|
|
#21
07-22-2006, 08:12 PM
|
|||
|
|||
Re: Blackjack Strategy
[ QUOTE ]
Is there any more of an indisputable fact in the world that so many people refuse to understand or believe? No non-counting betting strategy does anything to affect your win rate, period. It is a fact. Yet practically no one in the world believes it to be true. Man the casinos are good at marketing. [/ QUOTE ] I think if you doubled every hand that would be a very good way to your win rate. Not increase, just affect (negatively). 
|
|
#22
07-22-2006, 08:29 PM
|
|||
|
|||
|
Re: Blackjack Strategy
[ QUOTE ]
If you had unlimited money and the casino was willing to accept bet of any size, your system would in fact work. [/ QUOTE ] No, it wouldn't. This has been discussed in-depth in the probability forum.
|
|
#23
07-22-2006, 08:32 PM
|
|||
|
|||
|
Re: Blackjack Strategy
[ QUOTE ]
[ QUOTE ] the strategy above is just as flawed as any martingale system. Every bet has -EV. No batting strategy will allow you to beat the game. (Without counting of course). [/ QUOTE ] It's not designed to be a flawless system... but it's the best way to manage your money when playing blackjack. It will give you the most return on your money if you hit a hot streak, and will give you the least amount of loss when you hit a bad streak. The EXACT opposite is true for the Martingale. It is a very profitable way to bet IMO, as it gives you a chance to make some real money. I see thousands of hands played out a day, trust me [img]/images/graemlins/tongue.gif[/img] [/ QUOTE ] The best way to manage your money gambling and to last as long as possible is simple. Bet as little as possible each hand. Seriously, there is no "system" better than that.
|
|
#25
07-22-2006, 11:15 PM
|
|||
|
|||
|
Re: Blackjack Strategy
In ANY gambling game your long-term result win (or loss) will = Total of All Your Bets * Your (or House) Advantage. Shorter: Win = TotalBets * Advantage.
By definition, as rearranging this formula gives the definition of "Advantage" (Advantage = Win/TotalBets). If you bet a total of $1000, or $1,000,000, or any other amount, at a game where you have a -1% advantage, then you will, on average, lose 1% of it (doh - by definition!) The exact sequence of your bets which contribute to your TotalBets (the $1,000 or $1,000,000 or whatever) does not alter the Total (with another "doh!"). The total is the total, regardless whether it is arrived add by adding up bets which are increased after a win, or after a loss, and/or by how much you increase them (I think I learned this school when I was about 8 or 9, a+b = b+a, the total is the same regardless). When you play blackjack the cards you get, and the decision choices available to you, are not affected by the size of your bet. In other words, regardless of whether you bet small or big on a hand, the outcome will be the same. Thus the Advantage does not change (in simpler terms: your chance of winning or losing a hand is not affected by the size of your bet). So we have established that in the formula: Win = TotalBets * Advantage that altering the sequence of your bets affects neither the TotalBets nor Advantage values. So, surprise, Win won't be affected either. If you bet $x at a y% advantage game, you will, in the long-run, get a result of $x * y%. This is the definition of "Advantage" AND THERE IS NOTHING YOU CAN DO ABOUT THIS (unless you are capable of changing one of the laws of mathematics. If so you are a God, and probably have better things to do than play blackjack). The ONLY thing you can control, and only in the SHORT TERM - is the distribution of results. If you went to a casino and played a martingale system, and left as soon as you were at least $1 up, then you would walk away the winner something like 85%-90% of your visits (exactly how often depends mostly on the number of double-ups permitted by the maximum:minimum bet ratio available to you). Of course, a whole bunch of $1 wins, is somewhat ruined by a $1023 loss! And the sad news is that stopping gambling one day, and restarting the next, is mathematically indentical to playing non-stop. Win=TotalBet*Advantage regardless of when you go home to sleep. Here's an experiment: grab a coin and toss it a whole lot of times (a few hundred would be good). Record whether you toss a Head or Tail. (You can do this easily in Excel) Treat each Head as a win, and each Tail as a loss. All bets are a flat $1. After a series of bets (maybe only one long) has reached a $1 profit "go home" (ie. record that you had a winning session). Start the next session (from the next coin toss) as a new session. You'll eventually win EVERY session (because, sooner or later, there will be a run of luck sufficient to get you to $1 up). It might take a VERY long time (hundreds, thousands, even millions of coin tosses), but it is mathematically certain that eventually ever session will reach +$1 (such is the nature of a "random walk") Now - WITH EXACTLY THE SAME SERIES OF COIN TOSSES - repeat the accounting exercise, this time scoring a TAIL as a $1 WIN, and a Head