What are the odds of winning the lotto twice?

J K · 10-03-2011 04:59PM

robbie7730 wrote: »

Literacy.
1
a. Able to read and write.
b. Knowledgeable or educated in a particular field or fields.

2. Familiar with literature; literary.

3. Well-written; polished

Part b would suggest it is not purely about reading and writing.

The word is extended to other areas like computer-literate etc but no corresponding term exists. There is a perfectly good word in the english language already numeric. There is no necessity to say number literate, unless two words better than one.

keane2097 · 10-03-2011 05:00PM

J K wrote: »

Literacy is the wrong term,whether you like it or not.
Your post above where you credit yourself with intelligence and predicting the lotto numbers is hardly a reasonable argument. Someone above called you a Troll. You're either that or braindead. Well done young man.

I can ship you a dictionary out of my winnings if you like?

Srsly, the mark of a grown up is being able to admit when you're wrong, as you clearly are here.

At least have the good sense not to keep digging and embarrass yourself any further.

DoesNotCompute · 10-03-2011 05:06PM

The Govt should buy lotto tickets every week with what's left of the Pension Fund. 'Course the winnings would go directly to the IMF

fergalr · 10-03-2011 05:15PM

J K wrote: »

Literacy is the wrong term,whether you like it or not.
Your post above where you credit yourself with intelligence and predicting the lotto numbers is hardly a reasonable argument. Someone above called you a Troll. You're either that or braindead. Well done young man.

I mentioned good trolling skillz.
I actually only read one or two previous posts, and assumed keane2097 was trolling (not 'a troll', in the 'lives under bridge and is unpleasant' sense, but 'trolling' in the sense of trying to wind people up for a laugh).

Having read some of the earlier posts, I think its is even more likely, because keane2097 seems too smart to believe statements like this:

keane2097 wrote: »

No, I'm saying that the longer a number hasn't landed on in roulette, the more likely it is to come out in subsequent spins.

especially if he's been involved in this sort of gambling for a long time, as he claims.

I mean, this is exceedingly straightforward stuff.

1
Would we agree a roulette wheel is a mechanical device in which a ball lands in each numbered slot with effectively equal probability (1/37)? (I say effectively, because if the wheel isn't perfectly even, there might be small biases - but lets assume it is perfectly even, beyond the point of human measurement).

2
So the ball has a 1/37 chance of landing in each slot when you spin the wheel, for simple, mechanical reasons

3
So it doesn't matter how often the number came up before - its chance of coming up on this spin, is 1/37. For this to not be true, the machine would have to somehow 'remember' where the ball landed in previous spins. But it has no mechanical device to remember this.

Anyway, I think this is so obvious, that someone who argues persistently against it, and claims to have been involved in this field for a long time, is probably trolling.

keane2097 · 10-03-2011 05:39PM

fergalr wrote: »

I mentioned good trolling skillz.
I actually only read one or two previous posts, and assumed keane2097 was trolling (not 'a troll', in the 'lives under bridge and is unpleasant' sense, but 'trolling' in the sense of trying to wind people up for a laugh).

Having read some of the earlier posts, I think its is even more likely, because keane2097 seems too smart to believe statements like this:

especially if he's been involved in this sort of gambling for a long time, as he claims.

I'll take compliments anywhere I can get them, so ty :P

fergalr wrote: »

I mean, this is exceedingly straightforward stuff.

1
Would we agree a roulette wheel is a mechanical device in which a ball lands in each numbered slot with effectively equal probability (1/37)? (I say effectively, because if the wheel isn't perfectly even, there might be small biases - but lets assume it is perfectly even, beyond the point of human measurement).

2
So the ball has a 1/37 chance of landing in each slot when you spin the wheel, for simple, mechanical reasons

3
So it doesn't matter how often the number came up before - its chance of coming up on this spin, is 1/37. For this to not be true, the machine would have to somehow 'remember' where the ball landed in previous spins. But it has no mechanical device to remember this.

Anyway, I think this is so obvious, that someone who argues persistently against it, and claims to have been involved in this field for a long time, is probably trolling.

I agree with the stuff you're saying above. The thing is, over a large sample size each number will come out exactly as often as every other number.

The instances where a particular number doesn't come out for a large number of spins is what David Sklansky (esteemed gambling mathematician) refers to as "clumps of variance".

Over a large enough sample size, these clumps must eventually balance themselves out, i.e. if a ball doesn't come up for a large number of spins, there will be a corresponding number of spins where that particular ball appears unusually often.

