I'll steal this line from someone I can't stand: "Like the mosquito at the nudist colony, I don't know where to begin."
First off, I'll second what Josh said in his post.
@dpmorel If I could consolidate your advice, it sounds like your saying "gambling is a terrible idea for most people" and I would totally agree with that. As I said before:
I'm worried that people might read this thread and think that there's easy money to be made when actually winning money from a casino long term is difficult and requires work, math, computer simulations, and patience.
Now, again, I love our little community so I don't want to offend but a lot of the details you list afterwards are wrong.
I do believe the comparison between AP gambling and investing is valid.(AP = Advantage Play, ie you're a smart gambler who understands the odds, bankroll management etc)
Your first point is completely valid:
1. You need a big enough bankroll to ride the wave. Your bankroll is not infinity, therefore the odds are the odds of that night based on the cash you have and the cards you see. All odds quoted are the odds seen by people who study cards over a very long period of time. Those odds don't apply on any given night you play.
Totally agree. The same could be said of those studying the stock market, whenever you hear someone saying stocks have returned x% "in the long run", the long run they are referring to is a very, very long time. Indeed this is an advantage to gambling, I have no idea what the long run return of the stock market should be (and neither does anyone else) but I can calculate the exact long run return of counting a Blackjack game (ie one can calculate the exact normal distribution of the results and decide if you like them).
2. The downside:
-Investing the downside is step-wise (usually a few percentile) and takes days/months to realize significant loss (depending exactly on how you invest)
"Usually" is correct, unfortunately sometimes investing will punch you in the mouth (as we've seen recently). Also remember that here on ERE we're talking about investing HUNDREDS of THOUSANDS of dollars, a few % points is many thousands of dollars.
Gambling the downside is usually *everything* (you bring a bankroll for a reason), and it can happen lightening fast... before your brain and hormones have a chance to tell you something is wrong.
If this is the case then the gambler's bankroll management is *terrible* and they shouldn't be gambling at all. Just because this is the way many, nay most, people gamble does not mean there is no way to do it sensibly. And again, short term results do not tell you whether something is a good idea or not. If you know the math and you have a down swing you should not feel obliged to stop.
Anybody who condones gambling as a means to make a living or as a means to supplement income has never been a real gambler.
This is not true.
Its only a matter of time until you hit a cold blackjack table one night, and your cash runs out. I don't care what system you play.
Again, if you lose all your money at a "cold" table one night then you almost certainly were gambling WAY too much for your bankroll size. You can calculate what your bet should be for a given bankroll size using the Kelly criterion.
Honestly though, the technical discussion pulls away from the emotional and life arguments. Losing at gambling does not have the same emotional feeling as losing at investing. It is one of the most horrible feelings in the world to have to go back to your significant other and tell them you have lost a significant amount of money in one night.
It is also very, very, very hard to get back up on the horse and put money back on the table after you get hit hard one night. I no longer gamble anymore after being hit hard twice.
This is true, it can be emotionally difficult. Unfortunately, I find that losing a brick of cash hurts more than the good feeling you get from winning a brick of cash. It also hurts more than losing the same amount as depicted by digits on a screen because the market dipped a few points. Ultimately, you still lost the same amount of money and just need to keep telling yourself that. I'm sorry to tell you this, and again, I mean no offense but what you were doing when you were gambling before was probably terrible. This does not mean that there are no legitimate ways to win in the long run.
Your next post about blackjack:
BTW - the cheatsheet doesn't matter. I played at a casino where they gave them to you when you walked in.
Yes, you still lose in the long run by playing basic strategy with a cheatsheet (and if the casino gave you one it probably also has errors that will cause you to lose faster). The amount of that loss in a reasonable game where you're betting the minimum is extremely small though (with variance obviously) but if you're betting the minimum (eg $5 or $10) then the chance of a really horrible (ie many hundreds of $) loss is exceedingly small.
The only thing that matters in BlackJack is the bankroll.
The way BlackJack works is you dribble along going through runs of winning and losing. Then you get a chance to double down with great odds. This is where you make your money. You need to have had enough cash to survive the runs of losing so you have enough cash to make your stack back and surplus on the double down.
No, doubling down is part of basic strategy. It is already taken into account in calculating that a given game has a certain long term % loss. Even if you had an infinite amount of money to ensure that you could always double down you would still lose in the long run by that % of the total amount you have bet.
The way you make money in blackjack is by card counting. To simplify it: With card counting you bet the minimum when the odds are in the casinos favor (because the deck is rich in small cards) and you bet big when the odds are in your favor (because the deck is rich in tens and aces). Thus the math tells you that you win money in the long run from the casino (yes, with pretty scary variance).
