Baccarat is beatable statistically
Quote: wz60You don't need to flip a coin 200 million times to know that your statistical expectation is 0.5, may be 100000 is enough to converge to a practical close-enough value (say 0.49995) and any more sampling will not make a meaningful difference.
I don't need to flip any coins at all to know that the statistical expectation is 0.5...
You think all these many years AP Mathematicians who do know how to caculate EV (Expected Value) did not check all possible ways to beat Baccarat all these years and you came along and dicovered it?
Of course Baccarat, like Blackjack has memory (unlike roulette that has no memory) and each subsequent hand's probabilities change depending on cards that have passed. BUT unlike Blackjack the difference that past cards make to the change of probabilities is miniscule.
What you suggest, ie a particular pattern of BBP, PPB or whatever other combination you mention of course changes the probabilities of the next hand, but it changes them by a miniscule amount.
If you wanted a better way to find how past cards effect the probabilities, you would not this method BUT the actual cards that have been removed which is a lot more accurate than your method (because it takes into account speficic cards and not Bs and Ps which can be made of different combinations of cards)
You do not like terminology, but I will use some anyway. You can find the Effect of Removal (EOR) of each individual card and devise a Counting system (which is a linear) and find when the Baccarat, Banker or Tie bet become Profitable.
This was first done in 1979 (yes 35 years ago) by Peter Griffin (see the book the Theory of Blackjack) who proved that even with best linear counting system (counting ssytem using 10 different values of each card with count tags like 2,69 4.80 etc) and a very deep cut shoe (ie upto the last 6 cards) then you would probably find one profitable bet per shoe (at the very end) and if you spread like $10 to $1000 , you would make like $0,01 per hour.
Since Griffin (who was the first person to put the BJ Counting into proper formal math and prove mathematically that Counting works), many people checked this combinatorially, by simulation etc and found the exact same results. What do you think AP math guys do when not playing?
Forget linear, Do a perfect Composition dependent Strategy (meaning for each subset of cards at each point). Then you do get a meaningful number as a profit (provided very deep pem to say 6 cards) but still that number is nothing particularly good. And of course you can only do this by taking a computer to the casino. Reduce the pen to say 15 cards and the advantage become very small. (so even a compuer will not help).
You seem to suggest that a combination of BBP (not what you wrote but can be bother checking the specific pattern you mention) which will use maximum 18 cards will change the Banker Bet to advantage. You are completely wrong. Advantages with FULL and perfect play can only be found after taking into account at least 75% of the cards.
Tell me a SINGLE subset of cards after only 18 cards have passed from 8 decks that produces and advantage fof B,P or T.
To help you do this you can use the WoO Baccarat Calculator that can calculate for each specific subset.
https://wizardofodds.com/games/baccarat/calculator/
Quote: AceTwoWz60, you are trying to invent the wheel.
You think all these many years AP Mathematicians who do know how to caculate EV (Expected Value) did not check all possible ways to beat Baccarat all these years and you came along and dicovered it?
Of course Baccarat, like Blackjack has memory (unlike roulette that has no memory) and each subsequent hand's probabilities change depending on cards that have passed. BUT unlike Blackjack the difference that past cards make to the change of probabilities is miniscule
You may be right.
It doesn't hurt to confirm yourself if you can. It is not a bad idea to challenge status-quo, right?
Quote: AxiomOfChoiceI don't need to flip any coins at all to know that the statistical expectation is 0.5...
Your are correct on flip coin.
But to know how BPPPPPB vs BPPPPPP does, you either accept what others told you or you check yourself.
If you check yourself then you need to know what sampling size is large enough to close to the statistical expectation, right?
But before you do that you need to find out the Status Quo.
From your writings, it seems that you did not do a lot of checking of what the Status Quo is.
There is a plethora of reliable information on WoO and other respectable sides about Baccarat and that is not beatable linearly.
And of course there are many mumbo-jumbo websites for how to beat it with patterns etc.
Just a generic piece of advice.
