How Behaviors Spread or Fade in Groups

The idea

In classic game theory, players think through their choices (Choices That Depend on Other People). Evolutionary game theory drops that assumption. It imagines a population where behaviors that pay off get copied, taught or inherited more often, and behaviors that do poorly fade. What pays off depends on what everyone else is doing, so the mix keeps shifting. When a common behavior holds its ground against a few newcomers who try something different, it is called evolutionarily stable.

Why it attracts me

This is how culture often works. Nobody votes on what becomes normal in a team or a family. It grows out of what gets rewarded, copied and tolerated. Each person adapts to the others, and the whole system changes as they do (Systems Whose Parts Learn and Adapt).

An example

Imagine a factory where reporting a defect brings blame on whoever reported it. At first, a few people still report. Over months, the people who stay quiet have easier days, newcomers copy them, and reporting fades. No one decided to hide problems. The reward system selected that behavior. Change the reward, for example by thanking whoever finds a defect first, and the same process can run in the other direction.

Where it connects

This is one route by which a group drifts into a pattern where no one gains by changing alone (John Nash: When No One Gains by Switching). It is also why changing the rules can work better than urging people to try harder (Cooperation by Changing the Game).

Questions I'm still exploring

• Which habits in my family or team are spreading because they are rewarded, not because they are good?
• How small a change in the rules is enough to tip what becomes normal?
• Can a good behavior survive in a group that does not reward it?

Sources and further reading

• Stanford Encyclopedia of Philosophy, "Evolutionary Game Theory"
• John Maynard Smith, Evolution and the Theory of Games (Cambridge University Press, 1982)