The fools club
- Thread starter fredperson
- Start date
The nature of proof
I don't undestand why you place so much more faith in simulations than in analytical proof. A simulation could always be faulty in some way; a proof simply is what it is.
If someone comes up with a design for a perpetual motion machine, one could build that machine and demonstrate that it does not work, or if that is impractical to do, one could 'simulate' its operation in some way (perhaps just by a thought experiment) and achieve the same goal. Alternatively, one could explain the laws of physics which mean that perpetual motion is a fundamental impossibility. The latter option is even more attractive if faced with a thousand different designs of such machines.
So it is with gambling systems; rather than simulate each and every variation, you can prove that it's fundamentally impossible to affect the house edge except (in the case of games like blackjack) by taking note of the back-and-forth swings in advantage which occur from round to round.
I imagine this has all been said before, probably better than I am able to, but for what it's worth my understanding of the situation is this -
If a sim remains the only thing that can convince you, then I might be able to cook something up. I have, as a starting point, code that simply simulates flat betting. (I modifiied an existing program which was an open-source blackjack game, giving it the option to omit all of the user-interface stuff and simply play with itself, so to speak.
) I haven't looked at it in a while, and it would be non-trivial to adapt it, but I'll look into it.
If the comps outweigh the extra losses at the table, then that is obviously a good thing, but surely you could achieve the same result by flat betting at a level equivalent to your average bet under OG.
I don't undestand why you place so much more faith in simulations than in analytical proof. A simulation could always be faulty in some way; a proof simply is what it is.
If someone comes up with a design for a perpetual motion machine, one could build that machine and demonstrate that it does not work, or if that is impractical to do, one could 'simulate' its operation in some way (perhaps just by a thought experiment) and achieve the same goal. Alternatively, one could explain the laws of physics which mean that perpetual motion is a fundamental impossibility. The latter option is even more attractive if faced with a thousand different designs of such machines.
So it is with gambling systems; rather than simulate each and every variation, you can prove that it's fundamentally impossible to affect the house edge except (in the case of games like blackjack) by taking note of the back-and-forth swings in advantage which occur from round to round.
I imagine this has all been said before, probably better than I am able to, but for what it's worth my understanding of the situation is this -
We define and calculate house edge (HE) in blackjack as the expected value (EV) of the very first hand that is dealt from a new deck/shoe.
While the actual EV will vary up and down with each subsequent hand, depending on what cards come out, the ups and downs balance out so that it is valid to say that the average EV for each hand (and hence the HE for the game as a whole) is the same as this initial-hand calculation.
[A minor complication to the above is the cut-card effect: If a cut card is used, it produces a slight imbalance in the ups and downs, giving the house a tiny bit more of an edge.]
The net result of all this is that, for any series of hands, with any variation of bet size which is not in some way tied to knowledge of a shift in advantage, the amount you can expect to lose is HE * total amount bet. [Aaargh, another complication : EV is, by convention, defined in terms of the initial bet only (ignoring the extra bets for splits and double downs), so you actually expect to lose HE * total of all the initial bets made.]
Thus - However you choose to chop up your play into sessions or sequences, with stop losses or any other refinement, it is ultimately still just a single series of hands, each of which is subject to the same (average) EV. All you can ever do is trade between risk and reward over a period of time, balancing higher probabilities of being slightly ahead over the period, with larger potential losses when you are not.
While the actual EV will vary up and down with each subsequent hand, depending on what cards come out, the ups and downs balance out so that it is valid to say that the average EV for each hand (and hence the HE for the game as a whole) is the same as this initial-hand calculation.
[A minor complication to the above is the cut-card effect: If a cut card is used, it produces a slight imbalance in the ups and downs, giving the house a tiny bit more of an edge.]
The net result of all this is that, for any series of hands, with any variation of bet size which is not in some way tied to knowledge of a shift in advantage, the amount you can expect to lose is HE * total amount bet. [Aaargh, another complication : EV is, by convention, defined in terms of the initial bet only (ignoring the extra bets for splits and double downs), so you actually expect to lose HE * total of all the initial bets made.]