Mainstream games in the US (like Baccarat) have been analysed mathematically by WoO and many others over the years and if there were any advantages to be found they would be already be found (by only mathematical means).
Variations of games, new side bets etc, occasionnaly are found which you can be mathemetically.
Promotions (like rebates, BJ pays 2:1 etc) can be found which you can be mathematcially.
Other tecniques (ie edge sorting) together with math can be used to beat a game.
But Mainstream games, except BJ (without any promotion or comp, or without any variation in the game or without the use of another technique) cannot be beaten.
Quote: wz60Your are correct on flip coin.
But to know how BPPPPPB vs BPPPPPP does, you either accept what others told you or you check yourself.
If you check yourself then you need to know what sampling size is large enough to close to the statistical expectation, right?
As for the effect of removal on the player vs banker bet in baccarat, the math has been done time and time again and independently verified by many people who I trust. The chances of them all having made the same mistake is pretty much zero. I have better things to do with my time than simulate it and confirm what everyone already knows, especially when there are real plays out there that I could be simulating to try to find an edge.
By the way, your approach is terrible. You are looking for likelihood of streaks, in other words, how the winner of the previous hand affects the winner of the next hand. While there may be a tiny (tiny, tiny, tiny) correlation to be found there, it's a second-order effect. You should be looking at the effect of removal of sets of cards. It's clear (no simulation required) that if the composition of two shoes is the same, then the likelihood of player winning the same hand in those two shoes is the same, regardless of how that shoe composition was arrived at.
In other words, you are wasting a lot of time and effort trying to replicate work that has already been done over and over again, and been verified by many people, and you're doing it wrong.
"may be X is enough"sounds like you want to learn some basics of probability and statistics to find out.Quote: wz60You don't need to flip a coin 200 million times to know that your statistical expectation is 0.5, may be 100000 is enough to converge to a practical close-enough value (say 0.49995) and any more sampling will not make a meaningful difference.
may be not.
what you want is the formula to tell you how large a sample needs to be to have a certain confidence level and error.
This is high school level statistics. I learned it in college.
you need to first answer the questions:
how much error do you want to have?
how confident do you want to be in that your sample size does contain the true probability?
my Bac example:
B = success = .506825 (this being a known value. the point estimate many times is NOT known, a reason to simulate)
I want the sample data to have a 99.9% confidence level.
I ain't foolin' around here.
That means 999 out of 1000 samples will contain the true probability. (you can choose to have lower ones. 90% and 95% are very common)
that as a z-score = 3.290526731
sure one can have a smaller z-score and a lower degree of confidence.
Now how much error do you want??
.506825 +/- 10%? (0.10)
meaning you will accept a range of .406825 to .606825
too wide of a range? the sample size is only 270.64
.506825 +/- 1%? (0.010)
still too high?
the formula for minimum sample size = (z_score/error)^2*(p*(1-p))
now you can make your own table according to how much error and confidence you are willing to have.
would love to see it too!
should only take a few minutes
for my example
I want the error to be at most .0001
this range and have only a 1 in 1000 chance my simulation data will be outside that range (1-99.9%)
error+ 0.506 925
error- 0.506 725
This is the range I want. Yes!
This should ask for a very nice sample size.
plug in the values
and we need a sample size of at least 270,638,718.8
not shoes but trials
please check my math here as I was in a hurry
too many trials in the sim?
for some maybe, for others just right
but the error rate is smaller than a smaller z-score and an error rate of greater than .0001
If you want the error to be one decimal place more accurate (.001 instead of .01)
you should find that the sample size needs to be 100 times larger.
"any more sampling will not make a meaningful difference" but it could depending on how much error you are willing to accept.
a coin flip example and you are off on your own
I wants to be only 95% certain with a 1% error that my fair coin flip is 50/50
.49 to .51
9,603.65 = the minimum sample size gets me there
maybe less of an error but keep the same confidence interval of 95% (z-score = 1.959963985)
Times 9,603.65 by 100 should do it.
plug it in, try it out and see
study time, more here if a interested
http://www.mathsisfun.com/data/standard-normal-distribution.html
have fun!
and remember, Asian Baccarat players rule!
just ask Phil
Sally
Quote: Lemieux66I love when there is a new gun in town trying to change the world.