How much bankroll do you need? Well thats the problem, as I blogged above the bankroll size is totally dependent on the cards that the table serves that nights. No system in the world can tell you that. Odds/patterns only count over very long periods of time.
Yes, it is possible to lose everything you took with you to the casino. It usually doesn't make sense to take your entire bankroll for security reasons (you don't want to lose everything if you get mugged in the parking lot). Also, it makes sense to keep it in a bank earning interest rather than lugging it around on the very small chance that you lose your (smaller) trip bankroll.
ERE Gambler
I have no idea what the long run return of the stock market should be (and neither does anyone else) but I can calculate the exact long run return of counting a Blackjack game (ie one can calculate the exact normal distribution of the results and decide if you like them).
Actually you can't calculate the long run return for BlackJAck. Thats what I refer to as the difference between continuous and discrete math. The odds on your 1000th hand are the exact same as the odds in any given hand. Thats the gambler's bias, it is false to say you can ever calculate your long term earnings based on a compilation of discrete events.
The mathematical formula to calculate your odds is a sum of the odds of each individual hand. The odds of an individual hand are *unknown*. I will admit you can make the odds of an individual hand better by counting... but this is very very very different than saying "over the long term I will win 51 out of 49 times". No - that is statistically incorrect.
But thats all theory. In practice as I keep saying most gamblers eventually ahve this happen, and it is totally statistically possible:
W,W,W,L,L,L,L,L,L,L,L, W,W, L,L,L,L,L. That can be nights instead of hands. You tell me the bankroll strategy that prevents this? Keep making your bets smaller??? use a smaller bankroll? Bankroll mgmt is "stastical magic" because it presume that if you play long enough the odds will return in your favour. This is Gambler's Bias and is statistically incorrect.
Actually you can't calculate the long run return for BlackJAck. Thats what I refer to as the difference between continuous and discrete math. The odds on your 1000th hand are the exact same as the odds in any given hand. Thats the gambler's bias, it is false to say you can ever calculate your long term earnings based on a compilation of discrete events.
The mathematical formula to calculate your odds is a sum of the odds of each individual hand. The odds of an individual hand are *unknown*. I will admit you can make the odds of an individual hand better by counting... but this is very very very different than saying "over the long term I will win 51 out of 49 times". No - that is statistically incorrect.
But thats all theory. In practice as I keep saying most gamblers eventually ahve this happen, and it is totally statistically possible:
W,W,W,L,L,L,L,L,L,L,L, W,W, L,L,L,L,L. That can be nights instead of hands. You tell me the bankroll strategy that prevents this? Keep making your bets smaller??? use a smaller bankroll? Bankroll mgmt is "stastical magic" because it presume that if you play long enough the odds will return in your favour. This is Gambler's Bias and is statistically incorrect.
Let me try to explain this another way...
Each of these are a set of gambling events. Lets say hands at a table or nights of playing cards:
W,W,W,W,W,W,W,W,W,WW,W,W,W,W,W
L,L,L,L,L,L,L,L,L,L,L,L,L,L,L,L
W,L,L,L,W,W,L,L,L,W,W,L,L,L,L,W
W,W,W,L,W,L,W,L,W,L,W,L,W,L,W,L
L,L,L,L,L,W,L,L,L,L,W,L,L,L,L,W
Which of these is most statistically probable?
Each of these are a set of gambling events. Lets say hands at a table or nights of playing cards:
W,W,W,W,W,W,W,W,W,WW,W,W,W,W,W
L,L,L,L,L,L,L,L,L,L,L,L,L,L,L,L
W,L,L,L,W,W,L,L,L,W,W,L,L,L,L,W
W,W,W,L,W,L,W,L,W,L,W,L,W,L,W,L
L,L,L,L,L,W,L,L,L,L,W,L,L,L,L,W
Which of these is most statistically probable?
-
RightClawSouth
- Posts: 123
- Joined: Wed Jul 28, 2010 3:15 am
The odds of an individual hand are *unknown*.
This is absolutely not true. You can run simulations and/or do math to figure out the expected return of any bet. For basic strategy this guy has done the work. For counting you can get software that will run simulations for the strategy you want to use and you can figure out the odds.
This is absolutely not true. You can run simulations and/or do math to figure out the expected return of any bet. For basic strategy this guy has done the work. For counting you can get software that will run simulations for the strategy you want to use and you can figure out the odds.
-
RightClawSouth
- Posts: 123
- Joined: Wed Jul 28, 2010 3:15 am
Let me try to explain this another way...