Thus - However you choose to chop up your play into sessions or sequences, with stop losses or any other refinement, it is ultimately still just a single series of hands, each of which is subject to the same (average) EV. All you can ever do is trade between risk and reward over a period of time, balancing higher probabilities of being slightly ahead over the period, with larger potential losses when you are not.
If a sim remains the only thing that can convince you, then I might be able to cook something up. I have, as a starting point, code that simply simulates flat betting. (I modifiied an existing program which was an open-source blackjack game, giving it the option to omit all of the user-interface stuff and simply play with itself, so to speak.
) I haven't looked at it in a while, and it would be non-trivial to adapt it, but I'll look into it.If the comps outweigh the extra losses at the table, then that is obviously a good thing, but surely you could achieve the same result by flat betting at a level equivalent to your average bet under OG.
Nice to see you back, Fred. Yeah, I'm interested in joining too. (I think you mean it has its own thread here or something?) I believe what you say about having had 20 years of winning with your BJ progressions betting system. I believe that such a system is a possibility. I realize anyone can say whatever on the internet. Hell, I could say I am a clone of Don Schlesinger... But I could only get away with that for a short while until it became evident I was indeed a big liar, after I admitted that math gives me a headache.
Speaking of headaches, Sage, wtf is Richard Reid's math proof doing here on Voodoo? Can you please remove it, over to the Math Boy section where it belongs? I got no idea what it is about as it's totally beyond me. (And I use the term Math Boy affectionately, after Norm recently re-popularized that term here.)
Fred, I cannot go with the name Fools Club. If anyone thinks you (or any member of the FC) is a fool for inventing a gambling system that works, what the hell do you care? And who but a fool would want to join a Fools' Club? So for these reasons, I believe it should be renamed to something in line with the obvious genius of its very limited membership. How about:
1 The WTF? Club. 2 Math Boys Eat Your Heart Out Club. 3 Free Lunches Club. 4 Pitboss = Dumbass Club. 5 We Love S17 Club. 6 Voodoo Rules! Club. 7 Not the Ship of Fools Club. 8 Norm, You Were Wrong! Club. 9 Don't Count Club. 10 Wong = Wrong Club. Just ten off the top of my head. Oh, and 11 Progressive Progressions Club.
We already have something on the table, thanks to CPanther, Shad and Chitown. What they brought already, should be in the WTF? Club, definitely not the other joint.
Speaking of headaches, Sage, wtf is Richard Reid's math proof doing here on Voodoo? Can you please remove it, over to the Math Boy section where it belongs? I got no idea what it is about as it's totally beyond me. (And I use the term Math Boy affectionately, after Norm recently re-popularized that term here.)
Fred, I cannot go with the name Fools Club. If anyone thinks you (or any member of the FC) is a fool for inventing a gambling system that works, what the hell do you care? And who but a fool would want to join a Fools' Club? So for these reasons, I believe it should be renamed to something in line with the obvious genius of its very limited membership. How about:
1 The WTF? Club. 2 Math Boys Eat Your Heart Out Club. 3 Free Lunches Club. 4 Pitboss = Dumbass Club. 5 We Love S17 Club. 6 Voodoo Rules! Club. 7 Not the Ship of Fools Club. 8 Norm, You Were Wrong! Club. 9 Don't Count Club. 10 Wong = Wrong Club. Just ten off the top of my head. Oh, and 11 Progressive Progressions Club.
We already have something on the table, thanks to CPanther, Shad and Chitown. What they brought already, should be in the WTF? Club, definitely not the other joint.
London Colin said:
...
If a sim remains the only thing that can convince you, then I might be able to cook something up. I have, as a starting point, code that simply simulates flat betting. (I modifiied an existing program which was an open-source blackjack game, giving it the option to omit all of the user-interface stuff and simply play with itself, so to speak.) I haven't looked at it in a while, and it would be non-trivial to adapt it, but I'll look into it.
