Decision and Games

Excel Solver Treeplan
SilverDecisions Gambit SageMath

Maximin
Maximin
Minmax Regret
Minmax Regret
Maximax
The Principle of Insufficient Reason
Decision aking Under Risk
Expected Value Criterion
)Expected Monetary Value (EMV
Expected Value of Perfect Information
EVPI
Decision Trees

Treeplan
Rolling Back

SilverDecision
SilverDecision
Case Studies

Tree of the Game
Matrix of the Game
Two Persons Zero Sum Games

Dominance
Game Information
)(

6
6
9
9
9
9
9
11
12
15
16
18
21
22
21
22
23
24
24
27
40
42
44
44
53
67
92
94
99
100
116
120
129
153
153
154
157
158

Excel Solver
Equilibrium
Dominant Strategy Equilibrium
Nash Equilibrium

Gambit
Gambit
Gambit
Gambit
Extensive Form
Iterated dominance Equilibrium
Weak Iteration
Pareto Efficiency
Pareto Dominance
Gambit
Case Studies

)(

Shapley and Nash Method
N-Person Games

The Core
Stable Sets
The Nucleus
Shaplet Function Values
159
163
163
168
178
179
180
180
181
182
183
192
198
200
204
214
218
218
225
238
257
259
270
275
282
282
287
287
289
291
294
311
318
315
331
334
335
341
347
351

Sage
Sage
Sage
- Many-Person Games
Coalition Form and Characteristic Function
Supersdditivity Monotone Constant-Sum
Imputation and the Core

Essential Games
The Core

Sage
Matching Games
Combinatorial Games
Simple Take-Away Game
P P-positions and N-positions N
Subtraction Games
The Game of Nim
Nim Sum

Utility Theory
: LINGO
: SageMath
:
351
373
386
391
399
399
399
400
409
409
410
410
411
412
412
428
434
434
436
437
439
440
443
446
463
482
484
488
499

] .
" "
.

.
.
.
) (
.

Excel Excel Solver Treeplan
Gambit Sage
.SageMath

Linear Program Simplex Tableau
Excel Solver
) ( .

) (.

http://www.abarry.ws/
] .
2015/8/23

Decision Analysis
Alternatives
Uncertainity .
Optimal Payoff
.Decision Criterion
Maximizing Expected Profit
.
Maximizing the Utility Function

Risks .
Payoff Table Analysis

:
-1 .Discrete
-2 .
:
-1 .
-2 .
-3 ) States of Nature
) Mutually Exclusive
() and Collectively Exhaustive
(.
6
-4 .
1000 .
5 :
-1 .Gold
-2 .Bond
-3 .Stock
-4 .Certificate of Deposit
-5 .Stock Option Hedge
) (
.
.
: .

-1 .
-2 .
-3 .
-4 )( .
S1
S2
S3
S4
A1
p(1,1)
p(1,2)
p(1,3)
p(1,4)
p1
A2
p(2,1)
p(2,2)
p(2,3)
p(2,4)
p2
A3
p(3,1)
p(3,2)
p(3,3)
p(3,4)
p3

.
. . -
Decision
Alternativs
Gold
Bond
Stock
C/D account
Stock option
Decision
Alternativs
Gold
Bond
Stock
C/D account
Stock option
Larg rise
-100
250
500
60
200
States of Nature
Small rise
No change
Small fall
100
200
300
200
150
-100
250
100
-200
60
60
60
150
150
-200
Large fall
0
-150
-600
60
-150
Larg rise
-100
250
500
60
200
Small rise
100
200
250
60
150
Large fall
0
-150
-600
60
-150
No change
200
150
100
60
150
Small fall
300
-100
-200
60
-200
: Dominance :
A1
A2 A1 . A2

.
Decision Making Criteria

:
-1 : Future
.State-of-Nature
-2 :
.
-3 :
.
Decision Making Under Uncertainty
-1 .
-2 :
Maximin . Minimax Regret . Maximax . Principle of Insufficient Reasoning .
9
: Maximin
Worst-Case Scenario
. . . : -1 .
-2 " ".
Minimum
Payoff
-100
-150
-600
60
Large fall
0
-150
-600
60
States of Nature - The Maximum Criterion

Larg rise Small rise No change Small fall
-100
100
200
300
250
200
150
-100
500
250
100
-200
60
60
60
60
10
Decision
Alternativs
Gold
Bond
Stock
C/D
account
Maxmin
Maxmin
11
: Minimax Regret
:Minimax Regret
-1 .
-2 " "Lost Opportunity
"."Regret
-3 .
: -1 .
-2 Regret .
Maximum Regret . - .
12

Decision
Gold
Bond
Stock
C/D
account
Decision
Alternativs
Gold
Bond
Stock
C/D
account
States of Nature - The Payoff Table

Larg rise Small rise No change
-100
100
200
250
200
150
500
250
100
60
60
60
Small fall
300
-100
-200
60
Large fall
0
-150
-600
60
States of Nature - The Maximum Criterion

600
150
0
0
250
50
50
400
0
0
100
500
440
190
140
240
Large fall
60
210
660
0
= : :
-
. Stock
Decision
Gold
Bond
Stock
C/D
account
States of Nature - The Payoff Table

-100
100
200
250
200
150
500
250
100
60
60
60
500 - (-100) = 600
13
Small fall
300
-100
-200
60
Large fall
0
-150
-600
60
Decision
Alternativs
Gold
Bond
Stock
C/D
account
States of Nature - The Regret Table

-100
100
200
250
200
150
500
250
100
60
60
60
Small fall
300
-100
-200
60
Large fall
0
-150
-600
60
Minimum
Regret
600
400
660
440
. 600
.400 Bonds
14
Minimax Regret
15
16
: Maximax
.
.
) (.
: -1 .
-2 .

Minimum
Payoff
300
200
500
60
Large fall
0
-150
-600
60
States of Nature - The Maximax Criterion

-100
100
200
300
250
200
150
-100
500
250
100
-200
60
60
60
60
17
Decision
Alternativs
Gold
Bond
Stock
C/D
account
18
- Decision Making Under

Uncertainty - The Principle of Insufficient Reason
)(.
. : -1 .
-2 .

600 Gold -1
350 Bond -2
50 Stock -3
300 CD Acc -4
- .
19
20
Decision Making Under Risk

) ( .
.Expected Payoff
Expected Value Criterion
:)Expected Payoff = (Probability)(Payoff

. : " " Expected Monetary
) .Value ( EMV
)Expected Monetary Value (EMV
21

: -1 .
-2 .
Expected Value of Perfect Information
Gain Expected Return
:
)Expected Value of Perfect Information (EVPI
EVPI
" "
Stock
22
EVPI
EVPI = ERPI - EREV
Expected Return with Perfect Information :ERPI

ERPI = (probability)(Max payoff )
Expected Return of the EV Criterion :EREV
EV
23
Decision Trees
Nonsequential .Single Stage
.
Multistage.Decision Processes

Chronological . - Nodes .Branches
24
. )( .
3 . . : -1 Event Node .
Chance Node Uncontrollable
.
-2 ) Decision Node (
Controllable .
-3 )( Terminal Node )
(Combination .
.

) Mutually Exclusive (
) and Collectively Exhaustive ( .
Subjective .
Decision Branch .
25
) Mutually Exclusive
() and Collectively Exhaustive (.
.Event Branch : Terminal Values ) Payoff Value Outcome Value
.Endpoint Value

26
Treeplan
Treeplan Excel Add-In
Treeplan Excel .
Excel: -1 Tools => Add-Ins...:
27
Treeplan Excel Treeplan

Excel .
-2 Treeplan Excel
File => Open treeplan.xla
xla .Excel
Tools => Decision Tree... :
28
Ctrl-t : Treeplan
29
New Tree:
Ctrl-t
:
30
) Decision 1 Decision 2

31
32
)( Treeplan
http://www.treeplan.com
:
BGD
:
: -1 300000.
-2 500000.
-3 950000 .
-4 30000.
%40 . 260000.
3 20000 .
5000. :1- P(Consultant predict approval| approval granted) = 0.7
2- P(Consultant predict denial| approval denied) = 0.8
: -1 / .
-2 .
33

-1 .
-2 :
Treeplan
34
35
36

.
A B

. .

