Roulette Theory

[silverpig]

Originally posted by: alkemyst

Originally posted by: jman19
That's not what he said.

It sounds like another case of the gambler's fallacy to me - i.e. let's say you roll a dice 100 times and never roll a single 6. The gambler's fallacy would say "I haven't rolled a 6 in 100 rolls! That means I'm more likely to roll one in the future!"

That statement to me is saying your odds of getting a 6 in 100 rolls is no different than 101 rolls. In reality your odds get better each time. This question is fundamentally flawed as only in theory would you ever see someone NOT get at least one 6 in 100 rolls of a 6-sided die assuming no trickery is involved.

How would you do the math on this presentation?

Uh no, this is completely wrong. His original statement was correct. Your odds of rolling a 6 do not ever change. His original statement is something that people think a lot. "I've had 3 girls in a row, so the odds are my next one will be a boy!" That's exactly the same statement.

You can't take past independent events and use them to predict the outcome of the next independent event.

The odds of not rolling a 6 in 100 tries, then rolling a 6 on the 101st are the exact same as not rolling a 6 in 100 tries, then rolling a 2 on the 101st.

 

[Lorax]

how is it possible that some of ATOT, a fairly intelligent community, does not understand basic probability?

 

[DBL]

Originally posted by: aplefka

Originally posted by: Crusty
You go to Casinos to have fun, not win money. If you think otherwise you are an idiot.

Call me an idiot, but I've done pretty well for myself with poker. And I never expect anything less than to win money. Are you going to tell me that poker is all luck too?

C'mon, as a poker player, you should know that poker follows a different set of rules. You can certainly play winning poker in a casio over the long-run, although it is much harder than many people realize.

 

[Toasthead]

Thats why they have table limits

 

[jman19]

Originally posted by: alkemyst

Originally posted by: Special K

Originally posted by: alkemyst

Originally posted by: jman19
That's not what he said.

It sounds like another case of the gambler's fallacy to me - i.e. let's say you roll a dice 100 times and never roll a single 6. The gambler's fallacy would say "I haven't rolled a 6 in 100 rolls! That means I'm more likely to roll one in the future!"

That statement to me is saying your odds of getting a 6 in 100 rolls is no different than 101 rolls. In reality your odds get better each time. This question is fundamentally flawed as only in theory would you ever see someone NOT get at least one 6 in 100 rolls of a 6-sided die assuming no trickery is involved.

How would you do the math on this presentation?

I chose the number 100 somewhat arbitrarily. It really doesn't matter which number you pick - my point was that the outcome of the next die roll is not influenced by the previous ones.

However, as I said in a previous post:

Originally posted by: Special K
I guess we were just referring to different scenarios, although now I am confused - let's say a gambler wins if he rolls a 6, and loses otherwise. So he keeps rolling the die, hoping to get a 6. Let's say the first 20 rolls are not 6's. If we look at it the way I broke it down, then his probability of rolling a 6 on the 21st roll is 1/6, because each roll is independent. On the other hand, if we ask "what is the probability of not rolling any sixes in 21 rolls", the answer is (5/6)^21, or 0.0217.

Would anyone else like to comment?

I was replying to Jman to find out his take on it.

I don't know why you are now adding in the probability of not picking a 6...you are diluting your own question and making it overly complex.

The probability of not rolling a 6 in 21 rolls would be your 0.0217 result which would give about 98% probably to roll a 6 in 21 rolls.

If you really want to test it rolling a dice 21 times and recording the results over say 20+ times, 100 iterations at most; you should see close to this result. Use a real die and not a computer, many people mistakenly try to reproduce these questions in code, but computers don't really do random numbers so well with the methods most use.

There are some examples that work like this that are easily tested in a small group.

One of them is how many people have to be asked before two people in the room have the same birth month/day.

many will say 180ish...but in reality it's at 23 people you have the same odds of heads or tails...50% IIRC.

I took him to mean that the next roll was independant of the previous ones, which is true. There is a difference between saying "what are the odds of rolling 100 non-six numbers and then 6" and saying "what are the odds that the next dice roll is a six" - obviously these probabilities are very different. It doesn't matter how many rolls you've made to know what the probability of rolling a 6 is.

 

[jman19]

Originally posted by: spidey07

Originally posted by: jman19

Originally posted by: spidey07
All you math guys...