Each of these are a set of gambling events. Lets say hands at a table or nights of playing cards:
W,W,W,W,W,W,W,W,W,WW,W,W,W,W,W
L,L,L,L,L,L,L,L,L,L,L,L,L,L,L,L
W,L,L,L,W,W,L,L,L,W,W,L,L,L,L,W
W,W,W,L,W,L,W,L,W,L,W,L,W,L,W,L
L,L,L,L,L,W,L,L,L,L,W,L,L,L,L,W
Which of these is most statistically probable?
If there are the same number of Ws and Ls then they're equally likely.
Otherwise it depends on the odds of winning...
I have to head out now but I'll be back later to chat more...
(EDIT: changed to make more accurate
)-
RightClawSouth
- Posts: 123
- Joined: Wed Jul 28, 2010 3:15 am
@RCS - your first answer is the mathematically correct answer. They are all equally likely. They are all likely because a single discrete event has no bearing on the next event.
So forgetting card counting then, and applying the above law of math, which of these is more "risky":
-betting $10k once ..OR..
-betting 200 $50 bets
I'll answer for you, they are equally risky.
This is why "bankroll mgmt" is mathematically incorrect.
So forgetting card counting then, and applying the above law of math, which of these is more "risky":
-betting $10k once ..OR..
-betting 200 $50 bets
I'll answer for you, they are equally risky.
This is why "bankroll mgmt" is mathematically incorrect.
So now the big question is "what are the odds of winning any given hand".
Well, if you know nothing about what is in the dealer's deck then the odds of winning are *unknown*. In that case your odds look like this:
unknown, unknown, unknown, unknown, unknown,..
so the *real* mathematical chance of winning is... unknown...
If you count cards using any strategy then every once in a while you know something about the deck.
So therefore your odds are:
unknown, unknown, unknown, unknown,..., I know something, unknown, unknown,...
The rate at which you *know something* is the crux, but your odds still tend basically towards unknown.
Well, if you know nothing about what is in the dealer's deck then the odds of winning are *unknown*. In that case your odds look like this:
unknown, unknown, unknown, unknown, unknown,..
so the *real* mathematical chance of winning is... unknown...
If you count cards using any strategy then every once in a while you know something about the deck.
So therefore your odds are:
unknown, unknown, unknown, unknown,..., I know something, unknown, unknown,...
The rate at which you *know something* is the crux, but your odds still tend basically towards unknown.
So last is comparing this to the market. Some big differences in investing in the market vs gambling.
1. It is possible to *know* things about the market sometimes. Its not always an *unknown*. This is why inside information is illegal.
2. You don't have to lose on *unknowns* to play *knowns*. You can wait out the market until its "favourable". If you did that at a BlackJack table you'd get thrown out
3. When you invest your money in the market your money is continuously in the market. Your money will have the rate of return of what you invested in.
1. It is possible to *know* things about the market sometimes. Its not always an *unknown*. This is why inside information is illegal.
2. You don't have to lose on *unknowns* to play *knowns*. You can wait out the market until its "favourable". If you did that at a BlackJack table you'd get thrown out
3. When you invest your money in the market your money is continuously in the market. Your money will have the rate of return of what you invested in.
-
RightClawSouth
- Posts: 123
- Joined: Wed Jul 28, 2010 3:15 am
The probability of each of those lines is only the same if the probability of winning and losing is the same. This is the probability of getting each of those lines:
1. P(W)^16
2. (1-P(W))^16
3. P(W)^6 x (1-P(W))^10
4. P(W)^9 x (1-P(W))^7
5. P(W)^3 x (1-P(W))^13
Only in the situation where P(W) = 0.5 are those lines equally likely.
eg if P(W) = 0.52 then the likelihood of getting each of those lines is:
1. 0.0000286
2. 0.00000794
3. 0.0000128
4. 0.0000163
5. 0.0000101
Notice that I am NOT assuming that any previous discrete events are influencing future events.
1. P(W)^16
2. (1-P(W))^16
3. P(W)^6 x (1-P(W))^10
4. P(W)^9 x (1-P(W))^7
5. P(W)^3 x (1-P(W))^13
Only in the situation where P(W) = 0.5 are those lines equally likely.
eg if P(W) = 0.52 then the likelihood of getting each of those lines is:
1. 0.0000286
2. 0.00000794
3. 0.0000128
4. 0.0000163
5. 0.0000101
Notice that I am NOT assuming that any previous discrete events are influencing future events.
-
RightClawSouth
- Posts: 123
- Joined: Wed Jul 28, 2010 3:15 am
So forgetting card counting then, and applying the above law of math, which of these is more "risky":
-betting $10k once ..OR..
