Great article as always! The poker table being a microcosm of wider society is particularly interesting - I wonder if the converse is true (i.e. whether in times of “prosperity” or more diplomatic global behaviour, there would be less antics and more positive-sum thinking/play at the table?)
I think you've misstated the payoffs in the prinoners' dilemma. The classic situation is that, regardless of whether the other prisoner confesses, each prisoner is better off confessing.
I've usually seen this described as: the police believe this pair have committed dozens of burglaries, but only have enough direct evidence to convict them of breaking and entering the building where they were caught. If both remain silent, they'll stay in jail until the trial, and then each be sentenced to time served plus probation. But if just one rats out the other for the string of burglaries, the rat goes free immediately, while the other gets the book thrown at him, effectively life in prison. If both confess, they each get a plea-bargained sentence of five years in prison.
Thus, if my partner confessed, I can go from life in prison to five years in prison by also confessing. If my partner stayed silent, I can go from months in jail to home free by confessing.
The payoff matrix you describe gives incentive within the game to remain silent if I believe my partner is likely to do so. That doesn't land the full pressure that the traditional prisoners' dilemma payoff matrix presents.
You're absolutely right that there is a version of the game that is exactly the way you describe it. I chose a simpler payoff matrix to illustrate, but the core idea is the same in both matrices: you'd be better off collectively by cooperating, but the individual dominant strategy is to defect, so you're stuck in the double defection quadrant that neither side actually wants - and you have no way of getting into the "good" (cooperate/cooperate) quadrant in the absence of some sort of coordinating mechanism. The negative payoff of cooperation/defection is simply too high of a risk.
I watched Kabrhel play this past week and I couldn't figure out why it bothered me so much (apart that he is so annoying...) I've always tended to believe in win-win resolutions, and I was wondering how I would handle him at the table -- it probably wouldn't be good. If I had to switch to pure zero-sum behavior I would probably make mistakes.
There is some history in other sports -- in tennis in the 70's McEnroe and Nastase were disruptive players, and it helped them win. But eventually tennis penalized such behavior, and you don't see it anymore.
Apart from that, I've been a long time fan of Survivor on tv and wondered if there was game theory that could applied there. Interesting that Tit-for-two-Tats is the best strategy -- trust people in your alliance unless they deceive you twice. I'll keep that in mind if they ever select me to be on the show (believe me, I've applied many times...)
Thanks again for writing about such interesting subjects!
As to Kabrhel, I would would recommend Hawk Dove Game Theory strategy
As to why people adopt certain strategies like zero sum v collaborative, Hidden Games by Hoffman and Yoel has some very good insights.
As to historic of this dispute in game theory of collaborative v zero sum, see Bhattacharya’s book on von Neumann (but I’m suspect you may already have). The game theory section talks about von Neumann’s disapproval of Nash’s zero sum view of the world and how von Neumann’s world view was influenced by pre WWII Hungarian intellectual culture that emphasized collaboration.
I find it interesting that Nash (with others) invented a game called “So Long Sucker” that is also known as F You Buddy. It can only be won by first entering into coalitions and then breaking the prior agreements.
As Bhattacharya points out, the public view of Nash is largely shaped by the movie Beautiful Mind. Yet, the movie smooths out Nash’s negative qualities of his personality.
Great article as always! The poker table being a microcosm of wider society is particularly interesting - I wonder if the converse is true (i.e. whether in times of “prosperity” or more diplomatic global behaviour, there would be less antics and more positive-sum thinking/play at the table?)
That would be nice! I haven't played long enough to know, but it would make sense. Microcosms do tend to mirror greater social trends.
I think you've misstated the payoffs in the prinoners' dilemma. The classic situation is that, regardless of whether the other prisoner confesses, each prisoner is better off confessing.
I've usually seen this described as: the police believe this pair have committed dozens of burglaries, but only have enough direct evidence to convict them of breaking and entering the building where they were caught. If both remain silent, they'll stay in jail until the trial, and then each be sentenced to time served plus probation. But if just one rats out the other for the string of burglaries, the rat goes free immediately, while the other gets the book thrown at him, effectively life in prison. If both confess, they each get a plea-bargained sentence of five years in prison.
Thus, if my partner confessed, I can go from life in prison to five years in prison by also confessing. If my partner stayed silent, I can go from months in jail to home free by confessing.
The payoff matrix you describe gives incentive within the game to remain silent if I believe my partner is likely to do so. That doesn't land the full pressure that the traditional prisoners' dilemma payoff matrix presents.
You're absolutely right that there is a version of the game that is exactly the way you describe it. I chose a simpler payoff matrix to illustrate, but the core idea is the same in both matrices: you'd be better off collectively by cooperating, but the individual dominant strategy is to defect, so you're stuck in the double defection quadrant that neither side actually wants - and you have no way of getting into the "good" (cooperate/cooperate) quadrant in the absence of some sort of coordinating mechanism. The negative payoff of cooperation/defection is simply too high of a risk.
Great article!
I watched Kabrhel play this past week and I couldn't figure out why it bothered me so much (apart that he is so annoying...) I've always tended to believe in win-win resolutions, and I was wondering how I would handle him at the table -- it probably wouldn't be good. If I had to switch to pure zero-sum behavior I would probably make mistakes.
There is some history in other sports -- in tennis in the 70's McEnroe and Nastase were disruptive players, and it helped them win. But eventually tennis penalized such behavior, and you don't see it anymore.
Apart from that, I've been a long time fan of Survivor on tv and wondered if there was game theory that could applied there. Interesting that Tit-for-two-Tats is the best strategy -- trust people in your alliance unless they deceive you twice. I'll keep that in mind if they ever select me to be on the show (believe me, I've applied many times...)
Thanks again for writing about such interesting subjects!
Thank you! And yes, great point about tennis. When norms (and rules) change, behavior follows. And I think the sport improves as a result!
Thanks for great article as always.
As to Kabrhel, I would would recommend Hawk Dove Game Theory strategy
As to why people adopt certain strategies like zero sum v collaborative, Hidden Games by Hoffman and Yoel has some very good insights.
As to historic of this dispute in game theory of collaborative v zero sum, see Bhattacharya’s book on von Neumann (but I’m suspect you may already have). The game theory section talks about von Neumann’s disapproval of Nash’s zero sum view of the world and how von Neumann’s world view was influenced by pre WWII Hungarian intellectual culture that emphasized collaboration.
I find it interesting that Nash (with others) invented a game called “So Long Sucker” that is also known as F You Buddy. It can only be won by first entering into coalitions and then breaking the prior agreements.
As Bhattacharya points out, the public view of Nash is largely shaped by the movie Beautiful Mind. Yet, the movie smooths out Nash’s negative qualities of his personality.
Thanks for the thoughtful comment and recommendations! Great list :)