Week 9

    Cards (64)

    • Evolutionary symmetric games
      Games in finite and well-mixed populations of size N, modelled as birth-death processes
    • eC
      (1, 0)T
    • eD
      (0, 1)T
    • x
      Strategy profile, state of the population described by x = 0.4eC + 0.6eD = (0.4, 0.6)T
    • Payoff matrix
    • Replicator dynamics

      ẋ = x(1 - x)[x(a - c) + (1 - x)(b - d)]
    • When N → ∞, we recover replicator dynamics
    • When N < ∞, x(t) fluctuates, requiring stochastic modelling
    • Coexistence equilibrium
      x* = (d - b) / (a - c + d - b)
    • Coexistence equilibrium is metastable when b > d, c > a in a finite population
    • Birth and death rates are functions of the fitness of C and D
    • Moran process
      λi = (i/N)(N-i/N)(fC(i)/f̄(i))
      μi = (i/N)(N-i/N)(fD(i)/f̄(i))
      where f̄(i) = (i/N)fC(i) + (1 - i/N)fD(i)
    • Fermi process
      λi = (i/N)(N-i/N)2 / (1 + e(fD(i)-fC(i)))
      μi = (i/N)(N-i/N)2 / (1 + e-(fD(i)-fC(i)))
    • Dynamics with Moran and Fermi processes are the same in the weak selection limit (s << 1)
    • Fixation probability of C

      φC
      i
    • Fixation probability of D

      φD
      i
    • Fixation probabilities satisfy a 2nd-order linear map
    • Fixation probability of a single D
      φD = 1 - φC
    • Neutral dynamics (s = 0)

      • φC
      i = i/N
      φC = 1/N
    • Constant fitness (fC(i) = r, fD(i) = 1)

      • φC
      i = (rN-i - 1) / (rN - 1)
      φC = (rN-1 - 1) / (rN - 1)
    • Goal: Fixation probability in 2-player evolutionary games with 2 pure strategies in finite populations
    • Pure strategies
      Strategies where the player always chooses the same action
    • Infinite populations
      Populations where the number of individuals is very large
    • Weak selection
      Selection intensity is small (0<s<<1)
    • Fixation probability of C, ϕC, has an exact formula
    • Fixation probability of C, ϕC, has examples
    • Fixation probability of C, ϕC, has a weak selection limit
    • Birth-death process
      Stochastic process where individuals are born and die, leading to changes in the population composition
    • Absorbing boundaries

      States where the process cannot leave once entered (fixation of C or D)
    • In 2-player games with 2 pure strategies, the outcome is fixation of either C or D
    • Fitness-dependent Moran process

      Birth-death process where the birth and death rates depend on the fitness of the individuals
    • Fitness-dependent Fermi process

      Birth-death process where the birth and death rates depend on the fitness of the individuals via a Fermi function
    • Under weak selection, the fixation probability of a single C has a simple expression
    • Selection favours fixation of C
      When the fixation probability of C is greater than 1/N
    • Selection opposes fixation of C
      When the fixation probability of C is less than 1/N
    • Under weak selection and large N, selection favours fixation of C if a+2b > c+2d
    • Under weak selection and large N, selection opposes fixation of C if a+2b < c+2d
    • Coordination game
      Game where a > c and d > b
    • In coordination games under weak selection and large N, there is an unstable coexistence equilibrium at a frequency x* < 1/3 of C
    • In coordination games under weak selection and large N, selection favours fixation of C if the basin of attraction of x* is greater than 2/3
    See similar decks