. :
: .

) 3( .
:
.A .B A .B37
. :
.A .B Treeplan
-
38
39
Rolling Back

Expected Monetary Value ) EMV (
. EMV .
EMV A
0.4 :
Max(EMV) = 3.4
40
) ( Payoff
: 2 .EMV 2
.EMV=3.4
41
-1 EMV
-2 Maximum Possible Payoff
Maximax
-3 EVC .EVPI
Rolling Back
)
( Event Node
1 0.4 13 0.6
12 1
EMV(node 1) =0.4 x 13 + 0.6 x -12 = -2.0
EMV(node 2) =0.4 x -8 + 0.6 x 11 = 3.4
EMV(node 3) =0.4 x 5 + 0.6 x -1 = 1.4
EMV(node 4) =0.4 x 0 + 0.6 x 0 = 0.0
EMV 0 4
.-2 ,3.4 ,1.4 ,0
.EMV
42
EMV(node 0) = 3.4
EMV " .B
43
SilverDecisions
Open-Source
Decision Support Analysis Division Warsaw School of
Economics

Microsoft Silverlight .
:
-1 Microsoft Silverlight
http://www.microsoft.com/getsilverlight/GetStarted/Install/Default.aspx
-2
http://silverdecisions.pl/
44
-3 Run
45

.
:
-1
-
46
Decision Tree Designer
Label .
47
48
49
50
.Bold lines
:
51
52
Case Studies :
:1
4
:
20 :1 0.4 .
30 :2 20 0.3 .
40 :3 30 0.2 .
60 :4 100 0.1 .

:
Probability
0.20
0.20
0.30
0.20
0.10
Time
minutes
minutes
minutes
minutes
minutes
10
30
60
100
150

:
:
1 10 = 20 24= 0.4 X10 +
1 30 = 20 32= 0.4 X30 +
1 60 = 20 44= 0.4 X60 +
53
1 100 = 20 60= 0.4 X100 +

1 150 = 20 80= 0.4 X150 +
2 10 = 30 = 30 ) 20 (
2 30 = 30 33= 0.3 X10 + ) 20
(
2 60 = 30 42= 0.3 X40 + ) 20
(
2 100 = 30 54= 0.3 X80 + ) 20
(
2 150 = 30 69= 0.3 X130 + )20
(
3 10 = 40 = 40 ) 30 (
3 30 = 40 = 40 ) 30 (
3 60 = 40 46= 0.2 X 30 + ) 30
(
3 100 = 40 54= 0.2 X 70 + ) 30
(
3 150 = 40 64= 0.2 X 120 + )30
(
4 10 = 60 = 60 ) 100 (
4 30 = 60 = 60 ) 100 (
4 60 = 60 = 60 ) 100 (
4 100 = 60 = 60 ) 100 (
54
4 150 = 60 65= 0.1 X50 + ) 100

(
55
:2
.
800 1100
. 1 4 .
%50 .
100 .

maximin .maximum regret
:
:
56

P(Demand = 1) = 0.30
P(Demand = 2) = 0.10
P(Demand = 3) = 0.40
P(Demand = 4) = 0.20
57
:3

. 3 .
expected present worth profit
.
)(:
Average Number of Daily Customers
4
8
12
16
20
150
350
850
200
400
500
200
350
250
150
50
-100
100
-200
-400
A
B
C
Restaurant
Format

:
P(4) = 0.10, P(8) = 0.20, P(12) = 0.30, P(16) =0.30

:
.
58
:4

. 300 650 400
.
50 .
200
. 1 4
100
.
:
:
3 2 1300
2 1200 ) (( 50 - 650) X 2
200
1,200 + 200 - 1,300 = 100.

4
0
-400
0
400
800
3
0
-300
100
500
400
Demand
2
0
-200
200
100
0
1
0
-100
-200
-300
-400
0
1
2
3
4
Number
Ordered
(1 Maxmin
(2
59
(3
:
.
60
:5

. 3 .
.
.
:
Demand Growth for Chemical
Low
0
100
-300
-500
Medium
0
130
240
120
High
0
140
Do Nothing
Minor
150
200
Moderate
Major
Compound
Expansion
(1 minimize maximum
regret
(2 principle of
insufficient reason
(3 :
P(High Growth) = 0.20
P(Medium Growth) =0 .30
= 0.50

61
)P(Low Growth
(4
.
:
P(Expert predicts positive | High Growth) = 0.60
P(Expert predicts positive | Medium Growth) = 0.40
P(Expert predicts positive | Low Growth) = 0.20

:
.
62
:6
1 4 .
:
Total Cost
110,000
150,000
230,000
315,000
Number of Cars Ordered

1
2
3
4
90,000 .
75,000 .
5,000
.
:
Number of Cars Demanded by Customers
1
2
3
4
-20
-25
-30
-35
15
30
25
20
10
25
40
35
0
15
30
45
1
2
3
4
Number of
Cars
Purchased
(1
(2 :
Probability
.20
.20
.30
.30
63
Demand
1
2
3
4

(3
:
P(favorably inclined | 1 Cars demanded) = 0.20

:
.
64
:7
. 4

.
. :
High
600
540
490
470
State
Moderate
350
350
375
350
Low
200
250
300
300
Action
1
2
3
4
. 3
4 4 .
Max
600
540
490
Min
200
250
300
High
600
540
490
State
Moderate
350
350
375
Low
200
250
300
Action
1
2
3

- MAXIMIN
- MAXIMAX
- MINIMAX REGRET CRITERIA.
MAXIMIN
65
3 300 1 200
2 .250
MAXIMAX
1 600 2 540
3 .490
MINIMAX REGRET
:
i }A={1,2,3,4
) (
lij = max rmj rij

mA
Max Regret
100
60
110
High
0
60
110
State
Moderate
25
25
0
Low
100
50
0
Action
1
2
3
2 .60

P(Low) = 0.1
P(Moderate) = 0.5
P(High) = 0.4
66
Expected Monetary Value

Criterion .

Expected Value
435
416
413
)3 (High
0.4
600
540
490
)2 (Moderate
0.5
350
350
375
)1 (Low
0.1
200
250
300
State
Probability
)1 (A
)2 (B
)3 (C
Action
1 435.
:
:
.
67
68
69
:8

. )(
1000
8000 . )( 1500
12000 .

2000.
:
a1: bid on neither contract
a2: bid on the first contract but not the second
a3: bid on the second contract but not the first
a4: bid on both contracts
:
s1: both bids rejected

s2: bid on first contract accepted but not the second
s3: bid on second contract accepted but not first
s4: both bids accepted
16
rij , i = 1, 2,..., 4, j = 1, 2,..., 4 :
- - ) (
2 3
70
0 - 1000 = -1000
:
s4
s3
s2
s1
0
7
10.5
7.5
0
-1
10.5
9.5
0
7
-1.5
5.5
0
-1
-1.5
-2.5
States
a1
a2
a3
a4
Actions
.
0.8 0.5

P[Both are rejected] = P[First rejected Second rejected] = (1 0.8)(1 0.5) = 0.1
P[First accepted and second rejected] = 0.8(1 0.5) = 0.4
P[First rejected and second accepted] = (1 0.8)(0.5) = 0.1
P[Both are accepted] = 0.8(0.5) = 0.4

s4
0.4
0
7
10.5
7.5
s3
0.1
0
-1
10.5
9.5
s2
0.4
0
7
-1.5
5.5
s1
0.1
0
-1
-1.5
-2.5
71
States
Probabilit
ies
a1
a2
a3
a4
Actions
:Maximin ) ( a1
:Maximax a2
:Minimax regret a4
: EMV
EMV(1) = 0.1(0) + 0.4(0) + 0.1(0) + 0.4(0) = 0
EMV(2) = 0.1(-1) + 0.4(7) + 0.1(-1) + 0.4(7) = 5.4
EMV(3) = 0.1(-1.5) + 0.4(-1.5) + 0.1(10.5) + 0.4(10.5) = 4.5
EMV(4) = 0.1(-2.5) + 0.4(5.5) + 0.1(9.5) + 0.4(7.5) = 5.9
a4
:
72
73
74
75
:9
85000
.

