Conditional Dissociation. Cyp+By

Segismundo S. Izquierdo, Luis R. Izquierdo & Fernando Vega-Redondo


To use Cyp+By_applet, you will have to install NetLogo 5.3.1 (free and open source) and download the model itself. Unzip the downloaded file and click on Cyp+By_applet.nlogo


Cyp+By is a model designed to analyse the effect of conditional dissociation in the evolutionary emergence of cooperation. This section gives an informal and brief overview of Cyp+By. We use bold red italicised arial font to denote parameters (i.e. variables that can be set by the user).

In Cyp+By, there is a population of num-players players that are randomly matched in couples (partnerships) to play a 2x2 symmetric game. The 2 possible actions players can take are denoted C (for Cooperate) and D (for Defect). After each interaction, every individual is given the option to stay in her current partnership, or break it (thus getting a new randomly chosen mate for the next stage).

At each time-step, every individual has a fixed probability of dying; the expected number of time-steps each individual lives is exp-life. Whenever an individual dies, a new entrant comes in. In principle, new entrants adopt the strategy of an existing individual with probability proportional to her latest payoff, but mutations are also possible. Mutation occurs for each new entrant with probability prob-mutation, and when this occurs, the new entrant picks up a decision rule at random.


This section explains the formal model that Cyp+By implements. The information provided here should suffice to re-implement the same formal model in any sophisticated enough modelling platform. The figure below provides an informal sketch that illustrates the sequence of events within each time-step.


Sketch of the sequence of events within each time-step.
The "Remaining Pairs" and "Singles" at the end of a time-step are identical to the "Existing Pairs" and "Singles" in the next time-step.

The game

Consider a population consisting of an even number num-players of individuals who are matched in pairs. In each time-step, the two members of every pair play a symmetric 2x2 game once, where each of them can undertake one of two possible actions. These two possible actions are called C (for Cooperate) and D (for Defect). The action selected by each of the players determines the magnitude of the payoff that each of them receives in that time-step: CC-payoff, CD-payoff, DC-payoff, or DD-payoff, where AB-payoff denotes the payoff that an individual choosing action A obtains when her counterpart chooses action B.

After having decided to cooperate or not, both individuals simultaneously decide whether to stay with their current partner (S, for Stay), or break the partnership (L, for Leave). The Stay/Leave decision may be conditioned on the current partner's preceding action (C or D) -hence the name "conditional dissociation". If any one of the two individuals in a pair decides to leave, the partnership is broken and both individuals become single.

The strategies

Each individual carries three "genes" that shape her behaviour.

A strategy thus specifies a course of action at every possible juncture in an individual's life. We permit all 18 possible strategies or decision rules. Hence, there are (18·18+18)/2 = 171 strategy pairs.

The time line

The population process advances in discrete time-steps t = 1, 2, 3,... which are called "ticks" in the interface of the model. In each time-step, the following events occur in sequence (see Figure above):

  1. Random matching of singles. At the beginning of each time-step, single individuals are randomly matched in pairs (partnerships), with all matches being equally likely. At the beginning of the very first time-step all individuals are single.
  2. Pairwise interaction. All pairs play the 2x2 symmetric game described above once. Each individual's action (C or D) is determined by her decision rule, as applied to her immediate past experience, i.e., whether she is involved in a new partnership or not, and, in the latter case, the action that her partner took in the preceding time-step.
  3. Conditional dissociation. Every individual decides whether to leave or stay, according to her decision rule. Individuals involved in partnerships that are broken join the pool of singles entering the next time-step.
  4. Death. Individuals randomly die, an event that occurs independently for each individual with probability p in each time-step. Hence, the lifespan of an individual is geometrically distributed with mean-value exp-life = 1/p. The death of an individual renders her partner single, so those individuals whose partner died also join the pool of singles entering the next time-step.
  5. Replacement. Each dead individual is immediately replaced by a new entrant. This keeps the population size constant across all time-steps. These new entrants copy independently the decision rules of individuals in the population that started the current time-step; the probability that any particular individual's decision rule will be copied is proportional to the individual's payoff in the current time-step. Thus, only individuals who have played the game in this time-step can be copied (and all of them may be copied, even if they have died in this time-step). This copying is subject to random mutation, which occurs with a constant probability prob-mutation. In this event, the new entrant instead picks up a decision rule at random, with equal probability for all 18 decision rules. New entrants are also added to the pool of singles entering the next time-step. Thus, the pool of singles entering the following time-step will be composed of new entrants (imitators and mutants), separated individuals from broken pairs, and "widowed" individuals (those whose partner died). For num-players even, the number of singles is always even.




Population parameters


Definition of regimes

The model can track the number of time-steps that the system has been in one of two possible regimes and the number of transitions in between these two regimes. The regimes are called cc-regime and dd-regime. At any given time-step:




Importantly, all plots refer to the situation before any player dies and is replaced.


Cyp+By_applet is a model designed to analyse the effect of conditional dissociation in the evolutionary emergence of cooperation.
Copyright (C) 2009 Segismundo S. Izquierdo & Luis R. Izquierdo

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You can download a copy of the GNU General Public License by clicking here; you can also get a printed copy writing to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.

Contact information:
Segismundo S. Izquierdo
University of Valladolid, Spain.


This program has been designed and implemented by Segismundo S. Izquierdo & Luis R. Izquierdo.