Earlier, we had discussed the quadratic possibility.
What if you squared each number in the count, or maybe just square the actual count. Still rare, but it would be possible to get an advantage with a third of the deck remaining, unless this method doesn't work or the advantage threshold requires similar adjustment. Are there any sims on it?
MustangSally, can't help thinking of the 'wicked pickett' every time I see that handle, great singer, great song.
Quote: evoqueSuffice to say, having looked for an edge to this game for the last 10 years, I have not found one
That explains ALOT. LOL. Well at least you are honest there egalite. The problem/issue is not baccarat itself but YOU. Now this also explains why you arm twisted yours truly to join your crappy Real World Casino forum back in bf, your constant attack on gr8player for not wanting to reveal his play(s) and last but not least your stalking attempts towards Killer and his +50 what on earth steps flat betting scheme. You lambasts him publicly as egalite yet you follow him like some lost puppy under other username(s). Stalking and harassing Archer to the point of calling him and leaving threatening messages on his phone. Plus many others and for what? To obtain/steal their ideas to fill all the problems/issues in your own game which you can most likely solve yourself if you were meant for this which apparently is not the case. Your recent tilt problem in Australia and coming in here begging for help which you didn't receive indicate that is so.
Quote: SonuvabishEarlier, we had discussed the quadratic possibility.
What if you squared each number in the count, or maybe just square the actual count. Still rare, but it would be possible to get an advantage with a third of the deck remaining, unless this method doesn't work or the advantage threshold requires similar adjustment. Are there any sims on it?
I don't follow the logic. What's the point of squaring either of the RC or the tag values? A third of the deck remaining.....means 66% pen? Or they cut off 1/3 of a deck (Ie 17 cards)? How many decks?
No tool, it is called being REALISTIC, there is no edge which helps the player predict which side is going to win the next hand. Therefore you build an approach excluding this naive perceivement, and when you attain a level of enlightenment, guess what [pun intended], it works.Quote: DMSCRThat explains ALOT. LOL. Well at least you are honest there egalite. The problem/issue is not baccarat itself but YOU. Now this also explains why you arm twisted yours truly to join your crappy Real World Casino forum back in bf, your constant attack on gr8player for not wanting to reveal his play(s) and last but not least your stalking attempts towards Killer and his +50 what on earth steps flat betting scheme. You lambasts him publicly as egalite yet you follow him like some lost puppy under other username(s). Stalking and harassing Archer to the point of calling him and leaving threatening messages on his phone. Plus many others and for what? To obtain/steal their ideas to fill all the problems/issues in your own game which you can most likely solve yourself if you were meant for this which apparently is not the case. Your recent tilt problem in Australia and coming in here begging for help which you didn't receive indicate that is so.
As for the rest, you have a vivid imagination to say the least, obviously those bus journeys for a self-proclaimed expert guesser, are having an impact on your psyche.
(eat your heart out).
I am not interested in comprehensive statistic analysis, that is not my purpose. My purpose is to calibrate my cash machine ----- to confirm what I have learned from casino and to find out more that I don’t know. Making money is most important after all.
Well I may be wrong and I may be too lucky for shot time frame, my purpose is to get another modeler to independently verify what I have fund. To verify that:
1 My calculation is correct
2 The data is unbiased
3 The data size is sufficient to give results out of variations.
Regarding #3, without doing statistic analysis I know it by simply build a sensitivity table of shoe size with the results as following:
Shoe Size (k)/B-Win-Chance after 7P:
1 55.55556 %
5 53.05085
10 52.00507
50 51.16397
100 51.11412
150 51.13877
200 51.18468
250 51.10309
Well some may argue that if you run a data of 200 million shoes you will still converge to the status-quo number of 50.67, I don't think so.
