It’s very interesting! I used it with my "Aw, Craps!" app and it did pretty well. However, I did get two losing streaks where I’d maxed out the NOB bet ($75/$150) after a series of losses. It has target wins suggested, but not where to stop losing. The part about the trials says they had an average loss of $200 and change, but do you just stop when you lose your max bet, or what?

The author did say that if you’re employing the CPB betting method on a win streak to ride it out until a loss. That’s the answer to your "hit the target during a shooter" question.

Neural networks dramatically changed backgammon strategy. That paper is pretty well-written, too. The one thing I’d say is that for live craps, the bet sizes and Do/Don’t switching would make you look sorta weird. I also think I’d be bored – the author’s suggestion of Bac/mini-Bac is probably the right one.

Sorry to burst your bubble but this is not a method that can win in the long run.

It is mathematically impossible to turn a random game in which the player has a negative advantage into one in which he has a positive advantage. In other words you can not win with any mathematical system in a random game of craps, baccarat, roulette or any other casino game where the house has a positive advantage. The reason being, because the house is paying out less on wins than the true odds require, hence leaving the player with a negative advantage on every wager and decision that is made. Only the odds bet in craps has a zero advantage but you can not make that bet without an accompanying line wager. Since all line wagers are negative advantage wagers for the player, a line wager with odds still has a negative advantage for the player, albeit a smaller negative advantage than any other wager available in craps.

Anyone who understands what a neural network is and how it can be used would never suggest such folly as described in this book. Yes, neural networks can be useful in determining best playing strategy in games such as backgammon, chess or even poker. But that is because those are zero sum games. No player has an advantage or disadvantage over any other player in the game at the start of the game. That is what I mean by zero sum game. So by making decisions based on the best mathematical play at the time one is able to use a skill to defeat one’s opponent in each of those games. For example, neural networks are useful in designing Artificial Intelligence (AI) machines. Google has developed such an AI machine that plays the game of GO, a board game which originated in ancient China more than 2,500 years ago. In January, Google’s AlphaGo AI beat the European Champ Fan Hui. And now, the same AI beat the world’s top player of the last decade, Lee Sedol, after a three-and-a-half hour game. Those are appropriate uses of neural networks.

The same thing is not possible in casino games and specifically never in craps, baccarat or roulette. That is because each wager has a negative advantage. Once again, it is mathematically impossible to turn a random game in which the player has a negative advantage into one in which he has a positive advantage. It is not possible for a neural network to defy the laws of mathematics.

I am not going to go into detail about the flaws in this book with known mathematical laws. But I will cite one simple example in the hope that it is enough to convince folks that this book is not trustworthy and should not be followed.

The author claims on page 9 that there are 6 tricks one can use to gain an advantage over the house. Trick number 5 is the most obvious one that is totally false. It states:

5. You can modify your strategy based on table results and conditions.

Every table develops different trends at different times. Some tables favor bettors who play numbers, while others favor players who bet from the dark side, wagering on the don’t pass or don’t come. Most tables are choppy, favoring neither right nor wrong betters. You can adjust to the changing playing conditions as they occur. If the table is repeating numbers, you can modify your strategy to take advantage of this trend. If the table is ice cold, you can make still different moves. In short, you have the ability to bob and weave, duck and thrust, parry and counter punch. The table can’t react to anything. Every casino table game is like an inanimate object that must endlessly grind out numbers, while you circle and pounce.

In a random game of craps every toss of the dice is an independent event. That is a mathematical term which means, the occurrence of one event does not affect the probability of any other event. In other words, each toss of the dice has the same chance of producing a result of 2 thru 12 as all the previous tosses of the dice regardless of what has been thrown up to that point.

For example, let us say a table has a high preponderance of eights being thrown. Say there have been 10 eights thrown in the past 36 rolls. One would expect only around 5 eights to be tossed in 36 rolls on average in a random game. Just because you have seen twice as many as normal does not mean you can predict that the trend will continue for a bit longer and should put a lot of money on the eight or that in order for the numbers to balance themselves out there will be very few eights thrown in the next 36 rolls and you should lay against the eight for the near future. Even with 10 eights having been thrown in 36 rolls, the chance of throwing an eight on roll 37 and each subsequent roll after that will still and always be 5 chances out of 36. Yes, if you stay at that table for a long time and keep track of every single roll that occurs you will find that the eight will balance out to 5 out of 36 over the long run. But it could take many thousands of rolls before that occurs.

