Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

We toss 10 identical biased coins. Let $I$ and $J$ be the event that the $i$-th and $j$-th toss ($i,j\in\mathbb{N}_{[2,10]}$) are different from their predecessors, respectively. What's the easiest way to see that $I$ and $J$ are independent whenever $|i-j|>1$?

share|improve this question
    
Well, one way would be to verify that $P(I \cap J) = P(I)P(J)$ by direct computation. –  Alan Guo Oct 23 '12 at 4:00
    
@AlanGuo: You are right. But I'm expecting an explanation on how to understand it more intuitively or why exactly this rule fails when $|i-j|=1$ –  Voldemort Oct 23 '12 at 4:03
    
Is $i \in \{0, 1\}$? Can you be more specific with your notation? –  Alex Oct 23 '12 at 4:09
    
@AlexOlssen: $i,j\in\{2,3,...,10\}$ –  Voldemort Oct 23 '12 at 4:14
    
@AlexOlssen: $I$ and $J$ are events that those tosses are "different from their predecessors," instead of whatever their outcomes. –  Voldemort Oct 23 '12 at 4:20

1 Answer 1

up vote 2 down vote accepted

When $|i - j| = 1$ then one of the tosses is the predecessor of the second toss. Since the coins are biased ($p \in (0,1)$, and $p \neq \frac{1}{2})$, knowing that a coin agrees with its predecessor changes our probability of it being heads or tails.

That is $P(Heads |\text{ agrees with predecessor}) = \frac{P(Heads \text{ and agrees with predecessor})}{P(\text{agrees with predecessor})} = \frac{p^2}{p^2 + (1-p)^2} = \frac{p^2}{1 - 2p(1-p)} \neq p$ if $p \neq 0,\frac{1}{2},1$.

So, intuitively, if $p > \frac{1}{2}$, then if $I$ is true, then $i$ is most likely heads, and $i-1$ is most likely heads, and that means that $i-1$ is also more likely to agree with its predecessor (since $i-2$ is more likely heads), etc.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.