Quote: wz60Well I may be wrong and I may be too lucky for shot time frame, my purpose is to get another modeler to independently verify what I have fund.
Your purpose is to make ridiculous, incorrect claims and disagree with anyone who tells you that you are wrong.
You are wrong.
Quote: AxiomOfChoiceYour purpose is to make ridiculous, incorrect claims and disagree with anyone who tells you that you are wrong.
You are wrong.
I already said I MAYBE wrong ---- if you can provide EVIDENCE that betting against 7P has same winning chance as against 5P.
Quote: wz60I already said I MAYBE wrong ---- if you can provide EVIDENCE that betting against 7P has same winning chance as against 5P.
It's probably not exactly the same. There is probably some miniscule 2nd-order effect that you could identify over simulations of billions of shoes. There is no way it overcomes the house edge though.
There is plenty of work out there that shows exactly how often it is possible to make a bet in baccarat that has an edge. It is very, very, very rare, and that is with perfect information about shoe composition, not this streak nonsense.
"There is no way it overcomes the house edge" "There is plenty of work out there that shows exactly how often it is possible to make a bet in baccarat that has an edge" ---- this seems to contradict each other
"It is very, very, very rare" ----- how rare?
It looks that you at least opened your mind a little bit.
Quote: wz60"It's probably" ----- this doesn't strength your argument
"There is no way it overcomes the house edge" "There is plenty of work out there that shows exactly how often it is possible to make a bet in baccarat that has an edge" ---- this seems to contradict each other
"It is very, very, very rare" ----- how rare?
It looks that you at least opened your mind a little bit.
I'm not sure why you expect me to look up these answers for you.
Maybe you should actually read the literature that is out there instead of trying to reinvent the wheel after re-designing it to be square. Your approach to this problem is absolutely terrible.
Quote: AxiomOfChoiceI'm not sure why you expect me to look up these answers for you.
Maybe you should actually read the literature that is out there instead of trying to reinvent the wheel after re-designing it to be square. Your approach to this problem is absolutely terrible.
Yep. Terrible experiment specification. Then he doesn't even want to do the work himself. And, there's no need for this experiment at all since there's years and years and years worth of data and findings from previous well-specified studies to review.
What about a player win (or banker win) would impact the next hand? Is it just the fact that player won? How could that be so? Why could that be so? Is it because the player win included two 7s and an 8? Is P more likely than B to include two 7s and an 8? The cards actually matter. The past hand outcomes do not matter, at least not unless there's a correlation between hand composition and outcome. But even if that correlation exists at a statistically significant level, what is the reason for which we'd want to test based on outcome and not on card removal when card removal is the actual driving force behind shifts in the house edge? When I make a roast, the temperature in my living room goes up. But measuring the temperature in my living room is certainly not the best way to tell if it's almost dinner time.
Because it's like a bad accident, car/train wreck, burning building, choking victim, some crazy crime, no one likes those horrible things, but it's hard not to look. Some people have a heroic mentality and instead of walking away ,they rush into the burning building to help save innocent victims from danger.Quote: wz60If you don’t believe Baccarat at all why bother to read this thread and make comments at all, just leave those who are interested alone.
Bac "systems" players who think they can beat, or have found new math and or voodoo to beat baccarat, are at best dangerous to innocent people. If the math Hero's just let the evil system players run wild on the forums, the casualties would be great. (Or, they just like to argue with people who are full of crap and they love to prove them wrong)
The owner of this site has made it very clear he does not want "betting systems" brought to this site(they can't and they wont work) I believe he wishes people find a different home for discussing ways to beat the unbeatable. Why not respect that? Especially when other sites will welcome you with open arms? If you have simple math question or think you found something, Why not ask your question, once you have your answer, accept it and move on?
FYI: no one is interested, other then guys that are not interested. If not for the the Bac system haters, you would be talking to yourself. Anyone being nice is just humoring you.