If this wasn't the case, then there is no way a ball that doesn't come out for a large number of spins could ever "catch up" to the other balls in our larger sample, which of course we know it will.

The trick, therefore, is simply to spot these clumps of variance, and capitalise on them with appropriate wagers.

J K · 10-03-2011 05:49PM

keane2097 wrote: »

I can ship you a dictionary out of my winnings if you like?

Srsly, the mark of a grown up is being able to admit when you're wrong, as you clearly are here.

At least have the good sense not to keep digging and embarrass yourself any further.

I don't have to dig as I'm not in a hole. I can speak english properly unlike yourself.

Someone who believes they can predict the lotto based on the pattern of numbers drawn in previous games, thinks that I should be embarrassed?:D
At least you have a sense of humour.

fergalr · 10-03-2011 05:56PM

keane2097 wrote: »

I'll take compliments anywhere I can get them, so ty :P

I agree with the stuff you're saying above. The thing is, over a large sample size each number will come out exactly as often as every other number.

The instances where a particular number doesn't come out for a large number of spins is what David Sklansky (esteemed gambling mathematician) refers to as "clumps of variance".

Over a large enough sample size, these clumps must eventually balance themselves out, i.e. if a ball doesn't come up for a large number of spins, there will be a corresponding number of spins where that particular ball appears unusually often.

If this wasn't the case, then there is no way a ball that doesn't come out for a large number of spins could ever "catch up" to the other balls in our larger sample, which of course we know it will.

The trick, therefore, is simply to spot these clumps of variance, and capitalise on them with appropriate wagers.

I see why you'd say that all right.
I do agree that all those arguments are superficially convincing.

Unfortunately, as others have said, what you are describing is, in fact, the gamblers fallacy:
http://en.wikipedia.org/wiki/Gambler%27s_fallacy

Now, it's apparent you aren't going to be convinced of that, from just having an additional person - me - tell you its the gamblers fallacy. In some ways this is fair enough - you shouldn't necessarily believe something just because a lot of people say it.

The wikipedia article - which was already mentioned - is ok; many other websites do a good job at explaining the fallacy.

But this is a subtle thing to reason about, and there's a lot of back on forth on this thread already - as its subtle, I doubt you'll find analytical arguments I might make appealing.

So, how could I convince you?

Could we agree on an experiment, and an outcome, that if demonstrated, would convince you?

I don't suppose you know how to program a computer?

I could easily write a computer program, where I have a random number generator that picks one number randomly from 37 - this is what the roulette wheel does, right?
If I generate a big series of random numbers, the observed frequency of each number having come up will converge to 1/37.

I could write a program that runs this process many many times (say, millions).
And as it runs, I could simulate two betting strategies.
In the first betting strategy, a number would be randomly picked to bet on.
In the second betting strategy, the number that had least frequently come up so far, in the series, would be bet on.

If, after many runs, the second betting strategy did no better than the first one, would you be convinced that your reasoning is erroneous?

Can you program a computer? You could test this yourself?
Or, if you know how basic computer programming, but aren't familiar with writing code like this, I could write and send you the code, and you could run it yourself?

I mean, I could just write it, run it, and tell you the results - I know what they'll be ahead of time, but will do the experiment anyway! - but you might not believe me! :-)

What do you think?

A good test of whether we should believe something, that we believe, is to think about what would be required to convince us otherwise.
What would convince you otherwise?

If my computer program shows the second strategy does much better than the first - well, I'll start off looking for bugs in the program - extraordinary claims need extraordinary evidence, and all that - but I'll certainly start paying attention to whats going on! :-)
If this keeps happening, across a series of experiments, I'll be booking tickets to vegas. (But I know it won't...)

keane2097 · 10-03-2011 05:57PM

J K wrote: »

I don't have to dig as I'm not in a hole. I can speak english properly unlike yourself.

Someone who believes they can predict the lotto based on the pattern of numbers drawn in previous games, thinks that I should be embarrassed?:D
At least you have a sense of humour.

You're clearly wrong.

You decided to try and look clever by claiming that a phrase of mine was incorrect, and since you been shown that it is a perfectly acceptable, widely used and widely understood phrase you decided to try and save face with some fairly feeble exploration into the minutiae of the English language.

You, however, completely failed to show in any way at all that the phrase I originally used was anything but acceptable, which obv is to be expected.

Your line now is attempting to undermine my position on the issue by referring to my position on a completely unrelated issue. In English, this is referred to as a "red herring", or a deliberate attempt to divert attention.

It was your only real option when your original point had been shown to be incorrect (besides gracefully admitting your mistake obv), but unfortunately it was pretty transparent.