-betting 200 $50 bets
I'll answer for you, they are equally risky.
This is why "bankroll mgmt" is mathematically incorrect.
I'm sorry this is just wrong. Unless we have different definitions of what "risky" means. Betting $10,000 once is absolutely more risky than betting $50 200 times. This is true even if we simplify this to a coin toss. Are you seriously saying that you would be indifferent to betting $10,000 of your money on a coin toss compared to betting $50 on 200 coin tosses?
-betting $10k once ..OR..
-betting 200 $50 bets
I'll answer for you, they are equally risky.
This is why "bankroll mgmt" is mathematically incorrect.
I'm sorry this is just wrong. Unless we have different definitions of what "risky" means. Betting $10,000 once is absolutely more risky than betting $50 200 times. This is true even if we simplify this to a coin toss. Are you seriously saying that you would be indifferent to betting $10,000 of your money on a coin toss compared to betting $50 on 200 coin tosses?
I would say to that - yes, if I could bet 10k in one shot, and double up, then go for it.
You could toss that coin 50 times and I lose 49 out of 50. And just have my $200. I would give extreme examples that no-one (on this board) would ever commit, but I already killed a post topic once this week already.
If you take money out of the equation, and apply the risk to real life decisions, peoples answers change on the situation right?
By all your posts, you basically state that you need to run a simulation first to see if chances are greater than 50% to "win". Life, in general and to me, doesn't work like that, so I have a hard time accepting that claim.
At this point, I think we all can agree to disagree, and I chuckle a bit.
You could toss that coin 50 times and I lose 49 out of 50. And just have my $200. I would give extreme examples that no-one (on this board) would ever commit, but I already killed a post topic once this week already.
If you take money out of the equation, and apply the risk to real life decisions, peoples answers change on the situation right?
By all your posts, you basically state that you need to run a simulation first to see if chances are greater than 50% to "win". Life, in general and to me, doesn't work like that, so I have a hard time accepting that claim.
At this point, I think we all can agree to disagree, and I chuckle a bit.
@RCS - you are incorrect mathematically. Seriously.
Lets not use cards, I think a single dice helps clear up the example.
If I roll a dice once, the chances of me getting a 1 is 1/6.
If I don't get a one on the first roll, what is the chance of me getting a 1 on the second roll.
Now lets say I roll the dice 6 times. What is my chance of getting a 1 on the 6th roll? 1 in 6.
So, now lets examine all possibilities of 6 dice rolls:
6,6,6,6,6,6
1,1,1,1,1,1
1,2,3,4,5,6
3,4,3,4,3,4
etc.
etc.
Each of these is equally likely. Rolling a 1 6 times in a row is just as probable as rolling a 6 6 times in a row.
You can expand this out. If I roll a dice 1000 times, on the 1001th try, the odds of me rolling a 1 are still 1 in 6. Even if I rolled 1000 ones before that.
It doesn't matter what the probability of a single discrete event, within each event the likelihood is equally as likely.
Now again, apply this to blackjack...
w,w,w,w,w,w,w
l,l,l,l,l,l,l
w,l,w,l,l,w,l
They are all equally likely patterns. The odds don't have to be 50/50 for this to hold true.
This math is tough to get your head around. It was taught in a 3rd year stats course for me, so its non-trivial.
Lets not use cards, I think a single dice helps clear up the example.
If I roll a dice once, the chances of me getting a 1 is 1/6.
If I don't get a one on the first roll, what is the chance of me getting a 1 on the second roll.
Now lets say I roll the dice 6 times. What is my chance of getting a 1 on the 6th roll? 1 in 6.
So, now lets examine all possibilities of 6 dice rolls:
6,6,6,6,6,6
1,1,1,1,1,1
1,2,3,4,5,6
3,4,3,4,3,4
etc.
etc.
Each of these is equally likely. Rolling a 1 6 times in a row is just as probable as rolling a 6 6 times in a row.
You can expand this out. If I roll a dice 1000 times, on the 1001th try, the odds of me rolling a 1 are still 1 in 6. Even if I rolled 1000 ones before that.
It doesn't matter what the probability of a single discrete event, within each event the likelihood is equally as likely.
Now again, apply this to blackjack...
w,w,w,w,w,w,w
l,l,l,l,l,l,l
w,l,w,l,l,w,l
They are all equally likely patterns. The odds don't have to be 50/50 for this to hold true.
This math is tough to get your head around. It was taught in a 3rd year stats course for me, so its non-trivial.