...
If a sim remains the only thing that can convince you, then I might be able to cook something up. I have, as a starting point, code that simply simulates flat betting. (I modifiied an existing program which was an open-source blackjack game, giving it the option to omit all of the user-interface stuff and simply play with itself, so to speak.
) I haven't looked at it in a while, and it would be non-trivial to adapt it, but I'll look into it....
what language is it in?
It's one of the sample programs that comes with Eric Farmer's strategy engine. The plan is to convert it into a library that just represents game-logic, without any user interface elements, which application programmers can then use to code their own games and simulators.
I've heavily modified my local copy of the strategy engine too, adding various features.
But I've tended to work on all this in fits and starts, meaning that the code is not in a great state. There's still lots to do before I would ever embarrass myself by making it public.
The language is C++. I originally downloaded it from bjmath.com.
I've heavily modified my local copy of the strategy engine too, adding various features.
But I've tended to work on all this in fits and starts, meaning that the code is not in a great state. There's still lots to do before I would ever embarrass myself by making it public.
The language is C++. I originally downloaded it from bjmath.com.
If the comps outweigh the extra losses at the table, then that is obviously a good thing, but surely you could achieve the same result by flat betting at a level equivalent to your average bet under OG.[/QUOTE]
First off,thank you for taking the time to discuss this instead of simply repeating the 'Progressions are bad" mantra.
My biggest question would be how does one determine what ones average bet would be in a session of Oscars Grind, as you may start with a $5 bet but will often be betting $25, $30, or $40.
Secondly, if you are flatbetting ,$10, and find yourself $150 down,your chances of overcoming that are nil. Using OG, I come back from that almost every time.
As I have stated repeatedly, I use OG on a game that you can't count on.
Assuming that a flat bet would result in the long term results equalling the house edge( 6 Decks,H17, Surrender, DOA, DAS), then my results using OG are far superior to the expected results. If you want to call it good variance,feel free. I am now at the point where a number of sessions that completely wiped me out would still leave me with a net profit on a negative game,with about a months worth of free rooms and food to boot.
I don't believe flat betting would have gotten the same results.
When I am in Vegas, or even AC, I don't use OG as the games that can be counted and I use a modified version of KISS.
But not every BJ came can be counted. I don't think that means they must be avoided.
First off,thank you for taking the time to discuss this instead of simply repeating the 'Progressions are bad" mantra.
My biggest question would be how does one determine what ones average bet would be in a session of Oscars Grind, as you may start with a $5 bet but will often be betting $25, $30, or $40.
Secondly, if you are flatbetting ,$10, and find yourself $150 down,your chances of overcoming that are nil. Using OG, I come back from that almost every time.
As I have stated repeatedly, I use OG on a game that you can't count on.
Assuming that a flat bet would result in the long term results equalling the house edge( 6 Decks,H17, Surrender, DOA, DAS), then my results using OG are far superior to the expected results. If you want to call it good variance,feel free. I am now at the point where a number of sessions that completely wiped me out would still leave me with a net profit on a negative game,with about a months worth of free rooms and food to boot.
I don't believe flat betting would have gotten the same results.
When I am in Vegas, or even AC, I don't use OG as the games that can be counted and I use a modified version of KISS.
But not every BJ came can be counted. I don't think that means they must be avoided.
I thought I was repeating the mantra, just at greater length.
That's putting the cart before the horse. If you are playing for comps based on the total amount of action you give, you can pick an appropriate flat bet, based on the amount of action you wish to give per hour and the speed at which you can play.
My point was that a flat bet equivalent to your OG average would earn the same amount of comps, not that your OG average represents some kind of ideal amount to bet.
I can only repeat what I and others have been saying. Play for long enough and the house edge must emerge from your results, whatever roller-coaster ride OG provides along the way.
