Sep

24

The Prisoner's Dilemma is very well analyzed in the highly recommended very technical book, Evolutionary Dynamics by Martin Nowak.

The two by two payoff matrix:

                         Remain Silent       Confess

Remain Silent          -1                      -10

Confess                     0                       -7

shows payoffs to you if you and your colleague have committed a crime. The D. A says that if you confess and your colleague doesn't, you go free. But if you don't confess and he confesses you get 10 years in jail. But if you both confess you both get 7 years in jail. But if you both are silent, you both get just 1 year in jail because they can't prove anything.

The problem is that you do better by being disloyal to your partner. And so does he.

Rapaport has a very good solution to this problem if you play the game repeatedly. It has many applications to trading. If you are a flexion and you have inside information, perhaps from being one of the hundred people receiving economic releases in advance on a need to know basis, and your conspirator is a trend follower, or someone you are revealing the news to, as so often happens, you do better if you act but your colleague doesn't. Same for him. But if you both act, you'll move the market and the opportunity will be lost.

What other situations in markets can be modeled by the prisoners dilemma, and how do the solutions that Nowak and Wiki discuss illumine our trading, and enlighten us as to the disadvantages we face.

Tyler McClellan adds: 

Freeman Dyson published a short paper in the last year or so that supposedly showed a very unintuitive and until then unknown solution to this game.

"Iterated Prisoner's Dilemma Contains Strategies the Dominate and Evolutionary Opponent"

Stefan Jovanovich writes: 

The speculations about the Prisoner's dilemma too often omit the fact that the criminals both belong to the same tribe. The criminals' choices of rat/don't rat are bounded not only by the lesser/greater punishments by the prosecutors but also by the rewards/punishments offered by the gang/group. When the group's incentives are included in the calculations, the conspirators will, as wise guys, follow the logic of silence. That is why successful Federal prosecutions of organized crime that depend on informers have to offer the additional incentive of bribery. An offer of lesser punishment is not enough.

Pitt T. Maner III comments: 

And the game rules and risk/reward payouts in open systems would seem even more variable and subject to interpretation/enforcement depending on the players involved.

This article has two viewpoints on some recent data: "it's suggestive" vs. "overwhelming"

…"Does a burst of ETF trading in the same millisecond of the Federal Reserve's policy statement raise an eyebrow? Sure. Is it indicative of a leak or insider trading? Not necessarily. For that, you'd need something besides numbers on a chart."

And this is one of the latest papers on the subject which might be of interest:

"Penn Biologists Show that Generosity Leads to Evolutionary Success"

"Last year William Press and I proposed the 'extortion strategy' in the game of Prisoner's Dilemma, enabling one player to maintain a dominant position over the other," said Dyson, who is retired as a professor of physics at the Institute for Advanced Studies in Princeton, N.J. "One year later, Stewart and Plotkin turned our strategy upside down and showed that it enables one player to coax the other gently toward collaboration. They understood our strategy better than we did. They reached by rigorous mathematics the happy conclusion that, in a game between ruthless antagonists, generosity wins." '

Richard Owen writes:

It always seems to be that the merits of a mathematical discovery aren't enough by themselves. The closed system needs to be extrapolated to the wide world. Thus a specific proof about a mathematical game is assumed to show that "it pays to be generous in life." As if, without the mathematical imprimatur, this might be held in some doubt. That particular habit of taking results proved in a closed system and extrapolating them to the wide world is probably particularly relevant to the investment field.

Jim Sogi adds: 

I definitely like this author's approach to game theory using a spreadsheet to tally levels of factors similar to a plus minus decision list. The approach can definitely be used to quantify market information and decisions. It breaks down multi factored complex decisions into manageable quantifiable choices which are tallied to arrive at the big decision.


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search