Take the 2-minute tour ×
Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization. It's 100% free, no registration required.

There is a little problem in the working on this one I have encountered recently. Here it goes: Starting off with \$100, when you toss a coin, if there is a tail, you will get \$10 and lose $10 if it is a head. What is the probability that after 1000 tosses, the amount of money you have is between 90 and 110?

Here is the way I do it. Let $X$ be the difference in amount when you toss a coin. Hence, X could be -10 and 10 with the probability of 1/2 each. The expected value is 0 and the standard deviation is 10. Hence, I go on formulating the sum $S = X_1 + X_2 + \ldots + X_{1000}$ and figure out the $\mu(S) = 0$ and $var(S) = 100000$ and I do not know how to go from here? Any hints will be greatly appreciated.

share|improve this question
    
Are you allowed to keep tossing after you hit \$0? And if yes, and you get heads, will you go into debt or stay at \$0? This will significantly affect the answer. –  Ilmari Karonen Apr 14 at 1:18
    
That is a very insightful question, IImari. For this moment, we assume that the game is simple enough (you can be in debt) unless stop playing at \$0. It is another question to consider here. –  Viet Hoang Quoc Apr 14 at 1:23

2 Answers 2

The exact probability can easily be computed with the binomial distribution.

Let $X$ be a binomial random variable corresponding to the number of tails obtained in the 1000 tosses. You could do the following:
a) Determine the number of tails out of 1000 tosses that would result in the final amount of money being between 90 and 110. Denote $a$ the minimum number of tails, and $b$ the maximum number of tails.
b) The requested probability is $\Pr(a \leq X \leq b)$, which can be computed using the probability mass function of the binomial distribution.

Hint: can the final amount be 90? And 110?

share|improve this answer
1  
+1 especially for the hint. –  Dilip Sarwate Apr 13 at 14:48
    
That is an excellent approach. Cheers. –  Viet Hoang Quoc Apr 14 at 0:55

You should use Central Limit Theorem and deduce your answer from normal approximation: http://www.math.uah.edu/stat/sample/CLT.html

Basically you sum by CLT is just a normal distribution with the mean and variance you have already calculated.

share|improve this answer
1  
+1 The Binomial distribution gives $0.0252250$ while this Normal approximation gives $0.0252271$: they agree to over four significant figures. –  whuber Apr 13 at 15:36

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.