 hands, then you can play 20 shoes and never see a single repeat.

That doesn't tell the whole story. Only a novice would risk 8 bets in order to win a single bet. You need to be a bit savvy about it, there is no substitute than actually being the gaming table, experiencing the rough to smoothwall the rest. At the same time you don't want to be sitting at the table against shoes that present

no bet opportunities, which my own testing has sometimes shown, or only bet once or twice for an entire shoe. Yes, I occasionally test to determine worst case scenarios, so that I can anticipate LIARS, while appreciating if I ran it through a set of binary tables, the result would be

*'egalite'* all things being equal. Which would be perfect in the real world, because anything that returns a 50% strike rate is ideal for a negative progression, alas it is the LIARS which are bane of all systems

People need to focus more on controlling LAIRs, if a binary sequence has an mathematical expectation of occurring once per 2048 B/P decisions (once every 29 shoes). Then what is the expectation of a 1/256 sequence repeating within 70 hands? (please don't go be a novice thinking, oh I'll just bet the opposite, your staking plan won't handle such type pf play).

Acutely aware cards are just innate pieces plastic, have no memory, if you encounter the shoe from hell (already been there done that), then how many losing bets will such an freak event cost you? Once you sort that out, then introduce more bet options, so you are not sitting on your hands for most of the game and make sure your staking plan is up to scratch, because mathematically there can never be an edge only expectation.