. 5000
50-50 .
microwave cellular
infrared
.
:
Equipment Cost
4,000
5,000
4,000
Technology
Microwave
Cellular
Infrared

) research and development (R&D
. best-case worst-case

:
Possible R&D Costs
Case
Worst Case
Prob.
Cost
Prob.
0.4
60,000
0.6
0.8
70,000
0.2
0.9
80,000
0.1
76
Best
Cost
30,000
40,000
40,000
Microwave
Cellular
Infrared
.
:
:
.
77
78
:
.
. 25
15 .
%75 .%30

) (:
Demand
High
Low
$175
$ 95
$125
$105
79
Factory Size
Large
Small

(1

State of Nature
3
35
75
50
60
50
40
30
35
40
Decision

)(2
.
.

. :
Weather Condition
very cold
Normal
Light
10
7
3
8
8
6
4
4
4
)Payoffs (in $1000s
Size of Order
Large
Medium
Small
0.25 0.6
0.15.
(
maximax
80
minimax regret
EMV
)(3

30000 . .
stock mutual fund one-year
) .certificate of deposit (CD %8 .
%16 %9 %2-
.
0.1 0.85 .0.05
(1
(2
maximax
(3
maximin
(4
minimax regret
(5
EMV
81
)(4
3
Fixed Monthly Payment
200
300
Additional Cost Per Kilo

0.095 per Kilo.
0.061 for the first 6,000
kilos; 0.050 thereafter.
0.000 for the first 6,000
kilos; 0.14 per kilo
thereafter.
170
Plan
I
II
III
15000 35000
:
P(driving 15,000 kilos) = 0.1
(1 .
(2 ) maximax (
(3 maximin
(4 minimax regret
(5 EMV
82
)(5
.
2.45
.3.95 1.25 .

:
20
19
18
17
16
14
15
13
12
11
10
Demand
Probability 0.02 0.06 0.09 0.11 0.13 0.15 0.18 0.11 0.07 0.05 0.03
(1 .
(2 maximax
(3 maximin
(4 minimax regret
(5 EMV
(6 2.25
15 .

)(6
. 200
.
)(
83
.
:
Dust Damage
105
0.08
90
75
60
45
30
15
)(in 1000
Probability 0.25 0.08 0.10 0.12 0.15 0.12 0.10
:
-1 47000 %100
.
-2 25000 35000
.
-3
.
( 1 .
(2 maximax
(3 maximin
(4 minimax regret
(5 EMV
(6 .
)(7

. ) (
84

:
Market Demand
Low
Medium
High
400
400
400
200
500
500
-400
300
800
)(Payoffs in 1000
Size of Development
Small
Medium
Large
%21.75
%35.5 .%42.75
(1 maximax
(2 maximin
(3 minimax regret
(4 EMV
( 5 .
85
)(8
) (

:
High
0.0100
0.0350
0.3825
Actual Demand
Low
Medium
0.1600 0.0300
0.0350 0.2800
0.0225 0.0450
Forecasted Demand
Low
Medium
High
) : %82.25
(
(1 .
(2 EMV
(3 EMV
)(9
.
3
. .
:
Cost of Test
$1.75
$2.00
$2.40
Probability of Failure
0.125
0.075
0.140
.
86
Component
X
Y
Z
)(10

300000 9 .
:
-1 ) fixed rate loan (FRL %9.
-2 ) adjustable rate loan (ARL
%6 .
%7 0.1 %9 0.25 %11
0.65.
-3 ) adjustable rate loan (ARL
%4 3
%6 0.05 %8 0.3 %10 .0.65
%1 0.1 %1
0.2 %3 0.7.
(1
.
(2 .expected total interest
87
)(11

16 7 .

10 0.4 6 .0.6
0.15
0.35 .0.50
120 28
. 5
.
(1 .
(2 EMV
)(12
.
%60 .

. .
1.25 %25
1.45 %45 1.85
%85 .
2.2 .
1.15.
88
-1 .
-2 EMV
)(13
10 100 .
%90
. %95
.

)(14
.
240000 450000
. . 1500
.
. )( 125000 4
5000
. )( 85000 3
.
.%40
)( )( . )(
.
)( 20000
89
.
.
-1 .
-2 EMV
90
(15)
91
(16
92
(17
93
(18
(19
94
(20
95
(21
96
(22
(23
97
(24
98
(25
(26
99
(27
100
(28
101
(29
102
(30
103
104
Game Theory
: .
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
: .
162
:
.
Two-person Game Theory:
.

Payoff .
: ) (
.

Constant Sum .Variable Sum
:
: Players .
:Strategies .
:Strategic Game
.
:Payoff
.
) ( :Utility
.
" :Equilibrium Strategy"
.
:Equilibrium in a Game
) ( .
163
) ( :Rational Play
.
Constant-Sum :Zero-Sum
.
Games of Pure Conflict
.
:Mixed Strategy
.
:Simultaneous-Move Game )
(
.
Sequential-Move :Dynamic Games
) (
.
: Single-Stage Games
Single-Stage One-Shot
.Unrepeated Games
:Repeated Games
Multi-Stage n-Stage

.Meta-Strategies
164
Cooperative :Non-Cooperative Games

.

) . : (.
N N :N-Player Games .
2-Players Two-Person
N-Player ) N > 2 :
2-Players (.
:Game Move
:
: conscious Move .
:Random Move .
165
Game Information

. )
( . :
:Perfect Information .
:Incomplete Information "" Nature "" Chance
.
:Asymmetric Information
.

Two-Players Zero-sum Games

) - = 1 .( 2
.
A B m n
Payoff Matrix A:
166
B2 Bn
a12 a1n
a22 a2n

am1 amn
B1
a11
a21

am1
A i B j
A aij B . aij
167
A1
A2
Am

:
. A
A1 A2 . A3 B
B1 B2 B3
. B4
. :A
Rowmin
3
5 Maxmin
9
B4
B3
B2
B1
8 2 9 3
6 5 6 8
2 4 9 5
Column 8 5
9 8
A1
A2
A3
Minimax
max
Best of the Worst

.Maxmin A A1
B A 3.0%
.B .
A2 A 5% B .
A3 9% A .B
" ."Rowmin A
168
A2 Maximin
"."Rowmin
.B A
B Minimax
B . B2
A2 B2
A 5%
B . Value of the Game
5% A B . Saddle-point Solution

B B1 B3 B4 A
A2 B ) %6 .(8%
A ) A2 A1 B B4
9% A A3 B B3 2% .(A
:
.
169
Excel

A B
H .T .
) HH (TT A B B.
A
A B
BT rowmin
1
1
1
1
1
170
BH
AH
1
AT
1
colmax
1
Maximin = -1
Minimax = +1
Minimax
Maxinin
. A
AH B BT .A
A AT .

.
Minimax Maximin

Maximin value value of the game Minimax value

:
-1 :
) ( .
.
-2 : .
171
Graphical Solution
) (2 x n A.
A A1 A2 x1
1 x1 . 0 x1 1 B B1 Bn y1
yn y j 0 j = 1, 2,..., n
y1 + y1 + + yn = 1
y2
y1
yn
B2 Bn
B1
a11 a12 a1n

a
21 a22 a2 n
x1 : A1
1 x1 : A2
A j B
:
a2 j ) x1 + a2 j , j = 1,2,..., n
(a
1j
A x1
:
172
max min ( a1 j a2 j ) x1 + a2 j
j
x1