Have you ever gambled? Have you ever laid out a bet based on feel? Have you ever played?

It's fun to play with math, but call me a sucker again, you can feel it and react accordingly. Ever been around a hot craps table?

Honestly for those that are playing math...Have you EVER been at a hot craps table? If so, the what was the outcome?

In a limited number of games, the variability away from the true probability of an event can be rather large - that is the "heat" you are talking about. It's not some magical entity.

So, what was your outcome on a hot craps table? Did you make thousands?

-edit-
Call me stupid, call me what you will. But I'll keep up this hobby of mine and have a good time doing it.

I'm not calling you stupid, I'm explaining what a "hot" table means, is all.

 

[Kyteland]

Originally posted by: spidey07
All you math guys...

Have you ever gambled? Have you ever laid out a bet based on feel? Have you ever played?

It's fun to play with math, but call me a sucker again, you can feel it and react accordingly. Ever been around a hot craps table?

Honestly for those that are playing math...Have you EVER been at a hot craps table? If so, the what was the outcome?

I'm one of these "math guys," probably more so than others on these boards. I make these game for a living. I prove out the odds for the games and the company I work for sells them to casinos.

I also play the games all the time. It comes with the territory and allows me to make new, better games. I can play for free as much as I want in the office, but it just isn't the same as playing for real money. I'll be in Vegas for 7 days next week for our trade show, and I'll be gambling.

Everybody and their mother thinks they have ESP and can use it to gain the edge in a casino. I assure you that it's not true. If you kept real statistics on your play I could easily show you badly you actually do "based on feel." I'd bet that it's within 3 standard deviations of house edge.

I'd be happy to take some time to show you why these things are fallacies. Actually, this guy has probably covered every possible question you could ask.

 

[spidey07]

Kyteland,

I don't keep real statistics on my play but I DO have to keep records of winnings/losses for tax purposes. I'll keep using my ESP.

Professional handicappers (I'm just an amateur) are meticulous about the payout percentages and risk/reward. Heck that's all they are, statisticians.

But you can tell when a craps table is hot and when it's going bad. Same with blackjack. Bet big when it's good, and back down when it's not or go to a different table. Same with the ponies, if using one strategy isn't working and you see another trend for the day then go to that trend.

 

[alkemyst]

Originally posted by: silverpig
Uh no, this is completely wrong. His original statement was correct. Your odds of rolling a 6 do not ever change. His original statement is something that people think a lot. "I've had 3 girls in a row, so the odds are my next one will be a boy!" That's exactly the same statement.

You can't take past independent events and use them to predict the outcome of the next independent event.

The odds of not rolling a 6 in 100 tries, then rolling a 6 on the 101st are the exact same as not rolling a 6 in 100 tries, then rolling a 2 on the 101st.

For one having a girl vs boy is not really dependent on chance.

That said you have stated what I already explained.

The odds of getting a 6 in 101 rolls improves each time that the event does not happen.

The odds of rolling a 6 on the 101st roll is a separate condition.

However; in gambling you are betting on occurances over trials...each event is not independent at all.

I don't think anyone here is understanding how the math really works in these situations based on the absurd statements made.

It's all totally provable in usage and this is where the casinos figure out at what level to cap a game or to add another variable to offset the odds.

So the question that 'does your odds improve after 100 rolls of not getting a 6 of getting one if you keep rolling' is yes, however; the odds of getting it on a certain # roll are a not improved.

 

[alkemyst]

Originally posted by: spidey07
Kyteland,

I don't keep real statistics on my play but I DO have to keep records of winnings/losses for tax purposes. I'll keep using my ESP.

Professional handicappers (I'm just an amateur) are meticulous about the payout percentages and risk/reward. Heck that's all they are, statisticians.

But you can tell when a craps table is hot and when it's going bad. Same with blackjack. Bet big when it's good, and back down when it's not or go to a different table. Same with the ponies, if using one strategy isn't working and you see another trend for the day then go to that trend.

Thing is most will call a table 'hot' based on a couple pay outs. Everyone gets lucky, the pros are not playing on luck.

I am not sure where you are trying to go. You could just be lucky and that did not have anything to factor in to whether others can duplicate your efforts.

 

[silverpig]

Originally posted by: alkemyst

Originally posted by: silverpig
Uh no, this is completely wrong. His original statement was correct. Your odds of rolling a 6 do not ever change. His original statement is something that people think a lot. "I've had 3 girls in a row, so the odds are my next one will be a boy!" That's exactly the same statement.