For example, we know that if you flip a coin the chance of landing on heads (H) is 50%. We know if we flip a coin twice, the chance it will land on H then H is 0.25 or 25%. However, if the previous flip was H, we cannot say the next flip has a 25% of being H just because the chance of flipping HH is 25%.
I don't think that is precisely what's going on here....but I think it's rather similar.
There are some Baccarat haters indeed, one has to wonder where this came from...
Quote: wz60This would be my last response to some:
There are some Baccarat haters indeed, one has to wonder where this came from...
As long as you know you're gonna lose, theres nothing wrong with playing
Hopefully its your last response concerning Baccarat to anyone, if you are trying to insinuate there is a way to beat Baccarat other then what has already been proven. Why don't you read the Gr8player saga? It was made clear to him NONSENSE Bac talk is frowned upon and not welcome. I just don't understand why anyone would want to stay at a place they feel not welcome?Quote: wz60This would be my last response to some:
There are some Baccarat haters indeed, one has to wonder where this came from...
Quote: wz60Well I may be wrong and I may be too lucky for shot time frame, my purpose is to get another modeler to independently verify what I have fund. To verify that:
1 My calculation is correct
2 The data is unbiased
3 The data size is sufficient to give results out of variations.
Regarding #3, without doing statistic analysis I know it by simply build a sensitivity table of shoe size with the results as following:
Shoe Size (k)/B-Win-Chance after 7P:
1 55.55556 %
5 53.05085
10 52.00507
50 51.16397
100 51.11412
150 51.13877
200 51.18468
250 51.10309
Well some may argue that if you run a data of 200 million shoes you will still converge to the status-quo number of 50.67, I don't think so.
First of all, notice that the percentage is dropping below what is required to overcome the 5% commission, which, if I remember correctly, was your original claim.
Second, how many groups of 1000 shoes did you look at to come up with your 55.55556 number? One?
While there could be some merit in the claim that "betting B after 7P has a better chance of winning than betting B after 5P" (I don't think it's absolutely provable in an 8-deck game because of the sheer number of permutations of cards), neither one makes baccarat a player's advantage game. Besides, you have the disadvantage of having to wait until a sequence of BPPPPPPP occurs before you can bet. How often does that actually happen?
Quote: ThatDonGuyFirst of all, notice that the percentage is dropping below what is required to overcome the 5% commission, which, if I remember correctly, was your original claim.
Second, how many groups of 1000 shoes did you look at to come up with your 55.55556 number? One?
While there could be some merit in the claim that "betting B after 7P has a better chance of winning than betting B after 5P" (I don't think it's absolutely provable in an 8-deck game because of the sheer number of permutations of cards), neither one makes baccarat a player's advantage game. Besides, you have the disadvantage of having to wait until a sequence of BPPPPPPP occurs before you can bet. How often does that actually happen?
Let me make this clear: I am willing to admit that I am wrong if anyone can prove that I am wrong.
Then lets pursue things step by step.
I thought that you were willing to run simulations, using same 250k data I used, to find out winning chances of betting against 7P (BPPPPPPPB vs BPPPPPPPP),
you can use whatever calculation method you think is appropriate to derive your results. In this way we use the same base data and run the same simple pattern so we will be talking about same thing.
I use this particular simple pattern for the sake of discussion and that most reader can understand it easily, and I never said that it gives a winning chance to overcome 5% house commision. Initially I proposed to use BPB vs BPP as a test example but I think 7P will provide more clear difference.
Are you still willing to do it?
Quote: wz60Let me make this clear: I am willing to admit that I am wrong if anyone can prove that I am wrong.
Then lets pursue things step by step.
I thought that you were willing to run simulations, using same 250k data I used, to find out winning chances of betting against 7P (BPPPPPPPB vs BPPPPPPPP),
you can use whatever calculation method you think is appropriate to derive your results. In this way we use the same base data and run the same simple pattern so we will be talking about same thing.