2 x4 A:
B4
1
6
B3
B2
B1
2
4
A1
A2
. A
B :
A's expected payoff
. 0 x1 1
173
B's pure strategy
2 x1 + 4
x1 + 3
x1 + 2
7 x1 + 6
x1 = 0.5
3 4:
from line 3
from line 4
1
5
+
2
=
,
2
2
v=
7 1 + 6 = 5 ,
2
2
B B3 B4
y1 = y2 = 0 y4 = 1 y3
B A :
174
B' expected payoff
A's pure strategy
4 y3 -1
- 4 y3 + 6
B :
4 y3 1 = 4 y3 + 6
7
:
8
= y3
: .
A A1 A2
1 7
B B3 B4
8 8
: . .
175

x1 , x2 ,..., xm A :
m
m
max min ai1 xi , ai 2 xi ,..., ain xi

xi
i=1
i =1
i =1
x1 + x2 + + xm = 1
xi 0, i = 1,2,..., m
m
m
m
v = min ai1 xi , ai 2 xi ,..., ain xi

i=1
i =1
i =1
176
a x
ij i
v, j = 1, 2,..., n
i =1
: LP A
maximize
z=v
Subjet to
m
v aij xi 0,
j = 1, 2,..., n
i =1
x1 + x2 + + xm = 1
xi 0, i = 1, 2,..., m
v unrestricted
. v
: LP y1 , y2 ,..., yn B
n
n
min max a1 j y j , a2 j y j ,..., amj y j

yj
j =1
j =1
j =1
y1 + y2 + + yn = 1
y j 0,
j = 1, 2,..., n
: LP B
minimize w = v
Subjet to
n
v aij y j 0, i = 1, 2,..., m
j =1
y1 + y2 + + yn = 1
y j 0,
j = 1, 2,..., n
v unrestricted
177
) v ( .

maximize : v
st :
v 2 x1 4 x2 0
v 2 x1 3x2 0
v 3 x1 2 x2 0
v + x1 6 x2 0
x1 + x2 = 1
x1 0, x1 0, v unrestricted
Excel Solver
178
179

LP .B

:
r min
3
2
6
A LP
180
B3
B2
1 3
4 1
6 2
4 2
B1
3
2
A2
A3
5
c max
3
A1
maximize v
st
v 3x1 + 2 x2 + 5 x3 0
v x1 4 x2 + 6 x3 0
v + 3x1 + x2 2 x3 0
x1 + x2 + x3 = 1
x1 , x2 , x3 0
v unrestricted
LP B
minimize v
st
v 3 y1 + y2 + 3 y3 0
v + 2 y1 4 y2 + y3 0
v + 5 y1 + 6 y2 2 y3 0
y1 + y2 + y3 = 1
y1 , y2 , y3 0
v unrestricted
Excel Solver
:A
181
:B
Solver B
182

A B
.
B
. : A
B
produce groceries ba ker y
6
4
meat
2
produce 2
groceries 2
meat

A x1 x2
x3 . A :
max v
st
v 2 x1 + 2 x2 2 x3 0
7 x3 0
v 2 x1
v + 8 x1 6 x2 x3 0
v 6 x1 + 4 x2 + 3 x3 0
x1 + x2 + x3 = 1
x1 , x2 , x3 0, v unrestrected
183
Solver
:Excel
184
185
186
:Case Study
A ) B Single
(Simultaneous Move Game

:
Payoff from A to B
Bs Move
b
6
8
4
a
4
5
3
a
b
c
As Move
B ) (a ) (b A ) (a ) (b
) (c A ) (b B ) (a A B 5 .
A ) (c B ) (a B 3 .A
The Minimax Strategy
. A
) (b 8 B ) (a 5
.
Random Strategy
.
:
BMi = probability B makes move i, i = a or b,
187
AMi = probability A makes move i, i = a, b, or c.

B BMi B:
A ) (a Expected Loss:
4 BMa 6 BMb.
A ) (b Expected Loss:
5 BMa + 8 BMb.
A ) (c ExpecteL loss:
3 BMa 4 BMb.
.A
B BMi
Minimize the Maximum Expected Loss Minimax
Strategy : B BMi
A B . LB
B
) :(LINGO
;MIN = LB
;! Probabilities must sum to 1
;BMa + BMb = 1
;)! Expected loss if A chooses (a
;-LB + 4 * BMa - 6 * BMb <= 0
;)! Expected loss if A chooses (b
;-LB - 5 * BMa + 8 * BMb <= 0
;)! Expected loss if A chooses (c
;-LB + 3 * BMa - 4 * BMb <= 0
:
188
Global optimal solution found.

Objective value:
Infeasibilities:
Total solver iterations:
0.2000000
0.000000
2
Variable
LB
BMA
BMB
Row
1
2
3
4
5
Value
0.2000000
0.6000000
0.4000000
Reduced Cost
0.000000
0.000000
0.000000
Slack or Surplus
0.2000000
0.000000
0.2000000
0.000000
0.000000
Dual Price
-1.000000
-0.2000000
0.000000
0.3500000
0.6500000
( b) 0.6 ( a) B :
0.2 B 0.4
A .A
PA Expected Profit Minimum Maximizing
: A
MAX = PA;
! Probabilities sum to 1;
AMa + AMb + AMc = 1;
! Expected profit if B chooses (a);
-PA + 4 * AMa - 5 * AMb + 3 * AMc >= 0;
! Expected profit if B chooses (b);
-PA - 6 * AMa + 8 * AMb - 4 * AMc >= 0;
: A
Objective value:
Infeasibilities:
0.2000000
0.000000
3
Variable
PA
AMA
AMB
Value
0.2000000
0.000000
0.3500000
189
Reduced Cost
0.000000
0.2000000
0.000000
0.000000
Dual Price
1.000000
0.2000000
-0.6000000
-0.4000000
AMC
0.6500000
Slack or Surplus
0.2000000
0.000000
0.000000
0.000000
Row
1
2
3
4
: A ) (b 0.35 ) (c 0.65
) (a .0.2
A .B B
A .0.2 Expected Transfer 0.2

) ( 0.2
B .A Biased A 0.2 .

.Mixed Strategy
: . LP and Dual LP
190
Equilibrium
.
Equilibrium Strategy "" "" .
Equilibrium in a game .
Rational Play .
) (
) Simultaneous Moves Static-move ( Hidden-move

.
Pay-off Matrix .Strategic Forms
Sequential Moves Dynamic Games
. Extensive Forms
.Game Trees
) Single-Stage ( ) Repeated Games (
191

Dominant Strategy Equilibrium
.
:
.

Nash Equilibrium
:
.
.
Nash Strategies
A B
:
) p ( Ai , Bi A Ai B . Bi
) p ( Ai , Bi A Ai B . Bi
p ( Ai , B i ) - A Ai B . Bi
192
Strictly Dominant Strategy

Ai A A i Bi:
) p ( Ai , Bi ) > p ( A i , Bi
) p ( Ai , Bi ) > p ( A i , B i
B A
:
) p ( Bi , Ai B Bi A . Ai
) p ( Bi , Ai B Bi A . Ai
) p ( Bi , A i B Bi A . Ai ) (...
Strictly Dominant Strategy
Bi B Bi A i:
) p ( Bi , Ai ) > p ( B i , Ai
) p ( Bi , Ai ) > p ( B i , A i
193

Ai Bi :
) p ( Ai , Bi ) > p ( A i , Bi
) p ( Bi , Ai ) > p ( B i , Ai
) = >(
Weak Nash Equilibrium Strong Nash
.Equilibrium
B
raise price lower price
A raise price
5,5
1, 2
lower price
2,1
3,3
:
Bimatrix .
}{raise price, raise price }price, lower price
{lower
194
} {raise price, raise price

} {lower price, lower price
.
195
Gambit
Gambit
Extensive Form
.Strategy Form OpenSource
http://www.gambit-project.org/

view Strategic game
196
197
. A
.B
198
199
Compute Nash equlibria
200
. . .OK
201
202
203
:
.
204
Gambit

205
206
207
208

Create a new strategic game
209
210
: Prisoners Dilemma

. .
.
. :
.
:
prisoner 2
deny confess
prisoner1 deny
1, 1 10,0
confess 0, 10 5, 5
Gambit
211
: .
212
Extensive Forms
213
) ( Insert move
214
215

)
( . 1
3

) 2 ( .

.
:
216
Iterated-dominance Equilibrium

.

