Dominance and mixed strategies

Dominance and mixed strategies

Spec DF4–DF5 · AQA A-Level Further Maths: Revision notes

Studying this course? Keep your progress in one place.

Key facts

  • A dominated strategy is never better than another, so remove it; repeat until nothing more can be removed.
  • Rowan (maximiser) prefers larger entries; Colin (minimiser) prefers smaller entries.
  • A mixed strategy plays each strategy with a stated probability, and the probabilities sum to 1.
  • For a 2×22\times2 game, Rowan's optimal pp makes his expected pay-offs against the two columns equal.
  • For 2×n2\times n games draw the lines and take the highest point of the lower boundary.

Dominated strategies

A strategy is dominated if another is at least as good in every case, so a sensible player never uses it.

Remove dominated rows and columns to get a smaller game. Repeat, because removing a row can make a column dominated and vice versa.

Say which row or column dominates which, and quote the entry-by-entry comparison.

Rows (Rowan maximises)

  • Row AA is dominated by row BB
  • if every entry of BB is greater than or equal to the matching entry of AA

Columns (Colin minimises)

  • Column AA is dominated by column BB
  • if every entry of BB is less than or equal to the matching entry of AA

Worked example

Reduce (324152435)\begin{pmatrix} 3 & 2 & 4 \\ 1 & 5 & 2 \\ 4 & 3 & 5 \end{pmatrix} using dominance.

Column A=(2,6)A=(2,6) and column B=(1,5)B=(1,5) are Colin's. Which is dominated?

Mixed strategies

If there is no stable solution, a player should choose strategies at random with set probabilities.

A player who always makes the same choice can be exploited. A mixed strategy chooses each strategy with a stated probability, at random each time.

Rowan plays row 1 with probability pp and row 2 with probability 1−p1-p. His expected pay-off against each of Colin's columns is a linear expression in pp.

The optimal mixed strategy maximises the smallest of these. For a 2×22\times2 game that happens where the two expected pay-offs are equal, and the result is the value of the game.

  • Rowan's mixrow 1 with probability pp, row 2 with 1−p1-p
  • Optimal pp in a 2×22\times2 gameexpected pay-offs are equal

Rowan plays row 1 with probability 0.30.3 in a 2×22\times2 game. What is the probability of row 2?

Solving a 2×2 game

Write Rowan's expected pay-off against each column in terms of pp, then equate them.

For (5124)\begin{pmatrix} 5 & 1 \\ 2 & 4 \end{pmatrix} there is no stable solution: maximin =2=2 and minimax =4=4.

Rowan plays row 1 with probability pp and row 2 with probability 1−p1-p, and we equate his expected pay-off against each column. The graph shows the two expected pay-offs meeting at the optimum.

0.10.20.30.40.50.60.70.80.91123456xy(1/3, 3)Column 1: 3p + 2Column 2: 4 − 3p
Expected pay-offs meet at p = 1/3, value 3.

Worked example

Find the optimal mixed strategies and the value for (5124)\begin{pmatrix} 5 & 1 \\ 2 & 4 \end{pmatrix}.

For (3102)\begin{pmatrix} 3 & 1 \\ 0 & 2 \end{pmatrix}, Rowan plays row 1 with probability pp. Which equation gives his optimal pp?

Graphical method for 2×n games

Draw Rowan's expected pay-off against each column as a line in pp; the highest point of the lower boundary is optimal.

If Rowan has two strategies and Colin has nn, write Rowan's expected pay-off against each column as a line in pp and draw the lines on one set of axes.

Colin chooses the column that gives Rowan least, so Rowan's guaranteed pay-off is the lower boundary. His optimal pp is at its highest point, and the height there is the value of the game. Any line above that point is a column Colin never plays.

Example: (251416)\begin{pmatrix} 2 & 5 & 1 \\ 4 & 1 & 6 \end{pmatrix} gives lines 4−2p4-2p, 1+4p1+4p and 6−5p6-5p. The first two meet at p=12p=\frac12 with value 33, and the third is 3.53.5 there.

0.10.20.30.40.50.60.70.80.91123456xy(0.5, 3)Column 1: 4 − 2pColumn 2: 1 + 4pColumn 3: 6 − 5p
Lower boundary peaks at p = 1/2, value 3. Column 3 lies above.

If Rowan has mm strategies and Colin has two, use qq for Colin and find the lowest point of the upper boundary of the lines.

In the graph, which column does Colin never play, and why?

Try an exam question

The pay-off matrix to Rowan is (4123)\begin{pmatrix} 4 & 1 \\ 2 & 3 \end{pmatrix}. Show that there is no stable solution, and find Rowan's optimal mixed strategy and the value of the game.

[5 marks]

That's the notes covered.

Carry on to the next subtopic.

Try a question

Which statement about dominance in this game is correct?

1 mark, marked against the mark scheme

Revision notes