Now an interesting question might be how long is long enough, and how does that compare with the entirety of play that you are likely to get through in a lifetime?
But even if this or any other scheme is able to give a player good odds of being ahead at the end of a lifetime of (presumably infrequent) play, it is still just a case of trading risk and reward - a high probability of ending your career with a small profit, balanced by a small probability of ending it with a big loss.
Whether your results to-date represent good variance, or are in fact well within the normal bounds of what could be expected at this stage, I don't know. (Both because I don't know how many hands you've played, and because I don't have the statistical background to work it out, even if I did have that information.)
Regarding my offer to try to create a sim, I've been looking over some of the previous, occasionally heated debates in which you've participated. You seemed adamant that there is no value in running a sim over many millions of hands, but it really is pointless doing anything else.
Why simulate a weekend's play, when you can simply record your actual results for any given weekend? The whole point of a simulation is to record lots and lots of data, so that it smooths out the variance.
What I thought I might do is run successively longer sims, say -
I though I might also sit a flat-bettor at the same table, to help with the comparison.
How would you react to that? There's no point me doing this if you can't be persuaded that the results are meaningful.
That's putting the cart before the horse. If you are playing for comps based on the total amount of action you give, you can pick an appropriate flat bet, based on the amount of action you wish to give per hour and the speed at which you can play.
My point was that a flat bet equivalent to your OG average would earn the same amount of comps, not that your OG average represents some kind of ideal amount to bet.
I can only repeat what I and others have been saying. Play for long enough and the house edge must emerge from your results, whatever roller-coaster ride OG provides along the way.
Now an interesting question might be how long is long enough, and how does that compare with the entirety of play that you are likely to get through in a lifetime?
But even if this or any other scheme is able to give a player good odds of being ahead at the end of a lifetime of (presumably infrequent) play, it is still just a case of trading risk and reward - a high probability of ending your career with a small profit, balanced by a small probability of ending it with a big loss.
Whether your results to-date represent good variance, or are in fact well within the normal bounds of what could be expected at this stage, I don't know. (Both because I don't know how many hands you've played, and because I don't have the statistical background to work it out, even if I did have that information.)
Regarding my offer to try to create a sim, I've been looking over some of the previous, occasionally heated debates in which you've participated. You seemed adamant that there is no value in running a sim over many millions of hands, but it really is pointless doing anything else.
Why simulate a weekend's play, when you can simply record your actual results for any given weekend? The whole point of a simulation is to record lots and lots of data, so that it smooths out the variance.
What I thought I might do is run successively longer sims, say -
- one of 1 million hands, with the change in bankroll logged every 10,000 hands
- one of 20 million hands, with the change in bankroll logged every 200,000 hands
- one of 400 million hands, with the change in bankroll logged every 4,000,000 hands
I though I might also sit a flat-bettor at the same table, to help with the comparison.
How would you react to that? There's no point me doing this if you can't be persuaded that the results are meaningful.
How would you make a sim using Oscars Grind? Would you be able to program the sim to bet according to OG? There is no point in running a sim that doesn't duplicate it exactly. You can't simply say I'm going to sim 1,000 hands when a winning series occured on the 998th hand and the next two are losing hands. Now, if you sim 10,000 winning series,and then replicate those hands by flat-betting, it would be a fair sim. Can you do this?
If you can do this, I'll send you the rest of the parameters and we'll see if flat betting will actually give better results.
If you can do this, I'll send you the rest of the parameters and we'll see if flat betting will actually give better results.
That is the intention.
If you mean that each logged change in bankroll ought to correspond with the end of a sequence, then I agree that would be ideal but I'm not sure how easy it would be. But over many, many hands it really wouldn't matter if the bankroll was merely tracked every 10,000 (or some other number) hands; the underlying trend would still be clear. However, you are the client in this software project, so if you insist then I'll try my best to do it the hard way. I think the ideal would be to play a fixed number of hands, and then wait for the current sequence to end.