You can't take past independent events and use them to predict the outcome of the next independent event.

The odds of not rolling a 6 in 100 tries, then rolling a 6 on the 101st are the exact same as not rolling a 6 in 100 tries, then rolling a 2 on the 101st.

For one having a girl vs boy is not really dependent on chance.

That said you have stated what I already explained.

The odds of getting a 6 in 101 rolls improves each time that the event does not happen.

The odds of rolling a 6 on the 101st roll is a separate condition.

However; in gambling you are betting on occurances over trials...each event is not independent at all.

I don't think anyone here is understanding how the math really works in these situations based on the absurd statements made.

It's all totally provable in usage and this is where the casinos figure out at what level to cap a game or to add another variable to offset the odds.

So the question that 'does your odds improve after 100 rolls of not getting a 6 of getting one if you keep rolling' is yes, however; the odds of getting it on a certain # roll are a not improved.

The boy/girl thing, for the purposes of statistical illustration is purely 50/50 chance. In reality it's pretty close to 50/50, of course ignoring hermaphrodites etc.

The answer is still no. You odds don't change at all. What you are probably trying to say is that your odds of rolling a six in N rolls approaches 1 as N -> infinity. So this is true, but it still has no bearing on the fact that someone saying "I haven't rolled a 6 in 100 rolls, therefore I'm due to roll one soon!" is completely wrong.

His original statement still stands. You're talking about something else.

 

[DBL]

Originally posted by: alkemyst

Originally posted by: silverpig
Uh no, this is completely wrong. His original statement was correct. Your odds of rolling a 6 do not ever change. His original statement is something that people think a lot. "I've had 3 girls in a row, so the odds are my next one will be a boy!" That's exactly the same statement.

You can't take past independent events and use them to predict the outcome of the next independent event.

The odds of not rolling a 6 in 100 tries, then rolling a 6 on the 101st are the exact same as not rolling a 6 in 100 tries, then rolling a 2 on the 101st.

For one having a girl vs boy is not really dependent on chance.

That said you have stated what I already explained.

The odds of getting a 6 in 101 rolls improves each time that the event does not happen.

You really complicate a simple statement. The problem is you not being clear. The bold statement can be understood in two different ways. However in both cases, it is wrong.

1) The bold statement is wrong if we consider 101 rolls to be one event. If the odds of rolling a 6 in 101 rolls are 99.999% (an estimate) and no 6's are rolled in the first 101 rolls, then the odds of rolling a 6 in another 101 rolls is still 99.999%. No odds have improved.

2) The statement is still wrong if we consider each one of the 101 rolls to be an event. Each 101 events have the same probability (1/6) of rolling a 6. As each roll comes up w/o a 6 being rolled, the next roll (separate event) still has a 1 in 6 chance of coming up as a 6. No odds have improved.

 

[Kev]

Originally posted by: aplefka

Originally posted by: Crusty
You go to Casinos to have fun, not win money. If you think otherwise you are an idiot.

Call me an idiot, but I've done pretty well for myself with poker. And I never expect anything less than to win money. Are you going to tell me that poker is all luck too?

The only reason you can win at poker is by playing against people who are not as skilled as you. If you were to play against people with equal skill then the winner would be 100% based on luck.

 

[Kyteland]

Originally posted by: spidey07
Kyteland,

I don't keep real statistics on my play but I DO have to keep records of winnings/losses for tax purposes. I'll keep using my ESP.

Professional handicappers (I'm just an amateur) are meticulous about the payout percentages and risk/reward. Heck that's all they are, statisticians.

But you can tell when a craps table is hot and when it's going bad. Same with blackjack. Bet big when it's good, and back down when it's not or go to a different table. Same with the ponies, if using one strategy isn't working and you see another trend for the day then go to that trend.

That strategy will definitely maximize your happiness while playing the game, but it will not affect your long term profits/loss.

Do you also believe in the "third base" when playing blackjack? Long term that person has no effect on you. If you believe that they're throwing the table off because of their play you should definitely find another table. You'll be much happier. It won't change anything though, and I can prove it.

 

[DBL]

Originally posted by: Kev

Originally posted by: aplefka

Originally posted by: Crusty
You go to Casinos to have fun, not win money. If you think otherwise you are an idiot.

