basic question but need good explanation please

[poker12]

I have a friend who I am having a very hard time explaining basic probability to. He plays 10nl poker and he thinks that if you flip a coin and the first time it is heads, the next time there is a greater chance it will be tails.

Now he thinks it is a more advanced concept(he's an engineer) that you need to look at the entire sample size and if you flip the coin and 50 times in a row it is heads, then he will bet money that tails is coming soon. Here's a little excerpt from our chat:

[20:34] p: sure, statistic do work over a large pool, but it is applied in small sub-systems as well.
[20:34] p: and sure if the first two times you flip heads, I will put my money that the next one will be tails.
[20:34] p: an so forth.
[20:34] p: I keep seeing hearts being flopped, that soon I going to start seeing spades or less hearts.

[PairTheBoard]

You could ask him by what mechanism the coin knows what it's suppose to do according to what happened previously. However he'll probably counter sublimely with more statistical mumbo jumbo and at some point you will probably have to conclude you are arguing with a fool and let it go.

PairTheBoard

[1o BoY]

Tell him to use that theory playing roulette

[Gonso]

[ QUOTE ]
You could ask him by what mechanism the coin knows what it's suppose to do according to what happened previously. However he'll probably counter sublimely with more statistical mumbo jumbo and at some point you will probably have to conclude you are arguing with a fool and let it go.

[/ QUOTE ]

Accurate and to the point.

If he doesn't get it, and he really believes in what he's saying, set up some proposition bets. Flip a coin 10 times, then see if he'll lay you $3-to-$2 that he can predict the next flip and the 99 flips after (he's free to make his choice just before the flip).

[poker12]

Does this make sense? maybe a little?

I say that each situation is independent (i.e. it doesnt matter what the flop was before). and he says that they are dependent in a statistical pool.


[15:06] b: Although statistis are best examined with large samples of data; Once with the appropriate statistical information, one could make an educated guess on future events; be it 1,2,3....n.
[15:07] b: So if the flop is A,K,K twice in a row, I could safely say it will not happen again.
[15:09] b: if I see quads, and three of the cards are on the table. I could safely say that I will not see that again in the next 100 hands.......and if I do see it again, I will be more sure that it will not happen again.
[15:15] b: it is all basis on making educated guesses. I believe that observing the flop and building information, one could decied to play additional hands because they predict that the flop will be on the low end, high end, full of a suit, due for a str8..........ect. But the flop is a lot more trickier than a coin flip, becasue there are many more possibilities than a coin flip.

[T50_Omaha8]

[ QUOTE ]
he's an engineer

[/ QUOTE ]Sounds like engineering logic to me.

Standard

[SpaceAce]

Almost everything your friend says in those chats is ridiculous. It does not matter if an eight flops (or doesn't flop) every hand for a thousand hands, playing eights does not become more or less profitable. Likewise, it simply DOES NOT MATTER if a coin (assuming it is a fair coin) comes up heads 234,000 times in a row, the chance of the next flip coming up heads is still 50%.

Ask your friend this: If the coin comes up heads ten times in a row and then he puts the coin away for a year, when he pulls that coin out of his desk drawer a year later, does it still owe him some tails? How is the coin "remembering" it's flips? How long is this mysterious coin memory good for? A day? A month? Eternity?

Ask your friend this as well: how does he know how many times the coin came up tails before the coin came into his possession? How does your friend know that the coin didn't simply "owe" the world a few heads to balance out a deficit created by the last guy who flipped the coin? How does your friend know that the lack of eights in flops he has witnessed isn't just the balancing out from a previous glut of eights at that table?

Your friend is all washed up on this one.

SpaceAce

[SplawnDarts]

This kind of [censored] cannot be "explained" to someone who doesn't want to learn. You're wasting your time.

Your best bet for teaching him if you really want to is to bust his bad-at-math ass flipping coins. Start flipping a coin, and every time it comes up the same way twice (or three times or whatever), let him bet 11 to 10 that it won't do it again - that's more than enough vig to cover the bias in any standard US coin.

When he's broke, he might be more receptive to basic probability theory.

[PairTheBoard]

It wouldn't be too hard to write a program that "flips a coin" and tracks how often heads and tails come up after a run of 2 or more of the same type. To make it worth your while, ask him if he'd be willing to bet $1000 with you - your side of the bet being that when run with enough trials the program will prove him wrong. Of course you need to set precise conditions for what constitutes a proof ahead of time.

Also, since he seems to know a little about statistics you might ask him if he's ever heard of the concept of "Independent Trials". You might also mention that some pretty smart people here at 2+2 are unanimous in the opinion that he is an idiot.

PairTheBoard

[poker12]

Doesnt what he says in bold contradict the sentence before???? If it is 50/50 everytime how can it be predictable.


[18:25] b: First I feel that you have missed interpreted my information, and that it was posted out of context: Firstly, I was not trying to prove that a flip of a coin does not have a 50-50 chance that it will be heads or tails. It will always be 50-50 when flipping the coin. But I still believe due to the preceding events I could provide an educated guess as to what the next flip may be.
[18:25] b: Have you seen the comedy,"Rosencrantz and Guildenstern Are Dead (1991)," the opening is about the coin flip. One was loosing all his coins to his partner because he kept flipping tails, one could believe it and thought they may be stuck in time.

[18:51] b: It starts with character A, having a bag full of money, and character B having an almost empty bag of money.
[18:51] b: A says ," 98 heads"
[18:51] b: B continues on the idea that it is impossible, and gives A another coing.
[18:51] b: A says,"99 heads"
[18:52] b: B continues to pontificate the situation they are in, and gives another coin.
[18:52] b: A says, "100 heads"
[18:52] b: B continues to rationalize that they must be trapped in time.
[18:52] b: 101 heads.
[18:53] b: 102 heads
[18:53] b: .........so forth.
[18:53] b: the dialogue is interesting.
[18:56] r: \ok
[18:57] b: it is worth watching, it was well illustrated that the coin does have a 50/50 chance, and that it could also just keep flipping heads. And that is why he concluded that they must be an anomalistic situation, or time warp. because in reality we know that tails should be coming soon.
[18:57] b: even in out little experiment, I knew we were due for a head.

(we flipped 10 times 6 tails 4 heads...and he guessed the next flip would be head based on those results, so he proved me wrong)

[18:57] r: why in reality do we know that tails should be coming soon
[18:58] b: well, flip a hundred heads in a row.
[18:58] b: or flip 15 heads in a row.
[18:58] b: some time there is written theory, and application of the theory.
[18:58] b: I bet you cannot flip 15 heads in a row.