I use this particular simple pattern for the sake of discussion and that most reader can understand it easily, and I never said that it gives a winning chance to overcome 5% house commision. Initially I proposed to use BPB vs BPP as a test example but I think 7P will provide more clear difference.
Are you still willing to do it?
Where's the data set? If I can figure out how to analyze it then I will.
Well how thoughtful of you. I'm willing to admit there are no aliens as long as someone can prove there are none.Quote: wz60Let me make this clear: I am willing to admit that I am wrong if anyone can prove that I am wrong.
Quote: RSWhere's the data set? If I can figure out how to analyze it then I will.
See link in p4. Somewhere you also can find my testing code.
I and others have done so repeatedly, each time you shift your claim like a wack-a-mole.Quote: wz60Let me make this clear: I am willing to admit that I am wrong if anyone can prove that I am wrong.
wz60, this was your first post in this thread. You clearly claimed to be able to beat the 5% commission. I have added empasis to make this clear.Quote: wz60I use this particular simple pattern for the sake of discussion and that most reader can understand it easily, and I never said that it gives a winning chance to overcome 5% house commision.
Your methodology is poor. You take a set of shoes and then mine it for patterns. You will find patterns in any data. However in random data (like that of baccarat) those patterns have no useful predictive power outside of the data you discovered them in. Read up on the theory of cross validation.Quote: wz60I want to share my findings and I want anyone who can do modeling to challenge me.
First, lets ignore ties, meaning to exclude ties in the calculation.
With ties ignored, B has an edge of 50.68%, this still leads to a defeat for players since we need 51.28% to overcome the 5% commissions.
However we should note that this 50.68% edge is a flat overall number, can some bet selections provide higher edge and even >51.28 so a statistically win can be achieved?
The answer is yes and some selections do provide much better edges. Here are some examples:
1 You should not bet B after B --- continuation is always bad bet in Baccarat statistically, although each one has slightly different statistical expectation, e.g. bet 5B to continue is much better than 6B to continue (50.18 vs 49.28).
2 As you can see from above example, we actually have an edge even bet P after 6B ---- 100-49.28=50.72
3 Lets back to bet B strategy, here is a winning one: bet B every time after first P appears, this will give us a 51.41% edge, it is a winner since it is >51.28.
If you bet B after 2P, you will get a 51.24% edge
If you bet B after 3P, you will get 51.27% ---- very close to even
If you bet B after 8P, you have a big edge at 52.5%
I am sure that there are other B-betting selections that will also provide winning statistics.
The bottom line is: the widely recognized 1.06% house edge for Baccarat is very misleading compared with BJ ----- BJ’s number is after applying some strategies while the Baccarat 1.06% number is flat overall number. It can be reduced significantly by some good betting selections.
There is dependent strategy in baccarat, however it is far weaker than you seem to believe. Here is a good article on such work. B vs P is not predictive as I showed you on page 4-6 when you tried to shift your claims the first time.
I'm still waiting on your reply to my offer for you to actually flat bet this stuff over many hands.
Quote: endermikeI and others have done so repeatedly, each time you shift your claim like a wack-a-mole.
wz60, this was your first post in this thread. You clearly claimed to be able to beat the 5% commission. I have added empasis to make this clear.Your methodology is poor. You take a set of shoes and then mine it for patterns. You will find patterns in any data. However in random data (like that of baccarat) those patterns have no useful predictive power outside of the data you discovered them in. Read up on the theory of cross validation.Quote: wz60I want to share my findings and I want anyone who can do modeling to challenge me.
First, lets ignore ties, meaning to exclude ties in the calculation.
With ties ignored, B has an edge of 50.68%, this still leads to a defeat for players since we need 51.28% to overcome the 5% commissions.
However we should note that this 50.68% edge is a flat overall number, can some bet selections provide higher edge and even >51.28 so a statistically win can be achieved?
The answer is yes and some selections do provide much better edges. Here are some examples:
1 You should not bet B after B --- continuation is always bad bet in Baccarat statistically, although each one has slightly different statistical expectation, e.g. bet 5B to continue is much better than 6B to continue (50.18 vs 49.28).