Never mind, I'm sure you'll find another thread where you can outsmart someone on the internetz and feel good about yourself.

Good luck with it.

ya-ba-da-ba-doo · 10-03-2011 06:00PM

The question was 'what are the chances of winning the lotto twice' , not 'twice in a row'. So it is not the odds of winning it once squared as some people are saying, it is the odds of winning it once multiplied by two.

10-03-2011 06:02PM

keane2097 wrote: »

I agree with the stuff you're saying above. The thing is, over a large sample size each number will come out exactly as often as every other number.

Incorrect. Over a large enough sample size, the difference between the numbers becomes insignificant as a percentage of the total sample. That's what the law of large numbers tells us

keane2097 wrote: »

The instances where a particular number doesn't come out for a large number of spins is what David Sklansky (esteemed gambling mathematician) refers to as "clumps of variance".

And yet he plays poker and blackjack, games which mix skill and luck, not roulette (or the lotto)

keane2097 wrote: »

Over a large enough sample size, these clumps must eventually balance themselves out, i.e. if a ball doesn't come up for a large number of spins, there will be a corresponding number of spins where that particular ball appears unusually often.

If this wasn't the case, then there is no way a ball that doesn't come out for a large number of spins could ever "catch up" to the other balls in our larger sample, which of course we know it will.

The trick, therefore, is simply to spot these clumps of variance, and capitalise on them with appropriate wagers.

Coin-flip. Ten heads in a row come up (a "clump" of variance)

What are the odds of tails coming up next?
What are the odds of tails coming up in the next 10 flips?
What are the odds of heads coming up for another ten flips in a row?

Do any of these odds differ in any way from a coin which doesn't have a history of 10 heads?

Oh and feel free to replace 10 with 100 or 1,000 if you think the sample sized is too small. And if you're going to say a coin flip isn't relevant, at least try and explain why, and how "clumps of variance" don't apply to it

keane2097 · 10-03-2011 06:04PM

28064212 wrote: »

And yet he plays poker and blackjack, games which mix skill and luck, not roulette (or the lotto)

I said he's a gambling mathematician.

10-03-2011 06:13PM

keane2097 wrote: »

I said he's a gambling mathematician.

And? You're claiming he has this theory about clumps of variance in games of random chance. I'm pointing out that he doesn't appear to apply it, he plays games with a skill factor. He hasn't made his money off the roulette wheel

J K · 10-03-2011 06:16PM

keane2097 wrote: »

You're clearly wrong.

No, you are 100% wrong. You don't know what literacy means.

fergalr · 10-03-2011 06:20PM

keane2097 wrote: »

I said he's a gambling mathematician.

He is indeed.

And while we are appealing to authorities, he is the same David Sklansky , who, in his book 'Getting the best of it', on p295 onwards, as I just searched for in google books, writes:

Before discussing the law of averages, I must bring in the concept of independent events. It is extremely important. In fact, when a gambler places a bet, he is usually betting on independent events, whether he knows it or not. The definition of the term is almost self-explanatory. Each outcome depends on no other previous outcome. Where the roulette ball lands does not depend in any way on where it landed before. The chances of rolling snake eyes reamins at 1 out of 36 regardless of all previous rolls. More to the point; when you are betting on independent events, the fact that you won or lost your last bet has nothing to do with whether you will win the next.

[...]

There are two main techniques used in most fallacious systems. The first of them has to do with riding streaks. [...] The theory is that luck runs in streaks - hot and cold. As it happens this is true, but only after the fact.

[...]
Those who would get off a hot number because the streak can't continue never have a legitimate basis fo their action. [...] A good example of this is playing 'rushes in poker', i.e. playing more loosely and aggresively after winning a couple of pots in a row. Those who think that this rush has somehow increased their chances of catching better cards for the next few hands are dead wrong.

keane2097 · 10-03-2011 06:26PM

28064212 wrote: »

And? You're claiming he has this theory about clumps of variance in games of random chance. I'm pointing out that he doesn't appear to apply it, he plays games with a skill factor. He hasn't made his money off the roulette wheel

I'm not claiming anything, he has publicly talked about it numerous times. Go listen to the 2+2 podcast if you want.

Just because he doesn't apply it doesn't mean a whole lot.

He's also a lolbad poker player, and hasn't made the bulk of his money in any of the pursuits he most frequently writes on, he's made them on books.

Like this one.

Your logic is typically terrible by the way. When Einstein postulated his theories of relativity I doubt his intention had anything to do with Two and a Half Men, but that didn't stop it leading to the invention of the television.

keane2097 · 10-03-2011 06:28PM

J K wrote: »

No, you are 100% wrong. You don't know what literacy means.

