1 A die is rolled 24 times. Use the CLT to estimate the probability that: a. the sum is greater than 84 b. the sum is equal to 84
The expected value is 3.5 * 24 = 84 The variance is \(24 * \frac{35}{12} =70\) The standard deviation is $=$8.3666003
Since the expected value is 84, the odds of it being above or below that value are 50%.
The probability is then the area under the curve for the area +/- 0.06 S.D. from the mean. Using r’s pnorm function for the bounds 83.5 and 84.5, the difference in cumulative probability is 4.8% - thus the probability of obtaining the expected value of 84 is equivalent to about 5%.
pnorm(84.5, mean = 84, sd = sqrt(70))- pnorm(83.5, mean = 84, sd = sqrt(70))
## [1] 0.04765436
As the number of rolls and expected value increases, the likelihood of obtaining the expected value will decrease. For 1,000 rolls, that probability is 0.7%
stdev = sqrt(1000*35/12)
pnorm(3500.5, mean = 3500, sd = stdev) - pnorm(3499.5, mean = 3500, sd = stdev)
## [1] 0.00738687
2.A random walker starts at 0 on the x-axis and at each time unit moves 1 step to the right or 1 step to the left with probability 1/2. Estimate the probability that, after 100 steps, the walker is more than 10 steps from the starting position.
~2.3%. 10 steps is 2 standard deviations from the mean, whose area under the curve of a normal distribution is ~2.3%
n = 100
p = 0.5
q = 1 - p
var = n*p*q
sd = sqrt(var)
1 - pnorm(10, mean = 0, sd = sd)
## [1] 0.02275013
The expected loss from one round is $0.25. From 240 rounds, it is $60.
What is the probability that she lost no money? Zero. With an expected value of $60, and a standard deviation of sqrt(2400.250.75) = 6.7, a corresponding z score of 60/6.7 = ~8.96, there is approximately zero chance she will have any net winnings.
ExpectedValue = 0.25*2 + 0.75*-1
ExpectedLosses = ExpectedValue*240
n = 240
p = 0.25
q = 1- p
stdev = sqrt(n*p*q)
pnorm(0, mean = -60, sd = stdev)
## [1] 1
plot(density(rnorm(240, mean = -60, sd = sqrt(240*0.25*0.75))),
main = "Expected Winnings")