Yes, tables have trends but you can only observe what a trend was in the past. Then and only then can you say what the trend was on the table after it has occurred. It is mathematically impossible for a trend that has occurred to be a predictor of the next toss or trend that will occur in the future.

This is just one simple example to show that this author is not using valid mathematical laws for what he is writing in this book. Neural networking can not be used to develop the world’s most powerful gambling strategy when it comes to random games of chance! The Super Neural Strategy is not mathematically valid.

Terrific post, Skinny! That’s certainly where my suspicions lay, but I had trouble backing it up. I figured I’d find an easy counterexample in play. I also expected to see more on the method used to develop the betting scheme – science usually is happy to expose methods to scrutiny. He wastes all these pages going into betting systems that don’t work, but not how he arrived at his conclusions using the tools he claims to be employing.

Do you think the results he has in the book are doctored? Is 5100 trials just an insignificant sample?

Skinny, thanks for your in depth analysis of this flawed system. We are so lucky to have you on this message board to explain in simply terms why these things don’t work.

Rose and Doc miss you and hope your family is doing well.

Rose and Doc

Skinny, thanks for the reply. I never suggested that this system was a sure thing. I agree with you, no system is a sure thing. I presented the link to start a conversation and I succeeded. 😀

I’ve read many a system over the years, some I like, some I don’t and they all carry a risk of loss.

However, this is the most successful system I’ve read about on the Internet and/or book. Combine this with Dom’s 3 come bet system or your "Big Skinny" could be a powerful combination. 🙂

RFink13, you are certainly welcome. I enjoyed discussing the topic and thank you for that opportunity.

I don’t mean to be argumentative but I am not sure we are on the same page and I want to be very clear on this. Perhaps I am misunderstanding how you feel about it. However I get the impression you do not understand how strongly I am against this system.

You said,

I agree with you, no system is a sure thing.

My statement is much stronger than that. I am saying there is no system that has been or ever will be devised that can beat the random game of craps, baccarat or roulette over the long run.

They do not simply

carry a risk of loss

, they are inherently systems which will eventually lose you money over time without any chance of being a long term winner.

Not only do betting systems fail to beat casino games with a house advantage, they can’t even dent it. Roulette balls and dice simply have no memory. Every spin in roulette and every toss in craps is independent of all past events. In the short run you can fool yourself into thinking a betting system works, by risking a lot to win a little. However, in the long run no betting system can withstand the test of time. The longer you play, the ratio of money lost to money bet will get closer to the expectation for that game.

Furthermore you stated,

this is the most successful system I’ve read about.

I am saying the exact opposite, it is an unsuccessful system, it is "snake oil".

According to the Urban Dictionary, "snake oil salesmen would falsely claim that the potions would cure any ailments. now-a-days it refers to fake products." The author makes untrue statements and bases his system on the very same falsehoods. This is the very definition of deception.

"The Big Skinny" (TBS) and Dom’s 3 come bets are not systems in the sense that neither one of us claims the strategy can beat the random game of craps. In fact we state the exact opposite. They are both betting strategies that a controlled shooter with a proven edge can use to make low house edge wagers.

In my description of TBS I state,

"Skinny" wrote:
None of anything I have said in this post applies to random rollers. I am talking strictly about controlled shooters and mostly only those whose style you know.

The Big Skinny is pretty aggressive. I am putting it out here for those interested in increasing their profits who may be too conservative. But it requires the ability to throw repeating numbers. I have given the pros and cons of this method. Others have discussed their positive and negative experience with it as well. It is not a panacea.

I have stated numerous times TBS is a betting strategy that is aggressive with conservative elements to help control your losses on bad days. I have listed the pros and cons of TBS and advised folks to modify it to suit their tolerance for risk and bankroll. There is no betting strategy that can work under all conditions. But if you are familiar with your own individual skills and tendencies you can match a method to yourself that works for you the majority of the time.

