I have a rather complex game, whose expected value I need to find. Given is a binary random value with probability for success p (lets say an unfair coin with probability of landing heads p and tails 1-p, p<<0,5) The rules of the games are the following: you start by betting 1 euro, a multiplier = 1 and 15 tosses. Each time you land heads (less likely option) you receive 15 more tosses (added to the remaining one you have) and your multiplier increases by 1. Landing tails doesn't give you anything, you've just lost one toss. Your multiplier can grow up to 5, and each time you land a head after achieving a multiplier of 5 adds new 15 tosses, but does not increase the multiplier further. I am seeking for the expected multiplier of the game.
I started by building a Markov Chain table A[i,j] for the probability of spin j to have multiplier i: A[1,1] = 1, A[1,2] = 1-p, A[2,2] = p, A[1,3] = (1-p)*(1-p), etc... We have 15 initial spins then each A[i, 15+i] = 0. But I am still stuck of deriving the Avg MP from the table. Any help will be much appreciated!
Thanks!
EDIT: I forgot to say, at each toss you win your bet x Multiplier, so actually, I am looking for the EV of the game, but I think it can be presented as finding the average MP of the game and the average number of tosses, so in the end the EV should be their product. I know how to find the Avg number of tosses, and my main issue is the AVG MP.