as a $1 loss (ie. reverse them). You will the identical result, ie. you will every single session (sooner or later). So, regardless of whether Heads are winning bets, or Tails are winning bets, you can win EVERY SINGLE SESSION. Surely, if a series of coin tosses results in winning sessions with Heads being good, it must result in losing sessions with Heads being bad? Nope. Regardless, every session will be a winning one (actually the last session is probably still "in progress", but if you'd tossed enough coins you could still have hundreds of winning sessions and only one still unfinished, and that might only be down a few dollars, ie. it's not signficant). This apparent paradox isn't real. The reality is that your sample contains a fixed number of Heads and Tails (eg. 346 Heads, 334 Tails). The difference between these two numbers (eg. 12) is how much you won or lost. Dividing the sample into "sessions" does not alter the total. I will repeat this - it is important: The quantity of winning sessions does not affect the total win (or loss). Now let's introduce a Margingale. If you worked through the above experiment (or just thought about it carefully), you'll have seen that a losing session can take a LONG time to get back up to $1. Sometimes only a handful of coin tosses is required, sometimes dozens, sometimes hundreds. If you were keen enough to do a REALLY large test, some sessions that went south early might take thousands of coin tosses to eventually come back up. There is a very quick and easy way to get back to a $1 profit - Martingale. After a loss, double your bet. Now it only takes ONE SINGLE WIN to get to +$1, regardless of how many losses preceeded it. Win = +$1 Loss, Win = -$1 + $2 = +$1 Loss, Loss, Win = -$1 + -$2 + $4 = +$1 Loss, Loss, Loss, Win = -$1 + -$2 + -$4 + $8 = +$1 etc. Boy! Imagine that, rather than having to toss hundreds or even thousands of coins to slowly inch our loss back up to +$1, a dollar at a time, we can do it with one single winning toss. It is, in fact, impossible to get there any faster. Go back to the SAME series of coin tosses, and with Heads being a win, and doubling up after a loss. What is the effect on your results? If you work your way through the first few sessions you see that the sessions are VERY much shorter than the flat betting approach, because all it takes is one single Head to win. You will win every single session (because, sooner or later, usually sooner, you'll throw a Head), and every session wins you a net $1. Each and every Head means you've won $1, regardless of how many Tails are in front of it. So count the number of Heads in your sample, that how many dollars up you are. Now do exactly the same thing for Tails being the winning bet. Each and every Tail means you've won a $1. The number of tails in your sample is the total you've won. [The last session might still be in progress, but assuming your sample is very large this will be insignificant. If you stop the entire sample gathering exercise after the last two tosses are different, ie. either HT or TH, you will either finish the last session $1 up, or have the last, unfinished session $1 down. In other words, if there is an unfinished session, it is irrelevant.] This sounds like paradise - a certain way of winning money! Where's the flaw? There is only one flaw: some of those session will have several losses in a row. You cannot double your bets indefinitely, you will hit either the house limit or your bankrolls limit (let's say that limit is $1,000). 10 losses in a row takes you over $1,000. What are the chances of having 10 losses in a row? Annoyingly, about 1-in-1,000. In other words, by the time you've accumulated 1000 winning sessions of $1, you're about due for a loss of $1,000. Net result: $0. Bummer. You've had 1000 winning sessions, only 1 losing session, and you are net $0. What a waste of effort! [What did you expect on a even-odds coin toss game? Remember Win=TotalBet*Advantage, and what answer do you get when Advantage=0?] What happens if, by bad luck, the $1,000 losing session comes before you've played 1000 winning sessions? What happens if it came after only 200 winning sessions? Obviously you'd be down $800, with no more money to play. You go home a loser (hardly an unusual occurrence in a casino). But what about if the losing session did NOT come in the first 1000 sessions? Ah, now you are a winner. Let's say you've played 1100 sessions and still haven't had the $1000 disaster. You've probably had a few scares, when the bets got to $256 or $512, but 1100 winning sessions and no losing sessions means you are up $1100. What do you do now? You think you're a genius, you're unbeatable, you're invincible. You do NOT stop gambling; if anything you up the odds (you have the really smart idea that if you can win $1100 starting at $1 per bet, why not start at $2, or even $5? What you should do is buy your wife a very nice present (this advice is biased). Gamblers rarely do, it is almost impossible for them to quit while ahead. Obviously (they delude themselves) they