In case there is any confusion, what I hope to show by logging the bankroll approximately every 10,000 (or even more) hands is that the results of a sim of that duration vary too much to be of any meaning (just like in real life), but by effectively conducting a whole host of such mini sims, one after the other, we get to see the trend and the end result.
If you are saying that all you want to see is a single run of 10,000, then that begs the question "which one?". I'd be able to supply a variety, each giving very different answers.
If you mean that each logged change in bankroll ought to correspond with the end of a sequence, then I agree that would be ideal but I'm not sure how easy it would be. But over many, many hands it really wouldn't matter if the bankroll was merely tracked every 10,000 (or some other number) hands; the underlying trend would still be clear. However, you are the client in this software project
, so if you insist then I'll try my best to do it the hard way. I think the ideal would be to play a fixed number of hands, and then wait for the current sequence to end.In case there is any confusion, what I hope to show by logging the bankroll approximately every 10,000 (or even more) hands is that the results of a sim of that duration vary too much to be of any meaning (just like in real life), but by effectively conducting a whole host of such mini sims, one after the other, we get to see the trend and the end result.
If you are saying that all you want to see is a single run of 10,000, then that begs the question "which one?". I'd be able to supply a variety, each giving very different answers.
Although not endorsing his particular method, I agree with shadroch. One factor, among others, that blackjack simulations fail to consider are money management techniques that utilize stop loss and stop win controls. I agree there isn't a non-count system that can win if played out for a million consecutive hands. What the stop controls do is to allow you to leave a bad game thus avoiding a serious downtrend and capitilize on the good games. A winning non-count strategy is superior to a counting strategy. I know this is contrary to counting doctrines because the counters (most of whom are unsuccessful in this venture) have been brainwashed into thinking there is only one way to win over the long run (there was a time when most thought the world was flat). What many fail to acknowledge or understand is that a few players are intelligent enough to overcome the house advantage with methodologies they haven't conceived of or read about. Disagree? then take your best shot. Agree? Don't be afraid to say you do. :devil:
I am not a newcomer and I really resent the cheap shots, most of which seem to be coming from people that rarely actually play the game.. It's a shame so many of you ignore most of what I've been saying and concentrate on the same ****,over and over.
Maybe you can tell me how to count a CSM or a machine that suffles after every hand? It would be great if we could all play on a single deck thats dealt to the last card and be able to spread whatever amounts we wanted.
The reality is that that game doesn't exist so we have to make the best out of the games we have. When in Vegas, I play at the best table my bankroll allows me to, and I count.
When the only game in town is uncountable, one can adapt other methods or one can not play,secure in the knowledge that only by counting can one beat the game. I've chosen the former and the game has been very good to me.
Is that really that hard to get into your closed little minds?
Maybe you can tell me how to count a CSM or a machine that suffles after every hand? It would be great if we could all play on a single deck thats dealt to the last card and be able to spread whatever amounts we wanted.
The reality is that that game doesn't exist so we have to make the best out of the games we have. When in Vegas, I play at the best table my bankroll allows me to, and I count.
When the only game in town is uncountable, one can adapt other methods or one can not play,secure in the knowledge that only by counting can one beat the game. I've chosen the former and the game has been very good to me.
Is that really that hard to get into your closed little minds?
Why do people keep pushing such nonsense. Sims can sim anything. The landing on Mars was simulated. Nuclear explosions are simulated. Why would you think a simple stop loss or stop win can't be simulated. CVData can certainly sims this. And the EV will be the same as flat betting.
See http://www.blackjack-scams.com/html/prog__systems.html
See http://www.blackjack-scams.com/html/prog__systems.html
You start out by complaining of "cheap shots" and then tell people trying to help you of having "closed little minds."
Rejecting voodoo is not an indication of a closed mind, any more than not jumping off a roof because you believe in gravity is an indication of a closed mind. Stop losses, stop wins, progressions were proved over a century ago of having no validity.