Call me an idiot, but I've done pretty well for myself with poker. And I never expect anything less than to win money. Are you going to tell me that poker is all luck too?

The only reason you can win at poker is by playing against people who are not as skilled as you. If you were to play against people with equal skill then the winner would be 100% based on luck.

True but an important part of winning is to make sure that you are playing at a table with enough players of a lower skill level than yourself.

And if the whole table was of equal skill level, the only winners would be the dealer (tips) and the house (over enough hands of course). All the players will go broke. This is highly unlikely of course.

 

[spidey07]

Originally posted by: Kyteland

Originally posted by: spidey07
Kyteland,

I don't keep real statistics on my play but I DO have to keep records of winnings/losses for tax purposes. I'll keep using my ESP.

Professional handicappers (I'm just an amateur) are meticulous about the payout percentages and risk/reward. Heck that's all they are, statisticians.

But you can tell when a craps table is hot and when it's going bad. Same with blackjack. Bet big when it's good, and back down when it's not or go to a different table. Same with the ponies, if using one strategy isn't working and you see another trend for the day then go to that trend.

That strategy will definitely maximize your happiness while playing the game, but it will not affect your long term profits/loss.

Do you also believe in the "third base" when playing blackjack? Long term that person has no effect on you. If you believe that they're throwing the table off because of their play you should definitely find another table. You'll be much happier. It won't change anything though, and I can prove it.

nah, third base is where I always play anyway. Get to see more cards.

 

[alkemyst]

Originally posted by: DBL

You really complicate a simple statement. The problem is you not being clear. The bold statement can be understood in two different ways. However in both cases, it is wrong.

1) The bold statement is wrong if we consider 101 rolls to be one event. If the odds of rolling a 6 in 101 rolls are 99.999% (an estimate) and no 6's are rolled in the first 101 rolls, then the odds of rolling a 6 in another 101 rolls is still 99.999%. No odds have improved.

2) The statement is still wrong if we consider each one of the 101 rolls to be an event. Each 101 events have the same probability (1/6) of rolling a 6. As each roll comes up w/o a 6 being rolled, the next roll (separate event) still has a 1 in 6 chance of coming up as a 6. No odds have improved.

In case one that would be a 202 roll cycle...the odds would continue to improve.

Your case number two is exactly what I have stated already more than once.

Either way in practice you would see this, it's obvious you're not understanding the way this REALLY works in a casino while gambling.

 

[jman19]

Originally posted by: aplefka

Originally posted by: Crusty
You go to Casinos to have fun, not win money. If you think otherwise you are an idiot.

Call me an idiot, but I've done pretty well for myself with poker. And I never expect anything less than to win money. Are you going to tell me that poker is all luck too?

No, there is skill in poker - the long term "luck" of all players is the same, it is the skill component that decides who wins more.

 

[alkemyst]

Originally posted by: silverpig

The boy/girl thing, for the purposes of statistical illustration is purely 50/50 chance. In reality it's pretty close to 50/50, of course ignoring hermaphrodites etc.

The answer is still no. You odds don't change at all. What you are probably trying to say is that your odds of rolling a six in N rolls approaches 1 as N -> infinity. So this is true, but it still has no bearing on the fact that someone saying "I haven't rolled a 6 in 100 rolls, therefore I'm due to roll one soon!" is completely wrong.

His original statement still stands. You're talking about something else.

1) having a boy or girl is not a 50% event and has nothing to do with hermaphrodites really. It has to do with what the father carries genetically. Some fathers can only have boys or girls. The proper analogy is the coin toss...heads vs tails.

2) if you are at a casino and have not rolled a 6 in 100 rolls...your odds will continue to improve with further rolls even if days go by between those rolls. No one is understanding how this works...the only way you'd be back to 1/6 odds is if you were requiring it to happen on X roll.

This is undeniable. If you do not believe me go ask someone either in your math department or any math site.

 

[Kev]

Originally posted by: alkemyst

Originally posted by: silverpig

The boy/girl thing, for the purposes of statistical illustration is purely 50/50 chance. In reality it's pretty close to 50/50, of course ignoring hermaphrodites etc.

The answer is still no. You odds don't change at all. What you are probably trying to say is that your odds of rolling a six in N rolls approaches 1 as N -> infinity. So this is true, but it still has no bearing on the fact that someone saying "I haven't rolled a 6 in 100 rolls, therefore I'm due to roll one soon!" is completely wrong.