Here, I'll put a link up to elementary probability theory. Don't take my word, take the word of Elementary Probability Theory.
http://www.rpg.net/columns/rollthebones ... nes2.phtml
"Two events are independent if the outcome of one event does not influence the outcome of the other. For example, when you roll a die twice, the outcome of the second roll does not depend on the outcome of the first roll (the die doesn't remember the previous roll), so the events are independent"
This is why gambling $10k once is just as risky as gambling 200 $50 bets.
Look man, its not that I want to win the argument. Its just that people don't understand the real math behind gambling. The chances of you winning on any hand without card counting are *unknown*. The chances of you winning on any given night are also *unknown*. Hopefully there is enough math and probability theory in here to show why.
http://www.rpg.net/columns/rollthebones ... nes2.phtml
"Two events are independent if the outcome of one event does not influence the outcome of the other. For example, when you roll a die twice, the outcome of the second roll does not depend on the outcome of the first roll (the die doesn't remember the previous roll), so the events are independent"
This is why gambling $10k once is just as risky as gambling 200 $50 bets.
Look man, its not that I want to win the argument. Its just that people don't understand the real math behind gambling. The chances of you winning on any hand without card counting are *unknown*. The chances of you winning on any given night are also *unknown*. Hopefully there is enough math and probability theory in here to show why.
-
RightClawSouth
- Posts: 123
- Joined: Wed Jul 28, 2010 3:15 am
You are correct that the probabilities of each of those dice sequences are the same. This is because the probability of each number is the same (1/6).
I showed you with the win-loss sequences that the probabilities DO have to be the same (ie 50% chance of winning, 50% chance of losing) for wins as they are for losses for each of those to be equally likely.
Lets take it to an extreme, if the probability of winning was 100%, do you still think those sequences are equally likely?
I showed you with the win-loss sequences that the probabilities DO have to be the same (ie 50% chance of winning, 50% chance of losing) for wins as they are for losses for each of those to be equally likely.
Lets take it to an extreme, if the probability of winning was 100%, do you still think those sequences are equally likely?
-
RightClawSouth
- Posts: 123
- Joined: Wed Jul 28, 2010 3:15 am
Just to point out the formula. It is the rule of coincidence:
The rule of coincidence:
If two events are independent and have probabilities p and q, respectively, of happening, then the chance that both happen is p × q (p times q).
That is the math you want... these are incorrect formulaes for calculating risk of independent events:
1. P(W)^16
2. (1-P(W))^16
3. P(W)^6 x (1-P(W))^10
4. P(W)^9 x (1-P(W))^7
5. P(W)^3 x (1-P(W))^13
The rule of coincidence:
If two events are independent and have probabilities p and q, respectively, of happening, then the chance that both happen is p × q (p times q).
That is the math you want... these are incorrect formulaes for calculating risk of independent events:
1. P(W)^16
2. (1-P(W))^16
3. P(W)^6 x (1-P(W))^10
4. P(W)^9 x (1-P(W))^7
5. P(W)^3 x (1-P(W))^13
-
RightClawSouth
- Posts: 123
- Joined: Wed Jul 28, 2010 3:15 am
if p is the prob of winning then the probabilities are
1. (p)(p)(p)(p)(p)(p)(p)(p)(p)(p)(p)(p)(p)(p)(p)(p) = p^16
2. (1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p) = (1-p)^16
3. (p)(1-p)(1-p)(1-p)(p)(p)(1-p)(1-p)(1-p)(p)(p)(1-p)(1-p)(1-p)(p)(p)(1-p)(1-p)(1-p)(1-p)(p) = (p^6)(1-p)^10
etc, which is exactly what I posted before.
What do you think the correct formulas are?...
1. (p)(p)(p)(p)(p)(p)(p)(p)(p)(p)(p)(p)(p)(p)(p)(p) = p^16
2. (1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p)(1-p) = (1-p)^16
3. (p)(1-p)(1-p)(1-p)(p)(p)(1-p)(1-p)(1-p)(p)(p)(1-p)(1-p)(1-p)(p)(p)(1-p)(1-p)(1-p)(1-p)(p) = (p^6)(1-p)^10
etc, which is exactly what I posted before.
What do you think the correct formulas are?...
You're right, I goofed that... sorry. I'll eat my humble pie (with whip cream on top).
Annoyingly the dice example threw myself off. Its not that the odds have to be 50/50, its that all results have to be equally probably. whoops!
Actually though I think the unknown part still holds though and the overall argument that gambling is a bad way to make money holds true.
Annoyingly the dice example threw myself off. Its not that the odds have to be 50/50, its that all results have to be equally probably. whoops!
Actually though I think the unknown part still holds though and the overall argument that gambling is a bad way to make money holds true.