Game Theory and Evolution of cooperation Gilberto Cmara

Game Theory and Evolution of cooperation Gilberto Câmara, Earth System Science Center, INPE Licence: Creative Commons By Attribution Non Commercial Share Alike http: //creativecommons. org/licenses/by-nc-sa/2. 5/

Acknowledgments for using previous material q q q Martin Nowak (Harvard University, USA) Francisco C. Santos (Université Libre de Bruxelles, Belgium) Craig Callender (Philosophy, Univ California San Diego, USA) Ana Aguiar (INPE, Brazil) Tiago Carneiro (Federal University of Ouro Preto, Brazil) Guy Brasseur (NCAR, USA)

Game Theory GT is an analytical tool for social sciences that is used to model strategic interactions or conflict situations. Strategic interaction: When actions of a player influence payoffs to other players

Game Theory Explanation: What is the game to be played? Prediction: What outcome will prevail? Advice or prescription: Which strategies are likely to yield good results in which situations?

Where can we use Game Theory? Any situation that requires us to anticipate our rival’s response to our action is a potential context for GT. Economics, Political science, Biology

What is a Normal Form Game? Players: list of players Strategies: all actions available to all players Payoffs: a payoff assigned to every contingency (every possible strategy profile as the outcome of the game) John Kennedy and Nikita Khrushchev

Modeling two-party games Payoffs for each player depend on actions of both Two possible strategies: A party cooperates when he performs value-increasing promises, and defects when he breaches

Modeling choice in non-cooperative games Player 2 Cooperate Player 1 Defect Player 1 Both cooperates, Player 2 defects Player 1 defects, Player 2 cooperates Both defect

Silvio Santos e o jogo do “Sete e Meio” Dois jogadores se enfrentam na TV. Se dois jogarem “meio”, cada um ganha R$ 14 mil. Se um jogar “sete” e o outro “meio”, o primeiro ganha R$ 112 mil e outro não ganha nada Se os dois jogarem “sete”, não ganham nada.

Prisoners’ Dilemma Two suspects are caught and put in different rooms (no communication). They are offered the following deal: 1. If both of you confess, you will both get 3 years in prison 2. If you confesses whereas the other does not, you will get 1 year and the other gets 5 years in prison. 3. If neither of you confess, you both will get 2 years in prison.

The “chicken game” “Rebel without a cause” Two persons drive their cars towards a cliff. They must stop or both may die in the fall. The one that stops first will be called a "chicken, " meaning a coward.

The hawk-dove game (== chicken game) Two individuals compete for a resource (In biological terms, its value increases in the Darwinian fitness of the individual who obtains the resource) Hawk Initiate aggressive behaviour, not stopping until injured or until one's opponent backs down. Dove Retreat immediately if one's opponent initiates aggressive behaviour. Maynard Smith and Price, "The logic of animal conflict“ (Nature, 1973 )

The hawk-dove game (== chicken game) Encyclopedia Britannica

The stag-hunt game: conflict between safety and social cooperation Two hunters want to kill a stag. Success is uncertain and, if it comes, require the efforts of both. On the other hand, either hunter can forsake his partner and catch a hare with a good chance of success.

The stag-hunt game: conflict between safety and social cooperation C D C 10, 10 0, 6 D 6, 0 5, 5 Rousseau, in A Discourse on Inequality: “If it was a matter of hunting a deer, everyone well realized that he must remain faithful to his post; but if a hare happened to pass within reach of one of them, we cannot doubt that he would have gone off in pursuit of it without scruple. . . "

Generalizing. . . Cooperation requires at least two individuals: A: the one providing cooperation (DONOR) B: the one benefiting from cooperation (RECEIVER) Donor has a cost c to cooperate and confers a benefit b to other player C D C b–c -c D b 0 you Payoff matrix other

Terminology Player 2 T = Temptation to defect R = Reward for mutual cooperation P = Punishment for mutual defection S = Sucker's payoff

Generalizing. . . Payoff matrix R: mutual cooperation other P: mutual defection C D C R(1) S(-c) D T(b) P(0) S : sucker’s payoff T : temptation to defect you Taking R = 1 and P = 0

Generalizing. . . Payoff matrix R: mutual cooperation opponent P: mutual defection C D C 1 S D T 0 S : sucker’s payoff T : temptation to defect you Taking R = 1 and P = 0

Different ordering -> Different tensions greed fear C D C R S D T P Chicken game T > R > S > P Stag-hunt game R > T > P > S Prisoner’s dilemma T > R > P > S (Macy&Flache, PNAS 2002)

Different ordering -> Different tensions greed fear C D C R S D T P Chicken game T > 1 > S > 0 Stag-hunt game 1 > T > 0 > S Prisoner’s dilemma T > 1 > 0 > S (Macy&Flache, PNAS 2002)

Is the tragedy of the commons inevitable? Experiments show that cooperation emerges if virtuous interactions exist source: Novak, May and Sigmund (Scientific American, 1995)

Repeated prisioner´s dillema Four different strategies for repeated prisioner´s dillema source: Novak, May and Sigmund (Scientific American, 1995)

Repeated prisioner´s dillema Evolution of prisioner´s dillema comparing different strategies source: Novak, May and Sigmund (Scientific American,

Spatial Prisioner´s Dillema q q Nowak and May considered a large lattice with each cell occupied by one player. The players engage in one round of the Prisoner’s Dilemma game against each of their neighbors. Afterward, the next generation is formed: each cell is taken over by a copy of the highest-scoring strategy within the neighborhood.