are winners (as if it had anything to do with them, rather than blind luck), so quitting would be foolish. Easy money awaits... Casino's do not offer even-odds games. In reality the chance having a 10 run loss is greater than 1-in-1000 (in blackjack, for example, where you only win 46% of your hands (ignoring pushes), the chance of a 10 run losing streak is 0.54^10 = 2-in-1000. In double-0 roulette, it is (20/38)^10 = 1.6-in-1000. Chances are you will hit your bad streak before you have accumulated enough win to cover it. The easiest way to visualize a martingale's effect is as follows (this is based on the mathematical law that a+b = b+a, ie. the sequence in which you add numbers doesn't affect the total). Take your series of coin tosses, and martingale bets. Extract all those bets which were $1 (ie. the first bet of a session). Look at them. I guarantee that approximately half of them are winners, and half losers (it was a coin toss game, so obviously). Now extract all the $2 bets. I guarantee half of them are winners, and half losers. Now extract all the $4 bets.... No matter what you bet, your chance of winning is 50/50 (hardly surprising on a coin toss game). If you bet $4 (which means the previous $2 bet lost), you win half the time and lose half the time. Same for every other bet denomination. In other words, whether or not you won or lost the previous bet does not alter your chance of winning the next bet at all. The next bet will win 50% of the time. Add up the total amount you bet in your sample using the martingale system. Let's say it totalled $1,000,000. Now let's play again. Pull out your coin and bet a completely random amount, either $1, $2, $4, $8, $16 or any other such multiple, up to $1024. Toss the coin, record you win or loss, and the total amount bet so far. Repeat, and keep repeating until the total amount you have bet is $1,000,000 (same as the martingale total). Extract all those bets which were $1 (ie. the first bet of a session). Look at them. I guarantee that approximately half of them are winners, and half losers (it was a coin toss game, so obviously). Now extract all the $2 bets. I guarantee half of them are winners, and half losers. Now extract all the $4 bets.... In other words, betting martingale, or betting random, will produce the same win rate for every bet amount (50/50). And if every bet amount has the same chance of winning, then OBVIOUSLY, the grand total net total win will be the same too (hint: Win=TotalBet*Advantage). The order in which you've placed the bets makes no difference! Casinos win money for several reasons: 1) All their games have a house advantage 2) "Gamblers' Ruin" (this is what statisticians call the effect that in a gambling game, the person with the least money will the the first to hit a bad run, go bankrupt, and be unable to carry on playing). 3) Winners almost never stop. House advantage is powerful (of course). It is continually nibbling away (like rake at poker), and it adds up powerfully after a long time. Eg. a Martingale on a roulette table has a 1.6-in-1000 chance of hitting 10 losses in a row, even though it only pays 1000-to-1. You are almost twice as likely to hit a 10 run losing streak than you are to accumulate the funds to cover it. And that only gets worse with longer runs. Gamblers' Ruin is what kills Martingale system. Casino games are nearly always very rapid. A good blackjack player, on a good table, can play over 200 hands per hour. 5 hours is 1,000 hands, So a 10 run losing streak can occur twice in 5 hours on a BJ table. So, "System Sellors" (aka "System Purveyors", aka "Con Men", aka "Idiots") often modify Martingale. With Martingale one single win is enough to offset any number of previous losses. This is a huge comeback (impossible to do it faster). But the risk of ruin is very high. So a slower recovery is often proposed (I've seen Fibonacci series, half-Martingale's, and every other weird combination you could think off). All they do is choose a comeback method somewhere between my two examples of flat betting $1 and waiting for what might be hundreds or thousands of tosses to eventually get back to +$1, or martingale (get back in one single winning toss). Just as a coin toss game's Martingale had 1-in-1000 chances of a 1000-to-1 loss, so do all the other betting schemes cancel out. Always Win = TotalBet*Advantage. [Aside: there is a thriving little industry in Vegas where conmen will 'sell' you a betting scheme. They "Sell" the suckers an impressive looking (on paper) modified Martingale system. You don't need to pay them any money up front for this, they assure you, merely a small share of the easy profits you'll make with the wonderful "new" system. With Martingale (or similar) you will win 80%+ of your sessions. At the end of each day you give them 25% of what you won (keeping 75% to yourself). Hopefully they'll get a few payments from you before, as is certain, you have the big losing day. It's a wonderful little con, and works great from the conmen's point of view (they don't risk any