217
218
219
220
221
222
223
224
225
226
Weak Iteration
)(
227
228
:
:
(1
(2
(3
229
230
Pareto Efficiency

.
Pareto Domination
) ( 1 ) ( 2
) .(1
: pure conflict

. :
231
232
1-3
233
234
235
Gambit
236
237
Gambit
: . A
A1 A2 . A3 B
B1 B2 B3
. B4
. :A
B4
B3
B2
B1
8 2 9 3
6 5 6 8
2
4
9
5
238
A1
A2
A3
239
240
241
242

A B
H .T . )
HH (TT A B B.
: A
BT
1
1
Gambit
243
BH
1
1
AH
AT
: .
244

2 x 4 A:
B4
1
6
.
245
B3
B2
B1
2
4
A1
A2
246
: .
247

:Gambit
B3
B2
B1
3 1 3
2 4 1
5 6 2
:
248
A1
A2
A3
249
: .
250
:Case Studies
(1
.

. .
) ( :

.
:
251
252
253
254
255
: ) (8- 2
" " 52
256
257
258
:
:Excel Solver
259
260
:Gambit
261
262
263
264
265
266
267
268
269
: .
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
: .
398
:Sage Normal Form

Strategic Form
Sage
:

B
1
A
1 3, 2 1,1
2 0, 0 2,3
or
3 1
2 1
A=
, B =
0 2
0 3
sage: A = matrix([[3, 1], [0, 2]])

sage: B = matrix([[2, 1], [0, 3]])
sage: b_of_s = NormalFormGame([A, B])
sage: b_of_s
sage: b_of_s.obtain_nash(algorithm=enumeration)
[[(0, 1), (0, 1)],[(3/4, 1/4),(1/4, 3/4)],[(1, 0), (1, 0)]]
A ( s1 , s2 )
s1 A s2
s1 B s2
sage: for ne in
b_of_s.obtain_nash(algorithm=enumeration):
....:
print "Utility for {}: ".format(ne)
....:
print vector(ne[0]) * A * vector(ne[1]),

vector(ne[0]) * B * vector(ne[1])
Utility for [(0, 1), (0, 1)]:
399
2 3
Utility for [(3/4, 1/4), (1/4, 3/4)]:
3/2 3/2
Utility for [(1, 0), (1, 0)]:
3 2
.Gambit :
:
B
1
A
1 1, 1 1,1
2 1,1 1, 1
or
1 1
1 1
A=
, B =
1 1
1 1
sage:
sage:
sage:
sage:
A = matrix([[1, -1], [-1, 1]])

B = matrix([[-1, 1], [1, -1]])
m_p = NormalFormGame([A, B])
m_p.obtain_nash(algorithm=enumeration)
[[(1/2, 1/2), (1/2, 1/2)]]
sage: [vector([1/2, 1/2]) * M * vector([1/2, 1/2])

....:
for M in m_p.payoff_matrices()]
[0, 0]
sage: m_p.payoff_matrices()
(
[ 1 -1] [-1 1]
[-1 1], [ 1 -1])
:
:
sage: A = matrix([[0, -1, 1, 1, -1],
....:
[1, 0, -1, -1, 1],
....:
[-1, 1, 0, 1 , -1],
400
....:
[-1, 1, -1, 0, 1],
....:
[1, -1, 1, -1, 0]])
sage: g = NormalFormGame([A])
sage: g.obtain_nash(algorithm=enumeration)
[[(1/5, 1/5, 1/5, 1/5, 1/5), (1/5, 1/5, 1/5, 1/5, 1/5)]]
sage: g.payoff_matrices()
:
sage: A = matrix([[10, 500, 44],
....:
[15, 10, 105],
....:
[19, 204, 55],
....:
[20, 200, 590]])
sage: B = matrix([[2, 1, 2],

....:
[0, 5, 6],
....:
[3, 4, 1],
....:
[4, 1, 20]])
sage: g=NormalFormGame([A, B])

sage: g.obtain_nash(algorithm=lrs) # optional - lrslib
[[(0, 0, 0, 1), (0, 0, 1)]]
sage: g.obtain_nash(algorithm=lrs, maximization=False) #
optional - lrslib
[[(2/3, 1/12, 1/4, 0), (6333/8045, 247/8045, 293/1609)],
[(3/4, 0, 1/4, 0), (0, 11/307, 296/
sage: A = matrix([[3,3],
....:
[2,5],
....:
[0,6]])
sage: B = matrix([[3,2],
....:
[2,6],
401
....:
[3,1]])
sage: g = NormalFormGame([A, B])

[[(0, 1/3, 2/3), (1/3, 2/3)], [(4/5, 1/5, 0), (2/3,
1/3)], [(1, 0, 0), (1, 0)]]
sage: A = matrix([[160, 205, 44],

....:
[175, 180, 45],
....:
[201, 204, 50],
....:
[120, 207, 49]])
sage: B = matrix([[2, 2, 2],

....:
[1, 0, 0],
....:
[3, 4, 1],
....:
[4, 1, 2]])
sage: g=NormalFormGame([A, B])

[[(0, 0, 3/4, 1/4), (1/28, 27/28, 0)]]
sage: g.obtain_nash(algorithm=lrs) # optional lrslib
[[(0, 0, 3/4, 1/4), (1/28, 27/28, 0)]]
sage: g.obtain_nash(algorithm=LCP) # optional gambit
[[(0.0, 0.0, 0.75, 0.25), (0.0357142857, 0.9642857143,
0.0)]]
sage: player1 = matrix([[2, 8, -1, 1, 0],
402
....:
[1, 1, 2, 1, 80],
....:
[0, 2, 15, 0, -12],
....:
[-2, -2, 1, -20, -1],
....:
[1, -2, -1, -2, 1]])
sage: player2 = matrix([[0, 8, 4, 2, -1],

....:
[6, 14, -5, 1, 0],
....:
[0, -2, -1, 8, -1],
....:
[1, -1, 3, -3, 2],
....:
[8, -4, 1, 1, -17]])
sage: fivegame = NormalFormGame([player1, player2])

sage: fivegame.obtain_nash(algorithm=enumeration)
[[(1, 0, 0, 0, 0), (0, 1, 0, 0, 0)]]
sage: fivegame.obtain_nash(algorithm=lrs) #
optional - lrslib
[[(1, 0, 0, 0, 0), (0, 1, 0, 0, 0)]]
sage: fivegame.obtain_nash(algorithm=LCP) #
optional - gambit
[[(1.0, 0.0, 0.0, 0.0, 0.0), (0.0, 1.0, 0.0, 0.0,
0.0)]]
Here is an example of a 3 by 2 game with 3 Nash
equilibrium:
sage: A = matrix([[3,3],
....:
[2,5],
....:
[0,6]])
sage: B = matrix([[3,2],
....:
[2,6],
....:
[3,1]])
403
sage: g = NormalFormGame([A, B])

[[(0, 1/3, 2/3), (1/3, 2/3)], [(4/5, 1/5, 0), (2/3, 1/3)],
[(1, 0, 0), (1, 0)]]
sage: p1 = matrix([[1, 2], [3, 4]])

