
Advanced Statistics Advanced Probability and Statistics Math Forum 
 LinkBack  Thread Tools  Display Modes 
February 8th, 2015, 06:32 AM  #1 
Newbie Joined: Feb 2015 From: West Calder Posts: 1 Thanks: 0  Tricky Bayesian question using posterior predictive distributions, help!!!?
A game is played using a biased coin, with unknown p. Person A and Person B flip this coin until they get a head. The person who tosses a head first wins. If there is a tie, where both people took an equal number of tosses to flip a head, then a fair coin is flipped once to determine the winner. Person A plays until they get a head, then Person B plays afterwards. Using the Jeffrey's Prior for the Geometric distribution and the Posterior predictive distribution for Person B's score given we know Person A's score, what is the probability that Person A will win? My working so far: X=the number of tosses until person A gets a head Z=the number of tosses until person B gets a head The Jeffrey's prior for the Geometric distribution is: p(p) = p^1*(1p)^0.5 The posterior predictive distribution for Person B's score given Person A's score is: p(zx) = 2(2x1)/(2x+2z1)(2x+2z3) Finally: Prob(Person A winsx) = Prob(Z>x) + Prob(Z=x)*0.5 Now P(Z=x) is computed by substituting z for x in the formula p(zx). But how do I compute Prob(Z>x)? By the way, the final answer is: Prob(Person A winsx) = 2(2x1)^2/(2x+2z1)(2x+2z3) Can anyone please tell me how to do this, and if I'm doing it correctly? 

Tags 
bayesian, distributions, posterior, predictive, question, tricky 
Search tags for this page 
Click on a term to search for related topics.

Thread Tools  
Display Modes  

Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Prior And Posterior  wannabe1  Advanced Statistics  4  February 27th, 2013 05:36 PM 
A question about binomial distributions...  Artus  Advanced Statistics  1  October 17th, 2012 06:30 AM 
Posterior of the gamma and its conjugate  malaguena  Advanced Statistics  0  February 14th, 2011 12:11 AM 
Question: something with two uniform distributions  instantaneous  Advanced Statistics  0  October 25th, 2009 04:29 PM 
Bayesian or not Bayesian; this is my problem!  Marione  Advanced Statistics  0  April 24th, 2008 02:51 PM 