Why don't you tell me what you think it means, and I can continue to point out (in excruciating detail) how much of an idiot you're making yourself appear?

keane2097 · 10-03-2011 06:31PM

fergalr wrote: »

He is indeed.

And while we are appealing to authorities, he is the same David Sklansky , who, in his book 'Getting the best of it', on p295 onwards, as I just searched for in google books, writes:

I don't have any particular issues with any of that (although I'll admit I didn't read it all that thoroughly).

I would point out that that book is pretty old, and his theories that I'm talking about are probably a decade more up to date.

evercloserunion · 10-03-2011 06:32PM

Wibbs wrote: »

Apparently not that remote. I recall reading somewhere before that you've a slightly higher chance of winning the lotto if you've won it before. I know someone who won it twice myself. Not huge money, but a couple of 100 thou(I know that's loads, but not, "Im buying a small island in the far east, building my own rockets and if that cnut Bond shows up, I'm shooting him in the face on sight" money)

Is that just because you have enough money to buy tons of tickets or what?

10-03-2011 06:33PM

keane2097 wrote: »

Your logic is typically terrible by the way. When Einstein postulated his theories of relativity I doubt his intention had anything to do with Two and a Half Men, but that didn't stop it leading to the invention of the television.

It's statements like this which give me hope that maybe you are just trolling. Fine, forget the Sklansky stuff

Coin-flip. Ten heads in a row come up (a "clump" of variance)

What are the odds of tails coming up next?
What are the odds of tails coming up in the next 10 flips?
What are the odds of heads coming up for another ten flips in a row?

Do any of these odds differ in any way from a coin which doesn't have a history of 10 heads?

Oh and feel free to replace 10 with 100 or 1,000 if you think the sample sized is too small. And if you're going to say a coin flip isn't relevant, at least try and explain why, and how "clumps of variance" don't apply to it

keane2097 · 10-03-2011 06:35PM

28064212 wrote: »

It's statements like this which give me hope that maybe you are just trolling. Fine, forget the Sklansky stuff

Coin-flip. Ten heads in a row come up (a "clump" of variance)
What are the odds of tails coming up next?

What are the odds of tails coming up in the next 10 flips?

What are the odds of heads coming up for another ten flips in a row?

Do any of these odds differ in any way from a coin which doesn't have a history of 10 heads?

Oh and feel free to replace 10 with 100 or 1,000 if you think the sample sized is too small. And if you're going to say a coin flip isn't relevant, at least try and explain why, and how "clumps of variance" don't apply to it

Do you genuinely not understand my point of view at this stage?

fergalr · 10-03-2011 06:39PM

keane2097 wrote: »

I don't have any particular issues with any of that (although I'll admit I didn't read it all that thoroughly).

I would point out that that book is pretty old, and his theories that I'm talking about are probably a decade more up to date.

Well, its just it says the exact opposite to what you are saying here.
And the basic rules of probability haven't changed in the last few decades.
Independent events are still independent. Also, in fairness, its a lot clearer, and more sophisticated, than the mathematical reasoning you've discussed so far in this thread, no?
Perhaps you've misunderstood what you thought he wrote, more recently?

You are free to believe whatever you want, of course.
But things must be starting to look at little ropey now, right? Like, you've a lot of people pointing out you are making the gamblers fallacy. Many of them seem to know a lot of maths; and there's the author you were citing as an expert in this area, writing the exact opposite to what you are saying.

This must make you think 'hmmm' - no?

I mean, if not, thats your business - but if I were you, I'd be getting a bit cautious now, and deciding to do a bit more reading before gambling money on a system like this.

If you don't want to go down the computer route, you could literally spend a few days tossing coins, or spinning wheels, and seeing what would happen if you bet money when you thought you were in one of those 'clumps'.
If you look back on it after the fact, after a couple of 1000 trials, you won't find yourself winning big.

That's pretty boring way to spend a few days, but if you aren't willing to believe other simulations, it might be worth it, if you are going to bet money on this sort of thing.

10-03-2011 06:40PM

keane2097 wrote: »

Do you genuinely not understand my point of view at this stage?

Certainly not your system, given you haven't explained anything about it (other than to say it's vaguely related to the notion that a ball is 'due' to come up after it hasn't been out in a while)

Over a large enough sample size, these clumps must eventually balance themselves out, i.e. if a ball doesn't come up for a large number of spins, there will be a corresponding number of spins where that particular ball appears unusually often.