2 As you can see from above example, we actually have an edge even bet P after 6B ---- 100-49.28=50.72
3 Lets back to bet B strategy, here is a winning one: bet B every time after first P appears, this will give us a 51.41% edge, it is a winner since it is >51.28.
If you bet B after 2P, you will get a 51.24% edge
If you bet B after 3P, you will get 51.27% ---- very close to even
If you bet B after 8P, you have a big edge at 52.5%
I am sure that there are other B-betting selections that will also provide winning statistics.
The bottom line is: the widely recognized 1.06% house edge for Baccarat is very misleading compared with BJ ----- BJ’s number is after applying some strategies while the Baccarat 1.06% number is flat overall number. It can be reduced significantly by some good betting selections.
There is dependent strategy in baccarat, however it is far weaker than you seem to believe. Here is a good article on such work. B vs P is not predictive as I showed you on page 4-6 when you tried to shift your claims the first time.
I'm still waiting on your reply to my offer for you to actually flat bet this stuff over many hands.
You need to also quote what I said after the first post:
"I don’t claim that my findings are absolutely true since my sampling is limited (10000 shoes) and my algorithm may not be 100% correct although I am pretty sure they are correct. My purpose is to get those who CAN do modeling to confirm or invalid my findings."
I quoted your first post as it was. Then we did model it. We showed it was wrong. You admit that now?Quote: wz60You need to also quote what I said after the first post:
"I don’t claim that my findings are absolutely true since my sampling is limited (10000 shoes) and my algorithm may not be 100% correct although I am pretty sure they are correct. My purpose is to get those who CAN do modeling to confirm or invalid my findings."
You then tried to defend your nonsense saying:
Then we did model it. We showed it was wrong. You admit that now?Quote: wz60, in a different postAs a world-class expert in one of engineering modeling fields with >30 years experience I don’t feel a need to argue with you about terminology.
I don’t claim that my findings are absolutely true since my sampling is limited (10000 shoes) and my algorithm may not be 100% correct although I am pretty sure they are correct. My purpose is to get those who CAN do modeling to confirm or invalid my findings.
Your methodology is terrible. Please tell us what companies your >30 years of engineering modeling have contributed to so we can take precautions.
Quote: endermikeThen we did model it. We showed it was wrong. You admit that now?
I said lets go step by step and start with the 250k data.
I presented my results for betting against 7P, what is your results?
I told you already, you wouldn't get anymore for free. If you are willing to bet it I will take your action.Quote: wz60I said lets go step by step and start with the 250k data.
I presented my results for betting against 7P, what is your results?
BTW: I was you one who linked you to that data, you are welcome.
Quote: endermikeI told you already, you wouldn't get anymore for free. If you are willing to bet it I will take your action.
Well then lets wait others to check this data. You cannot convince me without presenting your results for same data I used.
This is your problem. You won't use data outside of where you got your initial results. Have you ever heard of cross validation?Quote: wz60Well then lets wait others to check this data. You cannot convince me without presenting your results for same data I used.
Quote: endermikeI told you already, you wouldn't get anymore for free. If you are willing to bet it I will take your action.
BTW: I was you one who linked you to that data, you are welcome.
I think I already thanked you for providing that data link.
1) pay me for the work (probably about $50)
2) make your signature line:
"I am a world-class expert in one of engineering modeling fields with >30 years experience. However I don't understand cross validation or how to properly mine data."
You did, I'm just trying to remind you that I am arguing for truth and mathematical reality, nothing more.Quote: wz60I think I already thanked you for providing that data link.
Quote: endermikeThis is your problem. You won't use data outside of where you got your initial results. Have you ever heard of cross validation?
Well we can get into that later, but lets just finish with first set of data and see what we got from it ---- see if there are clear difference between betting againt 5P and 7P, does that make sense?