Tragedy of the Commons (Hardin, 1968) Assume a common-property resource (exclusion is difficult and joint use involves subtractability) with no property rights. (Pasture open to all) Each herdsman tries to keep as many sheep as possible on the commons. Each tries to maximize gain.

Add those sheep! The rational herdsman concludes that he should add another sheep. And another…And so does each herdsman “Ruin is the destination toward which all men rush, each pursuing his own best interest…”

Tragedy of the Commons? Everybody’s property is nobody’s property (Hardin)

Prisioner´s Dillema as a Model for the Tragedy of the Commons 1. Suppose the commons can support 2 sheep at no cost and that each additional sheep put in the commons has a cost of 1/3 of its price due to overgrazing. 2. Assume two herdsman with one sheep on the commons each. 3. If a herdsman puts another sheep in the commons, he receives all the proceeds from the sale of each additional animal. His temptation is 4/3 and the sucker´s payoff for the other herdsman is -1/3.

Prisioner´s Dillema as a Model for the Tragedy of the Commons You are the herdsman. What are your options? Do you cooperate or defect? C D C 1 -1/3 D 4/3 1/3 you Payoff matrix other

Preconditions for the tragedy of the commons Lack of restraint on pursuits of self-interest Consequences are externalities (I don’t have to pay)

Externalities in the global commons Activity of one person has an impact on the well-being of another. Positive externalities (or external benefits): Benefits realized by those who didn’t pay for them. Negative externalities (or external costs): Costs borne by those who didn’t generate them. Byproducts that harm others. SUVs in USA Climate Change in Africa

Is the tragedy of the commons inevitable? Experiments show that cooperation emerges if virtuous interactions exist source: Novak, May and Sigmund (Scientific American, 1995)

CO 2 emissions (Pg. C y-1) Collective spatial action: global change 10 8. 7 8 Fossil fuel 6 9. 9 Pg. C 4 Land use change 2 1960 1970 1980 1990 1. 2 2000 2010 12% of total Le Quéré et al. 2009, Nature-geoscience, 2009

We need spatially explicit models to understand human-environment interactions Nature: Physical equations Describe processes Society: Decisions on how to Use Earth´s resources

Common pool resources (Elinor Ostrom)

The ultimate common pool resource

Governing the commons [Ostrom, Science, 2005]

Governing the commons: Ostrom´s conditions 1. Clearly defined boundaries. 2. Congruence between appropriation and provision rules and local conditions. 3. Collective-choice arrangements. 4. Monitoring and graduated sanctions. 5. Conflict-resolution mechanisms. 6. Minimal recognition of rights to organize. 7. Organized governance activities.

Ostrom on governing the commons “The challenge is how best to limit the use of natural resources so as to ensure their long-term economic viability. ” “Neither the state nor the market is uniformly successful in enabling individuals to sustain long-term, productive use of natural resource systems. ” “Optimal equilibrium with centralized control is based on assumptions concerning accuracy of information, monitoring capabilities, sanctioning reliability, and zero costs of administration. ”

Modelling collective spatial actions: the complex systems approach photo: Chico Albuquerque 1. 2. 3. 4. Situated individuals (persons, groups, agents) Interaction rules - communication Decision rules - actions Properties of space

How can cooperation happen? Nowak MA (2006). “Five rules for the evolution of cooperation” Science 314: 1560 -1563 (most highly cited multidisciplinary paper – ISI, 1 st quarter 2010) "I would lay down my life for two brothers or eight cousins“ (J. B. S. Haldane)

Five rules for evolution of cooperation b = benefit for the recepient c= cost for the donor

Conclusion Spatial models of human-environment interactions can support the study of common pool resources

We need social theories to understand humanenvironment interactions q Survey Moran, “Environmental Social Science: Human-Environment Interactions and Sustainability” (2010) q Social simulation Schelling, “Micromotives and macrobehavior” (1978). Batty, “Cities and complexity” (2005). q Game theory von Neumann and Morgenstern, “Theory of games and economic behavior” (1944) Nash, "Equilibrium points in n-person games“ (1950). q Evolutionary dynamics Maynard Smith, ”Evolution and theory of games” (1982) Axelrod, “Evolution of cooperation” (1988). Novak, “Evolutionary dynamics: exploring the equations of life” (2005). q Institutional analysis Ostrom, “Governing the commons” (1990).