Rejecting voodoo is not an indication of a closed mind, any more than not jumping off a roof because you believe in gravity is an indication of a closed mind. Stop losses, stop wins, progressions were proved over a century ago of having no validity.
So how do you explain the original Oscars success? Yes, I know his name was not really Oscar, or the success myself and many other have using his formula?
Everyone writes that progressives will fail because after a series of small wins,you'll have one horrendous loss. Now,if someone could explain how I will experiance a horrible loss when my stop limit is twenty units,I'm more than willing to listen. Two twenty unit losses in a row? Three? Four? Could that happen? Will it? It might,and then I'll have to decide if the money I've previously won and the comps I've earned are worth continuing.
As I've said, I use this mainly to be able to play and stay in one casino. A casino that has been very generous to me.
Now, if Mr London ( or anyone)shows me that flat betting a higher amount would get me the same results, I'm willing to look at the sim. On a table,i'm pretty sure OG would get better comps as pitbosses tend to focus more on larger bets, but I'm playing a machine that only counts money in,so thats a plus for the flat betting side of the debate.
I've been asking for someone to sim OG from about my first month here,and all I get is support from people who actually use it and disdain from people who never tried. The only OG sim I've ever heard of shows you will have a catostrophic loss once every 5,000 series but that is worthless to me as I would be long gone before that,due to my worthless stop limits.
Everyone writes that progressives will fail because after a series of small wins,you'll have one horrendous loss. Now,if someone could explain how I will experiance a horrible loss when my stop limit is twenty units,I'm more than willing to listen. Two twenty unit losses in a row? Three? Four? Could that happen? Will it? It might,and then I'll have to decide if the money I've previously won and the comps I've earned are worth continuing.
As I've said, I use this mainly to be able to play and stay in one casino. A casino that has been very generous to me.
Now, if Mr London ( or anyone)shows me that flat betting a higher amount would get me the same results, I'm willing to look at the sim. On a table,i'm pretty sure OG would get better comps as pitbosses tend to focus more on larger bets, but I'm playing a machine that only counts money in,so thats a plus for the flat betting side of the debate.
I've been asking for someone to sim OG from about my first month here,and all I get is support from people who actually use it and disdain from people who never tried. The only OG sim I've ever heard of shows you will have a catostrophic loss once every 5,000 series but that is worthless to me as I would be long gone before that,due to my worthless stop limits.
Who said it was successful? Why would you believe this? Probably one-third of people that enter a casino will win. That does not in any way suggest that they are using a winning strategy. It is simply standard deviation.
There is NO SUCH THING AS A STOP LOSS. Unless you NEVER play again as long as you live. Changing tables or pausing cannot stop a losing streak. How can anyone believe that you can force a halt to losses simply by pausing or moving?
You will not have to wait 5,000 hands. And, who is to say the big loss will not occur in the first ten minutes? The cards don't count how long you have been playing and change.
Simming progression systems is a waste of time. The answer is always the same. It's like simming 1+1 again and again and again hoping it will come out 3 if you keep trying. People have been asking me to sim progressions for many years. If I do and provide the always same results, then they just say sims are meaningless and usually toss in a stream of insults. I ain't doing it anymore.
There is NO SUCH THING AS A STOP LOSS. Unless you NEVER play again as long as you live. Changing tables or pausing cannot stop a losing streak. How can anyone believe that you can force a halt to losses simply by pausing or moving?
You will not have to wait 5,000 hands. And, who is to say the big loss will not occur in the first ten minutes? The cards don't count how long you have been playing and change.
Simming progression systems is a waste of time. The answer is always the same. It's like simming 1+1 again and again and again hoping it will come out 3 if you keep trying. People have been asking me to sim progressions for many years. If I do and provide the always same results, then they just say sims are meaningless and usually toss in a stream of insults. I ain't doing it anymore.