unknown distribution question

98 Views Asked by At

The question as follows: There's a gambling game with 100 rounds. First, a random number is randomized out of unknown continuous distribution. then, each round a random variable is randomized from same distribution. There are 2 gamblers: gambler A gambles the next random number would be bigger than the previous.

Question: what is the Expectation of number of rounds when gambler A was right?

The answer: $suppose\ X_{i}\ is\ the\ value\ of\ the\ number\ randomized\ in\ round\ i$ $and\ Y_{i}\ is\ an\ indicator\ that\ X_{i}>X_{i-1}:$

The Equation is as Follows:

$E(Y)=E(\sum_{i=1}^{100}Y_{i})=\sum_{i=1}^{100}E(Y_{i})=\sum_{i=1}^{100}P(X_{i}>X_{i-1})=\sum_{i=1}^{100}\frac{1}{2}=50$

and it's said that $P(X_{i}>X_{i-1})=\frac{1}{2}$ due to symmetry. I can't get it.

we are expected to show in our answers the whole way we get to the solution, with explanation for every equation.

I tried to convert P in terms of cumulative $F_{X}(x)$ but I'm confused which is my integration parameter and so on.

Thanks to everybody who tries to help!

1

There are 1 best solutions below

0
On

The simple answer is due to symmetry and independence of the random variables.
The Formulaic derivations:

$P(X_{i}>X_{i-1})\underset{(1)}{=}\int_{-\infty}^{\infty}P(X_{i}>x|X_{i-1}=x)\cdot f_{X_{i-1}}(x)dx\underset{(2)}{=}\int_{-\infty}^{\infty}(1-F_{X_{i}}(x))\cdot f_{X_{i-1}}(x)dx\underset{(3)}{=}\int_{-\infty}^{\infty}(1-F_{X_{i-1}}(x))\cdot f_{X_{i}}(x)dx\underset{(4)}{=}P(X_{i-1}>X_{i})$

1.Law of total probability

2.independence+ tail of comulative function

3.same dist.

4.obvious