sage: p2 = matrix([[3, 3], [1, 4]])
sage: g = NormalFormGame([p1, p2])
sage: g.payoff_matrices()
(
[1 2] [3 3]
[3 4], [1 4]
)
:Sage
.
1)
sage: g =
game_theory.normal_form_games.PrisonersDilemma()
sage: g.obtain_nash()
[[(0, 1), (0, 1)]]
sage: d = {(0, 0): [-2, -2], (0, 1): [-5, 0], (1,
0): [0, -5], (1, 1): [-4, -4]}
sage: g == d
True
404
2)
sage: g =
game_theory.normal_form_games.AntiCoordinationGame()
sage: g
Anti coordination game - Normal Form Game with the
following utilities: ...
sage: d ={(0, 1): [1, 5], (1, 0): [5, 1],
....: (0, 0): [3, 3], (1, 1): [0, 0]}
sage: g == d
True
[[(0, 1), (1, 0)], [(1/3, 2/3), (1/3, 2/3)], [(1,
0), (0, 1)]]
3)
sage: g =
game_theory.normal_form_games.BattleOfTheSexes()
sage: g
Battle of the sexes - Coordination game Normal Form Game with the following utilities: ...
sage: d = {(0, 1): [1, 1], (1, 0): [0, 0], (0, 0):
[3, 2], (1, 1): [2, 3]}
sage: g == d
True
[[(0, 1), (0, 1)], [(3/4, 1/4), (1/4, 3/4)], [(1,
0), (1, 0)]]
405
4)
sage: g = game_theory.normal_form_games.Chicken()
sage: g
Chicken - Anti coordination game Normal Form Game with the following utilities: ...
sage: d = {(0, 1): [-1, 1], (1, 0): [1, -1],
....: (0, 0): [0, 0], (1, 1): [-10, -10]}
sage: g == d
True
[[(0, 1), (1, 0)], [(9/10, 1/10), (9/10, 1/10)],
[(1, 0), (0, 1)]]
5)
sage: g =
game_theory.normal_form_games.CoordinationGame()
sage: g
Coordination game - Normal Form Game with the
sage: d = {(0, 1): [0, 0], (1, 0): [0, 0],
....: (0, 0): [10, 5], (1, 1): [5, 10]}
sage: g == d
True
[[(0, 1), (0, 1)], [(2/3, 1/3), (1/3, 2/3)], [(1,
0), (1, 0)]]
406
6)
sage: g = game_theory.normal_form_games.HawkDove()
sage: g
Hawk-Dove - Anti coordination game Normal Form Game with the following utilities: ...
sage: d ={(0, 1): [2, 0], (1, 0): [0, 2],
....: (0, 0): [-2, -2], (1, 1): [1, 1]}
sage: g == d
True
[[(0, 1), (1, 0)], [(1/3, 2/3), (1/3, 2/3)], [(1,
0), (0, 1)]]
7)
sage: g =
game_theory.normal_form_games.MatchingPennies()
sage: g
Matching pennies - Normal Form Game with the
sage: d ={(0, 1): [-1, 1], (1, 0): [-1, 1],
....: (0, 0): [1, -1], (1, 1): [1, -1]}
sage: g == d
True
[[(1/2, 1/2), (1/2, 1/2)]]
407
8)
sage: g =
game_theory.normal_form_games.PrisonersDilemma()
sage: g
Prisoners dilemma - Normal Form Game with the
sage:
d = {(0, 0): [-2, -2], (0, 1): [-5, 0], (1, 0): [0, -5],
....:
(1, 1): [-4, -4]}
sage: g == d
True
[[(0, 1), (0, 1)]]
9)
sage: g = game_theory.normal_form_games.RPS()
sage: g
Rock-Paper-Scissors - Normal Form Game with the
sage:
d = {(0, 1): [-1, 1], (1, 2): [-1, 1], (0, 0): [0, 0],
....:
(2, 1): [1, -1], (1, 1): [0, 0], (2, 0): [-1, 1],
....:
(2, 2): [0, 0], (1, 0): [1, -1], (0, 2): [1, -1]}
sage: g == d
True
[[(1/3, 1/3, 1/3), (1/3, 1/3, 1/3)]]
408
10)
sage: g = game_theory.normal_form_games.RPSLS()
sage: g
Rock-Paper-Scissors-Lizard-Spock Normal Form Game with the following utilities: ...
sage:
d = {(1, 3): [-1, 1], (3, 0): [-1, 1], (2, 1): [1, -1],
....:
(0, 3): [1, -1], (4, 0): [1, -1], (1, 2): [-1, 1],
....:
(3, 3): [0, 0], (4, 4): [0, 0], (2, 2): [0, 0],
....:
(4, 1): [-1, 1], (1, 1): [0, 0], (3, 2): [-1, 1],
....:
(0, 0): [0, 0], (0, 4): [-1, 1], (1, 4): [1, -1],
....:
(2, 3): [1, -1], (4, 2): [1, -1], (1, 0): [1, -1],
....:
(0, 1): [-1, 1], (3, 1): [1, -1], (2, 4): [-1, 1],
....:
(2, 0): [-1, 1], (4, 3): [-1, 1], (3, 4): [1, -1],
....:
(0, 2): [1, -1]}
sage: g == d
True
[[(1/5, 1/5, 1/5, 1/5, 1/5), (1/5, 1/5, 1/5, 1/5, 1/5)]]
11)
sage: g = game_theory.normal_form_games.StagHunt()
sage: g
Stag hunt - Coordination game Normal Form Game with the following utilities: ...
sage: d =
....:
{(0, 1): [0, 4], (1, 0): [4, 0],
(0, 0): [5, 5], (1, 1): [2, 2]}
sage: g == d
True
409
[[(0, 1), (0, 1)], [(2/3, 1/3), (2/3, 1/3)], [(1, 0), (1, 0)]]
12)
sage: g =
game_theory.normal_form_games.TravellersDilemma()
sage: g
Travellers dilemma - Normal Form Game with the
sage: d = {(7, 3): [5, 1], (4, 7): [1, 5], (1, 3): [5, 9],
....: (4, 8): [0, 4], (3, 0): [9, 5], (2, 8): [0, 4],
....: (8, 0): [4, 0], (7, 8): [0, 4], (5, 4): [7, 3],
....: (0, 7): [1, 5], (5, 6): [2, 6], (2, 6): [2, 6],
....: (1, 6): [2, 6], (5, 1): [7, 3], (3, 7): [1, 5],
....: (0, 3): [5, 9], (8, 5): [4, 0], (2, 5): [3, 7],
....: (5, 8): [0, 4], (4, 0): [8, 4], (1, 2): [6, 10],
....: (7, 4): [5, 1], (6, 4): [6, 2], (3, 3): [7, 7],
....: (2, 0): [10, 6], (8, 1): [4, 0], (7, 6): [5, 1],
....: (4, 4): [6, 6], (6, 3): [6, 2], (1, 5): [3, 7],
....: (8, 8): [2, 2], (7, 2): [5, 1], (3, 6): [2, 6],
....: (2, 2): [8, 8], (7, 7): [3, 3], (5, 7): [1, 5],
....: (5, 3): [7, 3], (4, 1): [8, 4], (1, 1): [9, 9],
....: (2, 7): [1, 5], (3, 2): [9, 5], (0, 0): [10, 10],
....: (6, 6): [4, 4], (5, 0): [7, 3], (7, 1): [5, 1],
....: (4, 5): [3, 7], (0, 4): [4, 8], (5, 5): [5, 5],
....: (1, 4): [4, 8], (6, 0): [6, 2], (7, 5): [5, 1],
....: (2, 3): [5, 9], (2, 1): [10, 6], (8, 7): [4, 0],
....: (6, 8): [0, 4], (4, 2): [8, 4], (1, 0): [11, 7],
....: (0, 8): [0, 4], (6, 5): [6, 2], (3, 5): [3, 7],
....: (0, 1): [7, 11], (8, 3): [4, 0], (7, 0): [5, 1],
410
....: (4, 6): [2, 6], (6, 7): [1, 5], (8, 6): [4, 0],
....: (5, 2): [7, 3], (6, 1): [6, 2], (3, 1): [9, 5],
....: (8, 2): [4, 0], (2, 4): [4, 8], (3, 8): [0, 4],
....: (0, 6): [2, 6], (1, 8): [0, 4], (6, 2): [6, 2],
....: (4, 3): [8, 4], (1, 7): [1, 5], (0, 5): [3, 7],
....: (3, 4): [4, 8], (0, 2): [6, 10], (8, 4): [4, 0]}
sage: g == d
True
sage: g.obtain_nash() # optional - lrs
[[(0, 0, 0, 0, 0, 0, 0, 0, 1), (0, 0, 0, 0, 0, 0, 0, 0, 1)]]
sage: g =
game_theory.normal_form_games.TravellersDilemma(5)
sage: g
Travellers dilemma - Normal Form Game with the
sage: d = {(0, 1): [2, 6], (1, 2): [1, 5], (3, 2): [4, 0],
....: (0, 0): [5, 5], (3, 3): [2, 2], (3, 0): [4, 0],
....: (3, 1): [4, 0], (2, 1): [5, 1], (0, 2): [1, 5],
....: (2, 0): [5, 1], (1, 3): [0, 4], (2, 3): [0, 4],
....: (2, 2): [3, 3], (1, 0): [6, 2], (0, 3): [0, 4],
....: (1, 1): [4, 4]}
sage: g == d
True
[[(0, 0, 0, 1), (0, 0, 0, 1)]]
411
SageMath

