Two MSU researchers quashed a spin on evolutionary game theory from 2012 that held coercion as a more favorable action than cooperation.
“In an evolutionary setting, these zero determinant strategies (those using coercion) will go extinct,” said Christoph Adami, professor of microbiology and molecular genetics. “You might succeed for a short time, but in the long run it can’t sustain.”
Game theory is used in many academic disciplines and focuses on two or more parties with the outcome of their strategic decision making.
For example, the prisoner’s dilemma portrays a game theory situation in which two prisoners are arrested for a crime. The police lack evidence to convict the prisoners to more than a year in prison, unless one of the prisoners rats on his partner.
They are questioned by the police in separate rooms and have several options: if one blames the other, he goes free and the other person serves three years in prison; if both prisoners blame each other, they both receive two years in prison; or if neither blames the other, they both receive a year in prison.
The spin on this assumes the prisoners can communicate, and the 2012 theory contests that one of the prisoners, the zero determinant, will coerce his partner into taking the fall.
Adami and his partner, microbiology and molecular genetics research associate Arend Hintze, tested the 2012 theory after it stirred the game theory community and countered their own research from 2010 that favored cooperation, Adami said.
“We plugged zero determinant strategy into the model and they disappeared after some time,” Hintze said. “We knew evolution should come up with the best solution. We have cooperation all over the place.”
Hintze and Adami found that when communication is introduced into game theory, after several attempts, cooperation holds the best outcome.
“They key to cooperation is communication,” Adami said. “It can also be used to dominate and extort. That’s not a behavior stable in evolution. It gives rise to extinction.”
Adami said zero determinants become extinct because they eventually play against each other and receive the most negative outcome. The only way zero determinants could survive is if they recognized one another, but in that case, so too could the cooperators evolve to recognize them. Adami called this a classic biological arms race.
15-year-old East Lansing resident Danielle Hanson and 15-year-old East Lansing resident Colin Graham played the prisoner’s dilemma game. Hanson tried to use coercion and Graham promised to take the fall, but he turned on her. Both got two years.
“It makes me feel angry,” Hanson said. “I can’t trust him anymore.”
Hintze said those who favor cooperation retaliate, but only when cheated.
“You can’t be entirely trusting, but (must be) able to develop trusting relationships,” he said.