matematica cooperazione

Why it’s better to collaborate: the experiment on the mathematics of cooperation

If a colleague speaks badly of us to the boss, what should we do? Instinctively one would feel nervous and, perhaps, reciprocate with the same weapon. But mathematics tells us that the opposite is better: in a famous one 1980 experiment based on the prisoner’s dilemma – a pillar of game theory, the branch of mathematics that seeks to rigorously explain human, animal and plant behavior – showed that under appropriate conditions the strategies that lead to the best results are cooperative ones rather than purely competitive ones.

The prisoner’s dilemma: is it better to cooperate or compete?

The prisoner’s dilemma is a famous “paradox” created in the 1950s, which compares two different behaviors in a difficult situation: collaboration and competition. Although it may seem counterintuitive, this dilemma was born in the economic field and developed with game theory, precisely to establish which strategies are most profitable in certain situations.

The situation is this: two prisoners they have to decide whether confess or less a crime without being able to consult before being interrogated. There are three possibilities presented to prisoners:

  • if both accuse their partner – that is, compete with each other – they receive a 2-year sentence;
  • if only one confesses and accuses the other, the other receives a 3-year sentence and the first is released;
  • if no one confesses – i.e. they collaborate with each other – both receive a sentence of less than 1 year.

Paradoxically, the prisoners will both tend to accuse the other to save themselves, thus obtaining a greater sentence (2 years) than they would have obtained by trying to protect the other (1 year).

Image

It has been observed, however, that the dynamic changes if the two prisoners – or players – know that they will have to see each other again in the futurebut above all if they don’t know when the last time they will see each other will be, which on balance is the most likely scenario compared to our daily experience.

What changes if we meet many times: the prisoner’s dilemma tournament

To try to establish what the best strategy was in cases similar to that of the two prisoners – that is, scenarios in which two individuals must understand whether to collaborate or compete to obtain the best possible result – in 1980 Robert Axelrod, full professor at the University of Michigan, held a tournament in which several research groups participated by sending their program with the solution.

The research groups participating in the tournament had one clear objective: score as many points as possible by playing the prisoner’s dilemma many times. Each university would have its own player play plan against all the others, and then add the results. Every match would last two hundred moves and for each move each university’s program would have to choose whether to collaborate or compete with the opposing university’s program.

Scores for each move were assigned as seen in the table: 3 points to both if they both decide to cooperate, 1 point to both if both decide to compete and, if so a decided to collaborate And the other Of competeto the first 0 points and to the second 5.

collaborate compete

With this information, each university presented its own program with a different strategy within it.

The 14 strategies applied during the tournament: peaceful and aggressive

They were introduced 14 strategies: 7 “peaceful”in which the player collaborates more than they compete, and 7 “aggressive”where the player competes more than they collaborate.

Among the aggressive strategies we find the program Friedman – The resentful – which follows the principle of “if you betray me once you are dead”. Initially he collaborates but, after the first betrayal, he punishes his opponent by competing forever.

Among the peaceful strategies there is instead Tit for Tatwhich starts by collaborating and then decides its move based on what the opponent did in the previous move. This strategy allows us to reciprocate cooperation, but also to counter aggression. There are more or less cooperative variants of this strategy. Jossfor example, 1 time out of 10 decides not to reciprocate cooperation and to be competitive.

Is it better to be peaceful or aggressive?

The collaborative strategy Tit for Tatdespite being extremely simple, managed to get to first place by scoring more than 500 points. Friedman he got thirty points less and Joss even 200 points less.

The incredible result of this tournament was that all the strategies in the top positions were those capable of reciprocating cooperation. The good ones, in short. In light of these results, Axelrod induced a second tournament. This time more than sixty different strategies arrived, but the However, he remained the winner Tit for Tat.

compete or collaborate victory

What are the characteristics of winning strategies

Even in this new tournament they noticed that the winning strategies had three characteristics in common:

  1. Good: Winning strategies do not become unjustifiably aggressive, at least until the final moves.
  2. Able to forgive: After punishing the opponent for being aggressive, they are willing to return to cooperation if the other player returns to cooperation.
  3. They don’t let their feet get pushed around: they are able to respond and survive very aggressive strategies, because they can reciprocate competitiveness with competitiveness.

It was also noted, however, that the world we live in defines the best strategy. In a world where there is only one Tit for Tat and everyone else is aggressive, Tit for Tat loses. But the crazy thing is that even just a small group of collaborative people is enough to make it happen Tit for Tat the best option. And if we assume that behavioral strategies are somewhat heritable, within a couple of generations the entire population becomes cooperative.

So how should I behave with my colleague? To quote Axelrod:

Be good, be forgiving, but don’t be a doormat.