.
Coalitions
Games in Coalitional Form
- Many-Prson Games
-1 .
-2 .
-3 Transferrable Utility Side
Payments .
) (
-4 .
Coalitional Form. Characteristic
Functions
n2
1 n
}N = {1,2,...n
S N S N
) 2 N Empty
(Coalition N Grand Coalition n = 2
{
}
8 } { ,{1} ,{2} ,{3} ,{1,2} ,{1,3} ,{2,3} , N n
) ( 4 ,{1} ,{2} , N n = 3
2 N 2n .
412
:
n n-person ) ( N , v
} N = {1,2,...n v
Characterstic Function 2 N
:
v ( ) = 0 (1
(2 S T S T =
) ) v ( S ) + v (T ) v ( S T (Superadditivity
) v ( S S N
S .
) (1 ) (2
) (
.
:
Monotone v ( S ) v (T ) T S
:
Constant-Sum
) v ( S ) + v ( S ) = v ( N S 2 N
Zero-Sum . v ( N ) = 0
413
:
3 I II III 1 2

(1 I 1
III
1
)( 0,3,1) ( 2,1,1
)( 4,2,3) (1,0,0
II 1
2
(2 I 1
III
2
)(1,0,0) (1,1,1
)(0,0,1) (0,1,1

II 1
2
:I
I1: first row
)(I1,II1,III1): (0,3,1
)(I1,II1,III2): (2.1.1
)(I1,II2,III1): (4,2,3
)(I1,II2,III2): (1,0,0
I2: second row
)(I2,II1,III1): (1,0,0
)(I2,II1,III2): (1,1,1
)(I2,II2,III1): (0,0,1
)(I2,II2,III2): (0,1,1
414
v ( ) = 0
) v ( N ) (1,2,1 )( 4,2,3
v ( N ) = 9 ) } v ({1 I
):(II,III
) ( II , III
1,1 1,2 2,1 2,2
I 1 0 2 4 1
2 1 1 0 0

)]] sage: A = matrix([[0, 2, 4, 1 ],[1, 1, 0, 0
)]sage: gI = NormalFormGame([A
)(sage: M = gI.obtain_nash
sage: M
]])[[(1/2, 1/2), (1/2, 0, 0, 1/2
sage: for ne in gI.obtain_nash():
)]print vector(ne[0]) * A * vector(ne[1
...
1/2
Gambit
. v ({1} ) = 1 2
) }v ({2
415
II1: first row

(I1,II1,III1): (0,3,1)
(I1,II1,III2): (2.1.1)
(I2,II1,III1): (1,0,0)
(I2,II1,III2): (1,1,1)
II2: second row
(I1,II2,III1): (4,2,3)
(I1,II2,III2): (1,0,0)
(I2,II2,III1): (0,0,1)
(I2,II2,III2): (0,1,1)
( I , III )
1,1 1,2 2,1 2,2
II
sage:
sage:
sage:
sage:
[[(0,
sage:
...
0
0
A =
gII
M =
M
1),
for
matrix([[3, 1, 0, 1 ],[2, 0, 0, 1 ]])

= NormalFormGame([A])
gII.obtain_nash()
(0, 0, 1, 0)], [(1, 0), (0, 0, 1, 0)]]
ne in gII.obtain_nash():
print vector(ne[0]) * A * vector(ne[1])
Gambit
v ({2} ) = 0
416
v ({3} )
III1: first row

(I1,II1,III1): (0,3,1)
(I1,II2,III1): (4,2,3)
(I2,II1,III1): (1,0,0)
(I2,II2,III1): (0,0,1)
III2: second row
(I1,II1,III2): (2.1.1)
(I1,II2,III2): (1,0,0)
(I2,II1,III2): (1,1,1)
(I2,II2,III2): (0,1,1)
( I , II )
1,1 1,2 2,1 2,2
III
1 3
sage: A = matrix([[1, 3, 0, 1 ],[1, 0, 1, 1 ]])

sage: gIII = NormalFormGame([A])
sage: M = gIII.obtain_nash()
sage: M
[[(1/4, 3/4), (0, 1/4, 3/4, 0)]]
sage: for ne in gIII.obtain_nash():
...
3/4
417
Gambit
v ({3} ) = 3 4
: II III I v ( {1,3} )
I,III: II1 first column
(I1,II1,III1): (0,3,1) 1
(I1,II1,III2): (2,1,1) 3
(I2,II1,III1): (1,0,0) 1
(I2,II1,III2): (1,1,1) 2
I,III: II2 second column
(I1,II2,III1): (4,2,3) 7
(I1,II2,III2): (1,0,0) 1
(I2,II2,III1): (0,0,1) 1
(I2,II2,III2): (0,1,1) 1
II
1 2
I , III
1,1
1 7
1,2
3 1
2,1
2,2
1 1
2 1
sage: A = matrix([[1, 7], [3, 1 ],[1, 1], [2, 1 ]])

sage: gI_III = NormalFormGame([A])
sage: M = gI_III.obtain_nash()
sage: M
[[(1/4, 3/4, 0, 0), (3/4, 1/4)]]
418
sage: for ne in gI_III.obtain_nash():

...
5/2
v ({1,3} ) = 5 2
Gambit
I III II v ( {1,2} ) III II I
v ({2,3} )
I,II: III1 first column

(I1,II1,III1): (0,3,1) 3
(I1,II2,III1): (4,2,3) 6
(I2,II1,III1): (1,0,0) 1
(I2,II2,III1): (0,0,1) 0
I,II: III2 second column
(I1,II1,III2): (2,1,1) 3
(I1,II2,III2): (1,0,0) 1
(I2,II1,III2): (1,1,1) 2
(I2,II2,III2): (0,1,1) 1
III
1 2
I , II
1,1
3 3
1,2
6 1
2,1
2,2
1 2
0 1
419
sage: A = matrix([[3, 3], [6, 1 ],[1, 2], [0, 1 ]])

sage: gI_II_III = NormalFormGame([A])
sage: M = gI_II_III.obtain_nash()
sage: M
[[(1, 0, 0, 0), (0, 1)]]
sage: for ne in gI_II_III.obtain_nash():
...
v ( {1,2} ) = 3
II,III: I1 first column

(I1,II1,III1): (0,3,1) 4
(I1,II1,III2): (2,1,1) 2
(I1,II2,III1): (4,2,3) 5
(I1,II2,III2): (1,0,0) 0
II,III: I2 second column
(I2,II1,III1): (1,0,0) 0
(I2,II1,III2): (1,1,1) 2
(I2,II2,III1): (0,0,1) 1
(I2,II2,III2): (0,1,1) 2
420
I
1 2
II , III
1,1
4 0
1,2
2 2
2,1
2,2
5 1
0 2
sage: A = matrix([[4, 0], [2, 2 ],[5, 1], [0, 2 ]])

sage: gII_III_I = NormalFormGame([A])
sage: M = gII_III_I.obtain_nash()
sage: M
[[(0, 1, 0, 0), (0, 1)]]
sage: for ne in gII_III_I.obtain_nash():
...
v ({2,3} ) = 2
v( ) = 0
v({1}) = 0.5
v({2}) = 0
v({3}) = 0.75
v({1,2}) = 3
v({1,3}) = 2.5
v({2,3}) = 2
v({N}) = 9
421
:

)1
I ,1
III
1
)( 2,1,1) (1, 4,3

) (1,3, 4 ) (10, 5, 4
II
I ,2
III
1
)( 1, 2,3
) ( 4,2,2
) (12, 6, 6 ) ( 1,3, 2
II
2
)2
I ,1
III
1
)(1,2,1
)( 3,0,1
)( 1,6, 3) ( 3,2,1
1
2
I ,2
III
2
)( 1,2,4 ) (1,0,3
)( 7,5,4 ) ( 3,2,1
422
II
1
2
II
Imputation and the Core

.
Superadditivity ) v ( N
.
.
.
Imputation
x = x , x ,..., xn .
1 2
xi i .
) . v ( N
:
x = x , x ,..., xn Group Rational

1 2
n
Efficient ) xi = v ( N
i=1

) ( . ) } xi v ({i
.
:
x = x , x ,..., xn Individually
1 2
Rational ) } xi v ({i . i = 1,2,..., n
.
423
} { x = ( x1, x2,..., xn ) : iN xi = v ( N ) and xi v ({i}) i N

