Game theory Flashcards

1
Q

Which players gains are shown on a payoff matrix

A

the player on the left / the vertical part

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

There is stable solution if…

A

the maximum value of the row of minima is the same as the minimum value of the column of maxima

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

playsafe strategy for horizontal player/ player 1

A

The maximum value of the column of row minima

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

playsafe strategy for vertical player/ player 2

A

the minimum value of the row of column maxima

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

maximum outcomes when finding playsafe

A

The maximum outcomes of each column

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

minimum outcomes when finding playsafe

A

The minimum outcomes of each row

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Dominance argument for vertical player/ player 1

A

If the outcomes of row R are always smaller than the outcomes of row P then P dominates R

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Dominance argument for horizontal player/ player 2

A

If the outcomes of column R are always smaller than the outcomes of column P then R dominates P

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Finding the optimal mixed strategy for 2x2 payoff matrix

A
  1. assign probabilities p and 1-p to the two options for player 1
  2. find expressions in terms of p for As payoff under each of Bs options
  3. Equate these two expressions to find p
  4. work out the value of the game for this value of p
  5. repeat for player B
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

player 1

A

The vertical player, the game is from their perspective

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

value of the game

A

the payoff from the strategies used

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

zero sum

A

one persons gains = net losses of other components

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

optimal mixed strategy 2xn matrix

A
  1. check for dominance if there is then just do 2x2 method
  2. otherwise assign probabilities p and 1-p to player 1s strategies
  3. find expressions in terms of p for As expected payoff under each of Bs options
  4. plot each of these expressions between 0 and 1 and shade the region under all the lines
  5. find the value of p at the highest point of the shaded area
  6. calculate the expected payoff for this value of p
  7. repeat for player B
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

optimal mixed strategy for nx2 payoff matrix

A

reflect the matrix in the leading diagonal and swap the signs then follow the 2xn method

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Finding the nash equilibrium

A
  1. circle the best choice for each player when the other player plays each of their strategies
  2. if there is a box which has two circles then that is the nash equilibrium
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

do all games have a nash equilibrium

A

no

17
Q

weak dominance

A

if two of the strategies give an equal value

18
Q

non strict nash equilibrium occurs when…

A

there is weak dominance

19
Q

formulating a game as an LP problem

A

positive (and non zero) by adding a constant to each value

  1. let the new value of this matrix = v, objective function maximise P = v- constant
  2. let p, q r equal the probability player 1 chooses each strategy
  3. find expressions in terms of p, q, r for As payoff under each of Bs options, these are the conditions (all ≤)
  4. solve using simplex
20
Q

objective function from a game theory problem

A

P = v - constant used to make the whole game positive

21
Q

if you need to solve game theory with LP from other perspective then

A

reflect matrix in leading diagonal

swap all the signs

22
Q

finding the LP conditions from a game theory problem

A
  • let p, q r equal the probability player 1 chooses each strategy
  • find expressions in terms of p, q, r for As payoff under each of Bs options, these are the conditions (all v ≤)
23
Q

LP conditions from game theory

A

v ≤ ap + bq + cr

24
Q

what time of LP problem is a game theory problem

A

maximise