+ Reply to Thread
Results 1 to 6 of 6

Thread: Number of trials in zero-sum game

  1. #1
    Points: 11, Level: 1
    Level completed: 21%, Points required for next Level: 39

    Posts
    3
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Number of trials in zero-sum game




    Hi all, this is my first post here! Forgive me if it's not in the right place.

    I have a question that's been bugging me because I tend to be pretty good at math/statistics but I guess I'm a bit rusty.

    Let's say you have $100 and the price to participate in the following game is $20:

    You have a 5% chance to win $80
    You have a 10% chance to win $60
    You have a 15% chance to win $40
    You have a 20% chance to win $20
    You have a 50% chance to win $0

    The expected value of your winnings will be 0.05*80+0.1*60+0.15*40+0.2*20 = $20 and therefore, given that the price to play is $20 as well, it is a zero-sum game. Suppose you are nevertheless interested in playing this game until you are bankrupt (lose the entire $100). My question is: what is the chance that you will be bankrupt in 10 trials or less? How do you calculate this? I suppose it's more the method of calculation I'm after, than the exact probabilities. 10 trials is just an example.

    Thanks in advance!

  2. #2
    TS Contributor
    Points: 22,410, Level: 93
    Level completed: 6%, Points required for next Level: 940

    Posts
    3,020
    Thanks
    12
    Thanked 565 Times in 537 Posts

    Re: Number of trials in zero-sum game

    The problem can be well described by the Markov chain,

    with state space \{0, 1, 2, \ldots \}

    (state x represent having 20x of current wealth)

    Specifically state 0 is the absorbing state represent bankruptcy.

    The transition probabilities are

    p_{i,i+3} = 0.05, p_{i,i+2} = 0.1, p_{i,i+1} = 0.15, p_{i,i} = 0.2, p_{i,i-1} = 0.5, i = 1, 2, 3, \ldots
    and 0 elsewhere among the transient states.

    The distribution of the wealth after n trials can be calculated by the nth power of the transition matrix.

  3. #3
    Points: 11, Level: 1
    Level completed: 21%, Points required for next Level: 39

    Posts
    3
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Number of trials in zero-sum game

    You, sir, are a hero.

    Am I right in assuming that the size of the state space is theoretically unlimited in this example? If so, how does that affect my calculations?

  4. #4
    TS Contributor
    Points: 22,410, Level: 93
    Level completed: 6%, Points required for next Level: 940

    Posts
    3,020
    Thanks
    12
    Thanked 565 Times in 537 Posts

    Re: Number of trials in zero-sum game

    Yes the number of states is countably infinite (so as the dimension of the transition matrix). Being rigorous, you may need to know some technical regularities / defining a matrix in that way. Anyway, if you are just interested the game in finite time settings, you can always truncate the transition matrix, up to the maximum possible wealth within the given number of trials.

  5. #5
    Points: 11, Level: 1
    Level completed: 21%, Points required for next Level: 39

    Posts
    3
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Number of trials in zero-sum game

    Got it thanks to you and Matlab. Cheers mate.

  6. #6
    Points: 46, Level: 1
    Level completed: 92%, Points required for next Level: 4
    gjay822's Avatar
    Location
    UK
    Posts
    2
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Number of trials in zero-sum game


    LOL, I didn't got the right answer in my own calculations. It is far as yours. I am still in need of studying math.

+ Reply to Thread

           




Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats