Re: Parrondo's paradoxical games

From: Hal Finney (hal@finney.org)
Date: Thu Jun 12 2003 - 15:39:12 MDT

  • Next message: Anders Sandberg: "Making sex tastier"

    Michael Butler writes:

    > On Thu, 12 Jun 2003 11:10:36 -0700, Hal Finney <hal@finney.org> wrote:
    > > The problem is that for his paradox to work, the games have to be set
    > > up so that the payoff depends on your cash balance. No real games work
    > > that way, so the paradox is artificial. In my posting above I came up
    > > with a trivial example which demonstrates that.
    >
    > Haven't reviewed your posting, but the Parrondo effect seemed to me to
    > resemble a kind of accumulations-of-beneficial-roundings process. Does that
    > square with your interpretation, Hal?

    I compared it with the investment strategy of asset rebalancing.
    Many investment advisors suggest that you periodically review your asset
    allocation and rebalance it in accordance with your investment strategy.
    For example, if you were 50% in bonds and 50% in stocks, then after a
    while price changes will alter these percentages. So you would transfer
    assets to bring your allocation back to what you wanted.

    You can show that using asset allocation, with a variety of assets that
    are going up and down somewhat independently, that you can do much better
    than the average gain of each individual asset. In fact you can make
    money doing this even if all the assets are on a long term downward trend
    (as long as they fluctuate up and down enough). That's a close analogy
    to Parrondo's trick of making money even on losing games.

    Hal



    This archive was generated by hypermail 2.1.5 : Thu Jun 12 2003 - 15:49:52 MDT