Two-Person Zero-Sum Game/Examples/Abstract 2
Example of Two-Person Zero-Sum Game
The two players are $A$ and $B$.
Player $A$ has $m$ strategies: $A_1, A_2, \ldots, A_m$.
Player $B$ has $n$ strategies: $B_1, B_2, \ldots, B_n$.
The game is zero-sum in that a payoff of $p$ to $A$ corresponds to a payoff of $-p$ to $B$.
The following table specifies the payoff to $A$ for each strategy of $A$ and $B$.
$B$'s strategy | ||
$A$'s strategy | $\begin{array}{r {{|}} c {{|}} c {{|}} c {{|}} }
& B_1 & B_2 & B_3 \\ \hline A_1 & -2 & 1 & 1 \\ \hline A_2 & -3 & 0 & 2 \\ \hline A_3 & -4 & -6 & 4 \\ \hline \end{array}$ |
This has an equilibrium point at $\tuple {A_1, B_1}$ with a payoff to $A$ of $-2$.
Proof
By inspection, strategy $B_1$ dominates strategy $B_3$.
Therefore by Elimination by Domination, $B_3$ can be eliminated.
$B_3$ having been eliminated, it is seen by inspection that $A_1$ dominates both $A_2$ and $A_3$.
$A_2$ and $A_3$ having been eliminated, it is seen by inspection that $B_1$ dominates $B_2$.
The optimum strategies are therefore $A_1$ and $B_1$, and the payoff to $A$ is $-2$.
Given $B_1$, every other strategy adopted by $A$ will result in a smaller payoff to $A$.
Given $A_1$, every other strategy adopted by $B$ will result in a larger payoff to $A$.
Hence, by definition, $\tuple {A_1, B_1}$ is an equilibrium point.
$\blacksquare$
Sources
- 1983: Morton D. Davis: Game Theory (revised ed.) ... (previous) ... (next): $\S 2$: The Two-Person, Zero-Sum Game with Equilibrium Points: Introductory Problems: Figure $2.2$