The gamblers fallacy - page 2. | Bitcointalksearch.org

monsterer

legendary

Activity: 1008

Merit: 1007

Quote from: Light on March 06, 2014, 07:01:29 AM

In layman's terms with the use of math, basically the past has no bearing or influence on your future results (for completely unrelated events ie dice rolls). So don't believe that just because something is so unlikely that it cannot happen when you lose 28 times in a row and try for a 29th.

That's really the reason I stopped dice betting, the more you play the more you accentuate the house edge against you, whereas in something like sports betting/poker you have greater control over your results and hence there is a correlation between your skill and your earnings. In effect you can improve your probability of winning in those, whereas in dice you cannot.

Agreed.

Importantly, and in laymen's terms:

This sequence:
<50, <50, <50, <50
is equally as likely as this sequence:
<50, >50, <50, >50
and this one:
<50, <50, >50, >50
and this one:
<50, >50, >50, >50
...and so on

Light

hero member

Activity: 742

Merit: 502

Circa 2010

In layman's terms with the use of math, basically the past has no bearing or influence on your future results (for completely unrelated events ie dice rolls). So don't believe that just because something is so unlikely that it cannot happen when you lose 28 times in a row and try for a 29th.

That's really the reason I stopped dice betting, the more you play the more you accentuate the house edge against you, whereas in something like sports betting/poker you have greater control over your results and hence there is a correlation between your skill and your earnings. In effect you can improve your probability of winning in those, whereas in dice you cannot.

monsterer

legendary

Activity: 1008

Merit: 1007

Quote from: spixel on March 06, 2014, 06:37:54 AM

Rolling <50 x amounts of times in a row still has its own probability, true a previous roll doesn't affect the next but this is obvious.

Rolling <50 1000 times in a row has the exact same probability of rolling any combination of < or > 1000 times in a row.

The question shouldn't be: "Would would rather bet $1,000 on rolling <50 once or <50 ten times in a row?", but instead:

"Would would rather bet $1,000 on rolling <50 once or any combination of under or over ten times in a row?"

<50 once has a probability = 1/2.
10 rolls (< or >) in a row has a probability = (1/2)^10 = 1/1024.

spixel

member

Activity: 98

Merit: 10

Rolling <50 x amounts of times in a row still has its own probability, true a previous roll doesn't affect the next but this is obvious.

Would would rather bet $1,000 on rolling <50 once or <50 ten times in a row?

monsterer

legendary

Activity: 1008

Merit: 1007

The gamblers fallacy is this:

Scenario: Dice game, 50/50 random chance, bet on <50 wins

"I've rolled >50 four times in a row now, probability of that happening is (1/2)*(1/2)*(1/2)*(1/2) = 1/16 (or 0.0625), the chances of rolling 5 times <50 are therefore 1/32, so I'm likely to win this time."

Seems logical enough? The reality is this:

The probability of <50 on any given roll is 1/2. It doesn't matter that you've rolled 4 times < 50, probability doesn't have a memory. To see why this is the case, consider this:

What's the probability of rolling 3 times < 50 and 1 time > 50? Guess what? (1/2)*(1/2)*(1/2)*(1/2) = 1/16.

What about 2 times > 50 and 2 times < 50, (1/2)*(1/2)*(1/2)*(1/2) = 1/16. The same probability.

It doesn't matter that you've rolled 4 times < 50, the probability of the next roll being > or < is still 50/50.

In fact, the probability of any combination of rolls occurring only depends on the number of rolls, not the previous outcomes.

Hope that helps some of you

Cheers, Paul.

Topic: The gamblers fallacy - page 2. (Read 1835 times)