I am willing to run this but you need to state that you understand statistical significance and practical significance.Quote: wz60Well we can get into that later, but lets just finish with first set of data and see what we got from it ---- see if there are clear difference between betting againt 5P and 7P, does that make sense?
I suspect that with just a little more effort, you could use a similar technique and get perfect results for your set of data on Baccarat shoes, or at least a lot better results than relying on encountering a sequence of BPPPPPPP or some such to determine what to bet.
Or you might consider trying your strategy yourself on a completely different set of data. I don't want to put a damper on your enthusiasm, but things didn't work out so well when I tried that with my "perfect" strategy. Then, I only won about half of the flips.
I mostly trust the result he got from the original 250,000 hands of data (or whatever it was). It's not hard to do that sort of analysis. The question should not be 'is this valid' but 'so what does that tell me?'.
Quote: endermikeI am willing to run this but you need to state that you understand statistical significance and practical significance.
Thanks for the offer and of cause I know the difference.
Let me remind you that you need to filter out all T's before simulation. Just like that a player sits in Baccarat table and ignores all ties.
So to be clear: you want me to parse the data from simulations 1-10 on WoO to confirm that betting B after a sequence of P results has distinctive results?Quote: wz60Thanks for the offer and of cause I know the difference.
Let me remind you that you need to filter out all T's before simulation. Just like that a player sits in Baccarat table and ignores all ties.
Quote: wz60Let me make this clear: I am willing to admit that I am wrong if anyone can prove that I am wrong.
Then lets pursue things step by step.
I thought that you were willing to run simulations, using same 250k data I used, to find out winning chances of betting against 7P (BPPPPPPPB vs BPPPPPPPP),
you can use whatever calculation method you think is appropriate to derive your results. In this way we use the same base data and run the same simple pattern so we will be talking about same thing.
I said what now?
Here's what I said when you asked me to use your data set:
Quote: ThatDonGuyNot at the moment, but I did run blocks of 250,000 random shoes, and after "only" 40 blocks, on B5P to stop, the bank percentage ranged from 50.4486 to 50.9663 (which includes the 50.62% you got), while on B7P to stop, the bank percentage ran from 50.3591 to 51.1366 (which includes the 51.10 that you got).
Quote: wz60I use this particular simple pattern for the sake of discussion and that most reader can understand it easily, and I never said that it gives a winning chance to overcome 5% house commision. Initially I proposed to use BPB vs BPP as a test example but I think 7P will provide more clear difference.
Are you still willing to do it?
I could, but why bother? Why would my result be any different than the one you came up with? Nobody is questioning the numbers you came up with - just that your one result somehow applies to baccarat as a whole.
What exactly am I supposed to be trying to determine? My 40 runs of 250K shoes resulted in a range of percentages for both the 5P and 7P cases that included your 250K-shoe results, but in none of those cases was there a percentage either way that would have made any money.
Also, the next time you want to claim that you never said that baccarat was beatable statistically, you might want to make sure it's not the subject of the thread you start first. (Although technically (and this has already been pointed out), it probably is beatable statistically, if you're able to get a complete running count of all cards played from a shoe up to that point - and note that Nevada gaming regulations specifically allow players to do this!)
One other thing - you asked me if I could do this:
Quote: wz60Can you provide a table (or better a graph) of the results for sampling size of:
10000
20000
50000
100000
.......
250000
500000
1000000
1500000
2000000
I am still unclear what you want me to do (and please be specific).
Quote: endermikeSo to be clear: you want me to parse the data from simulations 1-10 on WoO to confirm that betting B after a sequence of P results has distinctive results?
This is also the answer to ThatDonGuy:
Doing the test from a player's perspective:
You are going to a Casino and you are going to do Baccarat for 250k shoes, shoe by shoe. Ignoring all ties, you start from hand #9 to the end of the shoe. When you see the previous 8 hands are BPPPPPPP then you bet B. You win if it turns out to be B and you loss if P. You record all your bets and the results of wins and losses then you have the result for all 250k shoes.
Does this make sense?