Replace ball with tails and spins with coin-flips. How is the scenario I posted above not directly applicable?

keane2097 · 10-03-2011 06:46PM

fergalr wrote: »

Well, its just it says the exact opposite to what you are saying here.
And the basic rules of probability haven't changed in the last few decades.
Independent events are still independent. Also, in fairness, its a lot clearer, and more sophisticated, than the mathematical reasoning you've discussed so far in this thread, no?
Perhaps you've misunderstood what you thought he wrote, more recently?

You are free to believe whatever you want, of course.
But things must be starting to look at little ropey now, right? Like, you've a lot of people pointing out you are making the gamblers fallacy. Many of them seem to know a lot of maths; and there's the author you were citing as an expert in this area, writing the exact opposite to what you are saying.

This must make you think 'hmmm' - no?

I mean, if not, thats your business - but if I were you, I'd be getting a bit cautious now, and deciding to do a bit more reading before gambling money on a system like this.

If you don't want to go down the computer route, you could literally spend a few days tossing coins, or spinning wheels, and seeing what would happen if you bet money when you thought you were in one of those 'clumps'.
If you look back on it after the fact, after a couple of 1000 trials, you won't find yourself winning big.

That's pretty boring way to spend a few days, but if you aren't willing to believe other simulations, it might be worth it, if you are going to bet money on this sort of thing.

I don't think he's said anything like the direct opposite of what I've been saying.

I think you may be guilty of applying a context to his statements that isn't actually there.

The computer simulation idea is pretty cool, but I don't think it would be exactly the same as what I do (and have been doing, with a fairly large profit, for a while now).

J K · 10-03-2011 06:54PM

keane2097 wrote: »

Why don't you tell me what you think it means, and I can continue to point out (in excruciating detail) how much of an idiot you're making yourself appear?

You're starting to repeat yourself. Why don't you ask Robbie to help you with an answer again.

keane2097 · 10-03-2011 07:00PM

J K wrote: »

You're starting to repeat yourself. Why don't you ask Robbie to help you with an answer again.

This is another oft-used tactic in a losing battle, referred to as avoidance.

It really would make you look like less of an idiot if you just disengaged with the thread entirely.

We could always go back and have you explain why the phrase "mathematical literacy" sticks in your craw, as your efforts in this direction to date have been, shall we say, less than adequate?

fergalr · 10-03-2011 07:01PM

keane2097 wrote: »

I don't think he's said anything like the direct opposite of what I've been saying.

No?

What about:

keane2097 wrote: »

The instances where a particular number doesn't come out for a large number of spins is what David Sklansky (esteemed gambling mathematician) refers to as "clumps of variance".

Over a large enough sample size, these clumps must eventually balance themselves out, i.e. if a ball doesn't come up for a large number of spins, there will be a corresponding number of spins where that particular ball appears unusually often.

Sklansky wrote:

Each outcome depends on no other previous outcome. Where the roulette ball lands does not depend in any way on where it landed before.

It's pretty clear there's a contradiction there.

If you don't see one, then I don't think you are equipped to reason about this sort of thing at all. (or are trolling

)

I would advise you not to gamble.

10-03-2011 07:01PM

J K wrote: »

You're starting to repeat yourself. Why don't you ask Robbie to help you with an answer again.

Knock it off. He might be wrong about many things, but the phrase "mathematical literacy" isn't one of them. Numeracy is a "better" word for it, but mathematical literacy is not invalid

keane2097 · 10-03-2011 07:03PM

28064212 wrote: »

Knock it off. He might be wrong about many things, but the phrase "mathematical literacy" isn't one of them. Numeracy is a "better" word for it, but mathematical literacy is not invalid

Don't worry, I'm perfectly capable of dealing with that cretin on my own

Bruthal · 10-03-2011 07:04PM

J K wrote: »

You're starting to repeat yourself. Why don't you ask Robbie to help you with an answer again.

No one asked me for any help with anything, i pointed out that the term numerical literacy is not an incorrect term.

keane2097 · 10-03-2011 07:06PM

fergalr wrote: »

No?

What about:
It's pretty clear there's a contradiction there.

If you don't see one, then I don't think you are equipped to reason about this sort of thing at all. (or are trolling )

I would advise you not to gamble.

I'm not saying it depends on the previous outcome either.

I'm just saying that, over a large sample, there will be periods where a ball doesn't come out for a long time, and periods when it will come up (relatively) frequently.

The periods of frequency don't depend on the barren periods or vice versa, but both will still occur.

What are the odds of winning the lotto twice?

Comments