| Exam Board | AQA |
|---|---|
| Module | D2 (Decision Mathematics 2) |
| Year | 2012 |
| Session | January |
| Marks | 13 |
| Paper | Download PDF ↗ |
| Mark scheme | Download PDF ↗ |
| Topic | Groups |
| Difficulty | Easy -2.5 This is a misclassified question - it's about game theory (zero-sum games, pay-off matrices, optimal strategies) from Decision Mathematics, not group theory from Further Pure Mathematics. The question involves straightforward application of standard algorithms: defining zero-sum games, finding play-safe strategies by comparing row minima/column maxima, dominance reduction, and solving 2×2 mixed strategy problems using basic linear methods. These are routine D2 procedures requiring minimal mathematical sophistication. |
| Spec | 7.08a Pay-off matrix: zero-sum games7.08b Dominance: reduce pay-off matrix7.08c Pure strategies: play-safe strategies and stable solutions7.08d Nash equilibrium: identification and interpretation7.08e Mixed strategies: optimal strategy using equations or graphical method |
| Colum | ||||
| \cline { 2 - 5 } | Strategy | \(\mathbf { C } _ { \mathbf { 1 } }\) | \(\mathbf { C } _ { \mathbf { 2 } }\) | \(\mathbf { C } _ { \mathbf { 3 } }\) |
| \multirow{3}{*}{\(\operatorname { Roz }\)} | \(\mathbf { R } _ { \mathbf { 1 } }\) | - 2 | - 6 | - 1 |
| \cline { 2 - 5 } | \(\mathbf { R } _ { \mathbf { 2 } }\) | - 5 | 2 | - 6 |
| \cline { 2 - 5 } | \(\mathbf { R } _ { \mathbf { 3 } }\) | - 3 | 3 | - 4 |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Mark | Guidance |
| The gain of one player equals the loss of the other player | B1 | Must mention both gain and loss |
| The sum of the payoffs (to both players) is zero | B1 | Accept equivalent statement |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Mark | Guidance |
| Colum wishes to minimise the pay-off to Roz, so looks at maximum in each column | M1 | Must show column maxima: \(C_1: -2\), \(C_2: 3\), \(C_3: -1\) |
| Play-safe strategy is \(C_1\) (choosing minimum of these maxima = \(-2\)) | A1 | Must give reason: minimax = \(-2\) |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Mark | Guidance |
| Row minima: \(R_1: -6\), \(R_2: -6\), \(R_3: -4\) | M1 | |
| \(R_2\) is dominated by \(R_3\) (every entry in \(R_3 \geq\) every entry in \(R_2\)): \(-3>-5\), \(3>2\), \(-4>-6\) | A1 | Accept \(R_1\) dominated by \(R_3\): \(-3>-2\) fails — must show \(R_2\) dominated by \(R_3\) |
| Answer | Marks | Guidance |
|---|---|---|
| Answer | Mark | Guidance |
| Let Roz play \(R_1\) with probability \(p\) and \(R_3\) with probability \((1-p)\) | M1 | |
| Expected payoff against \(C_1\): \(-2p + (-3)(1-p) = p - 3\) | A1 | |
| Expected payoff against \(C_2\): \(-6p + 3(1-p) = 3 - 9p\) | A1 | |
| Expected payoff against \(C_3\): \(-p + (-4)(1-p) = 3p - 4\) | A1 | |
| Setting \(C_1 = C_2\): \(p - 3 = 3 - 9p \Rightarrow p = \frac{3}{5}\) | M1 | |
| Check \(C_3\) at \(p = \frac{3}{5}\): \(3(\frac{3}{5}) - 4 = -\frac{11}{5}\), which is less than \(-\frac{12}{5}\) so \(C_3\) not relevant | A1 | |
| Optimal strategy: play \(R_1\) with probability \(\frac{3}{5}\), \(R_3\) with probability \(\frac{2}{5}\), value \(= -\frac{12}{5}\) | A1 |
# Question 3:
## Part (a)
| Answer | Mark | Guidance |
|--------|------|----------|
| The gain of one player equals the loss of the other player | B1 | Must mention both gain and loss |
| The sum of the payoffs (to both players) is zero | B1 | Accept equivalent statement |
## Part (b)
| Answer | Mark | Guidance |
|--------|------|----------|
| Colum wishes to minimise the pay-off to Roz, so looks at maximum in each column | M1 | Must show column maxima: $C_1: -2$, $C_2: 3$, $C_3: -1$ |
| Play-safe strategy is $C_1$ (choosing minimum of these maxima = $-2$) | A1 | Must give reason: minimax = $-2$ |
## Part (c)(i)
| Answer | Mark | Guidance |
|--------|------|----------|
| Row minima: $R_1: -6$, $R_2: -6$, $R_3: -4$ | M1 | |
| $R_2$ is dominated by $R_3$ (every entry in $R_3 \geq$ every entry in $R_2$): $-3>-5$, $3>2$, $-4>-6$ | A1 | Accept $R_1$ dominated by $R_3$: $-3>-2$ fails — must show $R_2$ dominated by $R_3$ |
## Part (c)(ii)
| Answer | Mark | Guidance |
|--------|------|----------|
| Let Roz play $R_1$ with probability $p$ and $R_3$ with probability $(1-p)$ | M1 | |
| Expected payoff against $C_1$: $-2p + (-3)(1-p) = p - 3$ | A1 | |
| Expected payoff against $C_2$: $-6p + 3(1-p) = 3 - 9p$ | A1 | |
| Expected payoff against $C_3$: $-p + (-4)(1-p) = 3p - 4$ | A1 | |
| Setting $C_1 = C_2$: $p - 3 = 3 - 9p \Rightarrow p = \frac{3}{5}$ | M1 | |
| Check $C_3$ at $p = \frac{3}{5}$: $3(\frac{3}{5}) - 4 = -\frac{11}{5}$, which is less than $-\frac{12}{5}$ so $C_3$ not relevant | A1 | |
| Optimal strategy: play $R_1$ with probability $\frac{3}{5}$, $R_3$ with probability $\frac{2}{5}$, value $= -\frac{12}{5}$ | A1 | |
---
3 Two people, Roz and Colum, play a zero-sum game. The game is represented by the following pay-off matrix for Roz.
\begin{center}
\begin{tabular}{ l | c | c | c | c | }
\multicolumn{4}{c}{Colum} & \\
\cline { 2 - 5 }
& Strategy & $\mathbf { C } _ { \mathbf { 1 } }$ & $\mathbf { C } _ { \mathbf { 2 } }$ & $\mathbf { C } _ { \mathbf { 3 } }$ \\
\hline
\multirow{3}{*}{$\operatorname { Roz }$} & $\mathbf { R } _ { \mathbf { 1 } }$ & - 2 & - 6 & - 1 \\
\cline { 2 - 5 }
& $\mathbf { R } _ { \mathbf { 2 } }$ & - 5 & 2 & - 6 \\
\cline { 2 - 5 }
& $\mathbf { R } _ { \mathbf { 3 } }$ & - 3 & 3 & - 4 \\
\hline
\end{tabular}
\end{center}
\begin{enumerate}[label=(\alph*)]
\item Explain what is meant by the term 'zero-sum game'.
\item Determine the play-safe strategy for Colum, giving a reason for your answer.
\item \begin{enumerate}[label=(\roman*)]
\item Show that the matrix can be reduced to a 2 by 3 matrix, giving the reason for deleting one of the rows.
\item Hence find the optimal mixed strategy for Roz.
\end{enumerate}\end{enumerate}
\hfill \mbox{\textit{AQA D2 2012 Q3 [13]}}