His original statement still stands. You're talking about something else.

1) having a boy or girl is not a 50% event and has nothing to do with hermaphrodites really. It has to do with what the father carries genetically. Some fathers can only have boys or girls. The proper analogy is the coin toss...heads vs tails.

2) if you are at a casino and have not rolled a 6 in 100 rolls...your odds will continue to improve with further rolls even if days go by between those rolls. No one is understanding how this works...the only way you'd be back to 1/6 odds is if you were requiring it to happen on X roll.

This is undeniable. If you do not believe me go ask someone either in your math department or any math site.

What the hell are you talking about? Each roll is completely independent of the last, and with each roll you have a 1/6 chance of rolling a 6. The odds are 1 in 6 every single time.

If you see that the roulette table has landed on black 8 times in a row do you bet on red?

 

[preslove]

Originally posted by: Kev


What the hell are you talking about? Each roll is completely independent of the last, and with each roll you have a 1/6 chance of rolling a 6. The odds are 1 in 6 every single time.

Well, if you roll the die 100 times and 6 doesn't come up once, then a reasonable person would assume that the die is somehow loaded or warped & therefore bet against a 6 coming up

 

[jman19]

Originally posted by: alkemyst

Originally posted by: silverpig

The boy/girl thing, for the purposes of statistical illustration is purely 50/50 chance. In reality it's pretty close to 50/50, of course ignoring hermaphrodites etc.

The answer is still no. You odds don't change at all. What you are probably trying to say is that your odds of rolling a six in N rolls approaches 1 as N -> infinity. So this is true, but it still has no bearing on the fact that someone saying "I haven't rolled a 6 in 100 rolls, therefore I'm due to roll one soon!" is completely wrong.

His original statement still stands. You're talking about something else.

1) having a boy or girl is not a 50% event and has nothing to do with hermaphrodites really. It has to do with what the father carries genetically. Some fathers can only have boys or girls. The proper analogy is the coin toss...heads vs tails.

2) if you are at a casino and have not rolled a 6 in 100 rolls...your odds will continue to improve with further rolls even if days go by between those rolls. No one is understanding how this works...the only way you'd be back to 1/6 odds is if you were requiring it to happen on X roll.

This is undeniable. If you do not believe me go ask someone either in your math department or any math site.

With a fair die, your odds of rolling a 6 are the same every roll.

 

[Kev]

Originally posted by: preslove

Originally posted by: Kev


What the hell are you talking about? Each roll is completely independent of the last, and with each roll you have a 1/6 chance of rolling a 6. The odds are 1 in 6 every single time.

Well, if you roll the die 100 times and 6 doesn't come up once, then a reasonable person would assume that the die is somehow loaded or warped & therefore bet against a 6 coming up

Yeah but I was going under the assumption that there was no tomfoolery involved.

 

[TuxDave]

Originally posted by: alkemyst
2) if you are at a casino and have not rolled a 6 in 100 rolls...your odds will continue to improve with further rolls even if days go by between those rolls. No one is understanding how this works...the only way you'd be back to 1/6 odds is if you were requiring it to happen on X roll.

This is undeniable. If you do not believe me go ask someone either in your math department or any math site.

You sir have a case of gambler's fallacy. Either that or you are doing a pretty crappy job at explaining yourself. This is the truth and you can go to anyone in your math department or any math site to confirm.


(all coins are fair)
Truth)You just flipped a coin 100 times and got 100 heads. You still have a 50/50 chance of getting tails on the next flip.

Truth) Person A just flipped a coin 100 times and got 100 heads. Person B just started playing. The chance of getting a tail on the next flip is equal for both people.

Truth) Person A just flipped a coin 100 times and got 100 heads. Person B just started playing. The chance of getting a tail in the next X flips is equal for both people.

 

[JC86]

i'm surprised this thread has lasted this long. Looks like math and statistics really stokes the ATOT fire. I think regardless of how you wanna slice it, gambling is by design, a losing proposition that if you were to keep playing, eventually the odds will catch up to you. That being said, there are some betting methods that raise your odds of breaking even, or even winning a little bit. There is no point in trying to convince gamblers that there way of doing it is wrong no matter how strong of a mathematical or statistical case you can make.