OK, I get it. You said it’s a clunker and it most likely is. All systems are clunkers. I don’t understand what the difference is between a "system" and a "betting strategy". I would consider Dom’s 3 come bet and TBS "systems". Any mechanical predetermined betting plan in my book is a "system".

I understand that all systems in the long run are losers, but to quote John Maynard Keynes, " In the long run we are all dead".

In the short run anything can and will happen. I don’t expect any body nor any system to win all of the time. If I didn’t think I could win every so often I would not even play the game.

My opinion and it may be a small distinction but in my opinion a strategy is simply a method of betting or play.

Uncle John told you in 2003 that he uses and wins with XXX Craps betting. He thinks it is good and you should think about trying it. You decide to give it a shot and use it and tweak it to fit your wants and needs such as bankroll etc. and you make it work well for you over the years. You may even suggest it to your friends because it works so well for you and you want them to win too.

Any plan you take to the table as far as betting is a strategy in my opinion. I think most people think of "system" as something that someone is selling. If you try to sell your strategy for profit then I think it becomes a "system". This is how I think of it in my mind anyway.

"RFink13" wrote: OK, I get it.

That is good to hear. My work here is done 😀

"RFink13" wrote: I don’t understand what the difference is between a "system" and a "betting strategy". I would consider Dom’s 3 come bet and TBS "systems". Any mechanical predetermined betting plan in my book is a "system".

I agree in the macro sense based on the definition of system and strategy. But in the micro sense I was using the terms differently based on the same description that getagrip gave in the post above.

I think most people think of "system" as something that someone is selling. If you try to sell your strategy for profit then I think it becomes a "system".

My point was that we do not profess that our "system" or "betting strategy" can overcome the negative house advantage in a random game. Nor do we claim it is a foolproof means of winning and we certainly are not trying to sell it to anyone.

We are stating that they are ways for a controlled shooter with a proven edge to take advantage of the lowest house edge wagers available. I have also been very clear in stating that TBS is not for everyone. I developed it for myself and it fits well with my own tolerance for risk and bankroll. I have strongly urged others to modify it as they see fit for their own parameters. Once again, I have stated under which conditions it is strong and where its weakness are. It is up to each individual to decide for himself how to utilize the information.

"RFink13" wrote: to quote John Maynard Keynes, " In the long run we are all dead".

Yeah, I am more of a Milton Friedman kind of guy.

“There’s no such thing as a free lunch.”

“Anybody who was easily converted was not worth converting.”
― Milton Friedman

I’m going to do a Wizard of Odds job on the neural networking system. I’m in process of writing a computer program to run it through 10,000,000 sessions and see how well it does.

My gut felling tells me that Skinny is most likely be correct because the math is never wrong, but I’m curious.

RichInMich

Skinny:

Could a neural network figure out the least bad betting method in craps?

"Berto" wrote: Skinny:

Could a neural network figure out the least bad betting method in craps?

I don’t think so. I do not believe a neural network is of any use when it comes to craps because it is a random game in which each successive event is independent of every event that has preceded it.

A very simplistic definition of a neural network is an artificial intelligence system that can speed up the learning ability of the human brain to super human speeds. The fastest brain conceivable can not predict the next roll of the dice because it is completely random.

Computer models such as the one RFink13 is going to build to test this system can be useful in analyzing different betting methods. But those are not neural networks.

Skinny,

You called it, nailed it dead on. I ran a simulation for 100,000 , 1,000,000 and 10,000,000 sessions. The win rates varied based on the target amount as expected. Using the $40 target/$200 buy-in the book used, the win rate was about 78.3% (not quite the 95.4% the book stated). The interesting part is if you take the total loss divided by the total amount bet the answer was about -1.412% (almost the exact house edge on the pass line bet.)

Of course my model, used a random shooter. I’m interested to see how it fares with a skilled shooter, but that’s another project. 🙂

Another project is using the betting progression but betting all passes and all don’ts (with and without a skilled shooter). I want to see if filp-flopping like a politician makes a difference.

RichInMich

🙂 🙂 🙂 😉 😉 😉

I’ll stick with my standard signature which appears just below this.