Law of Averages...

pat venditto · 10-22-10, 04:31 AM

Still 50/50 chance on a coinflip regardless if you lost 200 times in a row. A large sample size would balance out the win/loss percentage to more realistics figures.

OmgUrMom · 10-22-10, 06:24 AM

if your up 4-0 odds are you will finish 7-3 most likely. each roll is independent of the other rolls but if you won/lost 200 50/50 s in a row then i would begin to think the game was rigged

UntilTheNDofTimE · 10-22-10, 06:29 AM

Originally posted by pat venditto

Still 50/50 chance on a coinflip regardless if you lost 200 times in a row. A large sample size would balance out the win/loss percentage to more realistics figures.

Sharpest post Ive ever seen you make, right on

UntilTheNDofTimE · 10-22-10, 06:31 AM

Read the "Law of Total Variance" wikipedia explains it very well here...

Law of total variance - Wikipedia

http://en.wikipedia.org/wiki/Law_of_total_variance

Although both variables are dependent on the coin. The sides themselves represent a independent variable, neither is dependent on itself. Therefore like patty said if the coin goes heads 5,000 times in a row it in no way chances the odds that it will land on tails.

UntilTheNDofTimE · 10-22-10, 06:33 AM

Belief that an event is "due" to happen: For example, "The roulette wheel has landed on red three consecutive times. The law of averages says it's due to land on black!" Of course, the wheel has no memory and its probabilities do not change according to past results. So even if the wheel has landed on red 10 consecutive times the probability that the next roll will be black is still 48.6% (assuming a European wheel with only one green zero: it would be exactly 50% if there were no green zero). Similarly, there is no statistical basis for the belief that a losing sports team is due to win a game or that lottery numbers which haven't appeared recently are due to appear soon. This sort of belief is called the gambler's fallacy.[1]

statnerds · 10-22-10, 07:16 AM

Originally posted by UntilTheNDofTimE

Similarly, there is no statistical basis for the belief that a losing sports team is due to win a game or that lottery numbers which haven't appeared recently are due to appear soon. This sort of belief is called the gambler's fallacy.[1]

But if the losing team is bad enough, they will be under priced by an over reactive market.

I'm thinking strictly NFL here, where no team is as good as everyone thinks, and no team is as bad as anyone thinks.

As for losing 4 straight, it might take another 700 rolls to correct a 4 roll variance.

UntilTheNDofTimE · 10-22-10, 07:46 AM

Yes but were talking coin flipping here not pricing off market numbers. But of course i agree with u there.

jjgold · 10-22-10, 09:36 AM

Everything comes to around 50%

Law of averages always holds true long term

UntilTheNDofTimE · 10-22-10, 05:40 PM

Originally posted by jjgold

Everything comes to around 50% Law of averages always holds true long term

But not for 1 individual spin, the past dosent represent the future on a true coin flip,

chilidog · 10-22-10, 05:46 PM

Originally posted by jjgold

Everything comes to around 50%

Law of averages always holds true long term

Not in the case of your spreadsheet ...

gryfyn1 · 10-22-10, 06:43 PM

Originally posted by jjgold

Everything comes to around 50%

Law of averages always holds true long term

Sure it does, but the long term only applies to an infinite number of results - any measurable sample is not affected by the law of averages.

Shonner · 10-22-10, 06:56 PM

Martingale is a losing proposition.

Do the math.

If you start with $100 and keep doubling up it is pretty realistic that you will eventually be broke.

I once played with this in excel using probabilities and I am convinced that the Martingale (keep doubling until you win) is by far the worst gambling strategy ever developed. You will go broke doing it.

Bet 1 = 100
Bet 2 = 200
Bet 3 = 400
Bet 4 = 800
Bet 5 = 1600
Bet 6 = 3200
Bet 7 = 6400
Bet 8 = 12800

That right there is 25.5K lost.

Have you ever won or lost 8 bets in a row? I have, many many of times.

Double your bet until you win and you're surely to go broke on a bad losing streak.

Sunde91 · 10-22-10, 07:16 PM

1)Law of Averages is a pseudo-theory supported by nothing but is still believed because of its intuitive appeal/practicality
2)Each event is independent. Losing a coin flip on heads 100 times doesn't make the next flip more likely to be tails, obvious.
3)Losing 100 in a row has a low probability, but if it's already happened it's after the fact, the next flip is still 50% like the rest.

As for the spread being 50%, I don't believe that at all. Theoretically, we say the spread is "50% odds", but practically, no way. It's just not possible to create a perfect 50/50 spread. It fluctuates with the market, there could be faulty statistics, injuries, the psyche of a team could be off, a team could be over/undervalued.

Let's say a game opens -3, but then the star QB is ruled out and it goes to +7. Was the -3 initially "50%", but the +7 is now the "true odds" of 50%? Based on what? How is it magically adjusted to the "right number"? And then what happens if the market it pounds the other side and it moves to +8? It's just ridiculous and arbitrary.

GiveMeaBJ · 10-23-10, 01:35 AM

I understand guys, I get it. Each independent roll/spin/coin flip is 50/50. But in the long term if it is supposed to even out shouldn't it be more likely that you will eventually win after each losing roll you make. Example, I bet $1 on each coin flip and lose 1000 times in a row. Now in the long term I should win 5 of those. I know the next flip is completely 50/50 but shouldn't it be coming out my way more often over the next 1000?

pat venditto · 10-23-10, 05:25 AM

Decent thread here some knowledge taught.

milwaukee mike · 10-23-10, 09:23 AM

Originally posted by GiveMeaBJ

I understand guys, I get it. Each independent roll/spin/coin flip is 50/50. But in the long term if it is supposed to even out shouldn't it be more likely that you will eventually win after each losing roll you make. Example, I bet $1 on each coin flip and lose 1000 times in a row. Now in the long term I should win 5 of those. I know the next flip is completely 50/50 but shouldn't it be coming out my way more often over the next 1000?

NO
every new spin/flip starts a new sample size

the coin/dice/etc has no memory

vegas loves the roulette number trackers cuz it gets people to look and say "wow it's come up red 6 times in a row it has to be black eventually" which in some ways is true but it can keep going red until you're broke

50/50 odds on a coin flip (assuming it is a fair coin and it is flipped fairly) no matter what happened in the past

eidolon · 10-23-10, 11:30 AM

Lets say you are on Mount Everest, and it is made out of 50% peanut butter and 50% jelly. You are hungry so you take a bite out of the mountain blind folded. You get a taste of peanut butter, and you take another bite, peanut butter again. Your whole life you are taking bites out of the mountain, but you will never come close to eating the whole mountain; so you could be around 1-1 for peanut butter to jelly ratio, but if you are lucky and enjoy jelly, maybe you will eat more jelly in your life time; or maybe you would end up eating more peanut butter in your life time, and hopefully you aren't allergic.