of their money, and don't do any work, merely chasing down their daily payments which is often surprisingly easy from the idiots who think a few winning sessions are due to the wonderful new system). Best of all, most humans are unable to understand that Win=TotalBet*Advantage and Martingale's do not alter this truth in any useful respect.] To improve your win you either have to: 1) Improve your Advantage. This means either to bet more when you are more likely to win (equally to bet less when you are more likely to lose), and/or to play differently to increase the chance of your winning. In BJ counting cards is one such method (as is marking cards, there are a diverse range of ways of improving Advantage). Because, in roulette, BJ, and many others, whether or not you won or lost the previous bet has NO effect on your odds on the next bet, so ANY betting scheme that operates on previous winning or losing data is equivalent to random betting. 2) Appropriately alter your TotalBet. If you are playing a -EV game, the best way to increase Win is to make TotalBet=0. If you are playing a +EV game, you can increase Win by increasing TotalBet. Eg. If you are a competent card counter, and can win on a $5 table, then you'll win twice as much on a $10 table - all else being equal (there are many, non-mathematical issues, such as casino heat). However, Gambling is not so much about maximizing win, as it is about maximizing win/risk. This is a whole topic to itself that I won't start here. To the OP (and any other person who cannot see past the smoke and mirrors): NO betting scheme will affect your long-run expected result. Always Win = TotalBet * Advantage. If you don't learnt how to change Advantage you have no hope of affecting your total Win. Short-term results can be affected by betting systems. But these are useful ONLY if you STOP (and NEVER gamble again in your entire life) after reaching your goal. If I was a gambler, I'd bet against that happening. Something like 80-90% of the gamblers in your local casino will lose today. You can do dramatically better than them (by 30-40%), by walking in, placing one single bet, and walking out regardless of the result, never to return. If you want to gamble, and win, you have to work on improving Advantage. You have to make it so that, simply, more of your bets win than lose. Rearranging the sequence of bets (via a Martingale, or whatever), does nothing constructive. To think otherwise is delusional.
|
|
#26
07-22-2006, 11:51 PM
|
|||
|
|||
|
Re: Blackjack Strategy
Sandra - you rock.
I've explained it similarly before although without all the x and y stuff that tends to turn people off sometimes (I'm easily confused). But one aspect of what Sandra is saying that I have also explained goes like this: If you bet a total of $10k on a game with a 1% disadvantage it doesn't matter whether you bet a little bit at a time, or mix it up, or whatever. Your expectation is still to lose $100 (1% of $10k). So if you bet exactly $100 on every bet for a total of $10k in bets your expectation at the end is -$100. If you bet it in a progressive-type way of $100, $200, $300, $400 (starting over whenever you think it's best, whether you won or lost or are 'playing with house money' or whatever) then once your total of bets reaches $10k your expectation is still -$100. Notice that if you mix it up in ANY way it doesn't matter. formula A: 100, 200, 300, 400, 500, 100, 200, 300... will have the same expectation as formula B: 300, 100, 500, 900, 50, 200, 400, etc It doesn't freaking matter what order you place those bets in. you can try to avoid this effect as much as you like by varying it in whatever way you choose. You can't escape this fact and whatever logic you choose to try to justify any progressive betting scheme as being 'better' than another is going to be flawed. Once you reach a grand total of $10k in bets at a 1% disadvantage your expectation is -$100. Most of the arguments promoting such progressive-betting 'schemes' are not unlike a fish at your poker-table saying that he's "due" for aces (and is suddenly more likely to get them) because he hasn't gotten them in his last 348 hands.
|
|
#28
07-23-2006, 12:10 AM
|
|||
|
|||
|
Re: Blackjack Strategy
So Martingale system doesnt work to win lots of money easily?
[img]/images/graemlins/shocked.gif[/img] [img]/images/graemlins/shocked.gif[/img] [img]/images/graemlins/shocked.gif[/img] [img]/images/graemlins/shocked.gif[/img] [img]/images/graemlins/shocked.gif[/img]
|
|
#30
07-23-2006, 03:20 AM
|
|||
|
|||
|
Re: Blackjack Strategy
I understand that 1% -ev is 1% of all bets lost but if I am entering a session with $20 and if I win I play until I am bored (usually when I start losing) If i lose my $20 I go do something else
So far it seems to be working very very well (I haven't played poker today just blackjack and I have made far more then I could in a week of poker at the .50/100 limit..) Also does anyone know the odds of being dealt a winning blackjack.. i.e. you have blackjack and dealer does not
|
 |
|
|
Linear Mode