)
x = x , x ,..., xn ) }xi v ({i

1 2
n
i = 1,2,..., n ) . xi = v ( N
i=1
.
Essential Games

n
n
) v ({i} ) = v ( N ) . v ({i} ) < v ( N
i=1
i=1
The Core
:
x S
v ( S ) > iS xi S x
S .
:
C
424
( {
C = x = x , x ,..., xn : iN xi = v ( N ) and iS xi v ( S ) S N
1 2
: .
.
.
:

v ( {1,2} ) = 4
v ( {1,3} ) = 3 v ({1,2,3} ) = 8
v ({1} ) = 1
v ( ) = 0 v ({2} ) = 0
v ({3} ) = 1 v ({2,3} ) = 5
x , x , x
1 2 3
x 1, x 0, x 1
1
2
3
x + x + x = 8,
1 2 3
:
v ) v ( S S
) v ( S ) = f ( S . f
The Shapley Value
v ) ( v ) = ( v ) , ( v ) ,...,n ( v
1
2
) i ( v i .
:
) (Fairness ) : ( v
(1 iN i ( v ) = v ( N ) : Efficiency
425
(2 :Symmetry i j ) } v ( S {i} ) = v ( S { j
S i j ) i ( v ) = j ( v
(3 : Dummy Axiom i ) } v ( S ) = v ( S {i
S i i ( v ) = 0
(4 :Additivity u v
) (u + v ) = (u ) + ( v
) )((1
:
.
:
= , ,...,n
1 2
( S 1)!( n S )! v ( S ) v ( S {i} ) , i = 1,..., n
!n
= ) i ( v
SN
iS
S . i
) } v ( S ) v ( S {i S ) i
)} ( v(S-{i i .S
module
SageMath
) : (SageMath
Sage
426
v ( ) = 0
v ({1} ) = 1
v ({2} ) = 0
v ({3} ) = 1
v ({1,2} ) = 4
v ({1,3} ) = 3
v ({2,3} ) = 5
v ({1,2,3} ) = 8
:
https://sagecell.sagemath.org/
:
integer_function = { ():0, (1,):1, (2,):0, (3,):1,
(1,2,):4, (1,3,):3, (2,3,):5, (1,2,3,):8}
integer_game = CooperativeGame(integer_function)
integer_game
integer_game.shapley_value()
integer_game.is_monotone()
integer_game.is_superadditive()
427
8 7/3
17/6 .
:
i_fun = { ():0, (1,):0.5, (2,):0, (3,):0.75,
}(1,2,):3, (1,3,):2.5, (2,3,):2, (1,2,3,):9
)l_g = CooperativeGame(i_fun
l_g
428
l_g.shapley_value()
l_g.is_monotone()
l_g.is_superadditive()
429
430
:
sage: integer_function = {(): 0,
....:
(1,): 6,
....:
(2,): 12,
....:
(3,): 42,
....:
(1, 2,): 12,
....:
(1, 3,): 42,
....:
(2, 3,): 42,
....:
(1, 2, 3,): 42}
sage: integer_game =
CooperativeGame(integer_function)
sage: letter_function = {(): 0,
....:
(A,): 6,
....:
(B,): 12,
....:
(C,): 42,
....:
(A, B,): 12,
....:
(A, C,): 42,
....:
(B, C,): 42,
....:
(A, B, C,): 42}
sage: letter_game =
CooperativeGame(letter_function)
....:
(A,): 6,
....:
(B,): 12,
....:
(C,): 42,
431
....:
(A, B,): 12,
....:
(A, C,): 42,
....:
(B, C,): 42,
....:
(A, B, C,): 42}
sage: letter_game = ooperativeGame(letter_function)

sage: letter_game.is_monotone()
sage: letter_game.is_superadditive()
sage: letter_game
sage: letter_game.shapley_value()
sage: payoff_vector = letter_game.shapley_value()
sage: letter_game.is_efficient(payoff_vector)
sage: letter_game.nullplayer(payoff_vector)
sage: letter_game.is_symmetric(payoff_vector)
sage: payoff_vector = {A: 0, C: 35, B: 3}
sage: letter_game.is_efficient(payoff_vector)
sage: letter_game.nullplayer(payoff_vector)
sage: letter_game.is_symmetric(payoff_vector)

....:
(A,): 6,
....:
(B,): 12,
....:
(C,): 42,
432
....:
(A, B,): 12,
....:
(C, A,): 42,
....:
(B, C,): 42,
....:
(B, A, C,): 42}

sage: letter_game.shapley_value()
{A: 2, B: 5, C: 35}
sage: letter_game.is_monotone()
sage: letter_game.is_superadditive()
sage: letter_game.is_efficient({A: 2, C: 35,
B: 5})
sage: letter_game.nullplayer({A: 2, C: 35, B:
5})
sage: letter_game.is_symmetric({A: 2, C: 35,
B: 5})
sage: letter_game.is_efficient({A: 0, C: 35,
B: 3})
sage: letter_game.nullplayer({A: 0, C: 35, B:
3})
sage: letter_game.is_symmetric({A: 0, C: 35,
B: 3})
....:
(A,): 6,
....:
(B,): 12,
....:
(C,): 42,
433
....:
(A, B,): 12,
....:
(A, C,): 42,
....:
(B, C,): 42,
....:
(A, B, C,): 42}
sage: letter_game =
sage: letter_game.is_efficient({A: 14, B: 14,
C: 14})
....:
(A,): 6,
....:
(B,): 12,
....:
(C,): 42,
....:
(A, B,): 12,
....:
(A, C,): 42,
....:
(B, C,): 42,
....:
(A, B, C,): 42}

sage: letter_game.is_efficient({A: 10, B: 14,
C: 14})
sage: long_function = {(): 0,
....:
(1,): 0,
....:
(2,): 0,
....:
(3,): 0,
....:
(4,): 0,
....:
(1, 2): 0,
....:
(1, 3): 0,
434
....:
(1, 4): 0,
....:
(2, 3): 0,
....:
(2, 4): 0,
....:
(3, 4): 0,
....:
(1, 2, 3): 0,
....:
(1, 2, 4): 45,
....:
(1, 3, 4): 40,
....:
(2, 3, 4): 0,
....:
(1, 2, 3, 4): 65}
sage: long_game = CooperativeGame(long_function)

sage: long_game.is_efficient({1: 20, 2: 20, 3: 5,
4: 20})
....:
(1,): 6,
....:
(2,): 12,
....:
(3,): 42,
....:
(1, 2,): 12,
....:
(1, 3,): 42,
....:
(2, 3,): 42,
....:
(1, 2, 3,): 42}
sage: integer_game.is_monotone()
435
....:
(1,): 6,
....:
(2,): 12,
....:
(3,): 42,
....:
(1, 2,): 10,
....:
(1, 3,): 42,
....:
(2, 3,): 42,
....:
(1, 2, 3,): 42}
sage: integer_game.is_monotone()
....:
(1,): 0,
....:
(2,): 0,
....:
(3,): 0,
....:
(4,): 0,
....:
(1, 2): 0,
....:
(1, 3): 0,
....:
(1, 4): 0,
....:
(2, 3): 0,
....:
(2, 4): 0,
....:
(3, 4): 0,
....:
(1, 2, 3): 0,
....:
(1, 2, 4): 45,
....:
(1, 3, 4): 40,
....:
(2, 3, 4): 0,
....:
(1, 2, 3, 4): 65}
436

sage: long_game.is_monotone()

....:
(1,): 6,
....:
(2,): 12,
....:
(3,): 42,
....:
(1, 2,): 12,
....:
(1, 3,): 42,
....:
(2, 3,): 42,
....:
(1, 2, 3,): 42}
sage: integer_game.is_superadditive()
sage: A_function = {(): 0,
....:
(1,): 6,
....:
(2,): 12,
....:
(3,): 42,
....:
(1, 2,): 18,
....:
(1, 3,): 48,
....:
(2, 3,): 55,
....:
(1, 2, 3,): 80}
sage: A_game = CooperativeGame(A_function)

sage: A_game.is_superadditive()
437

....:
(1,): 0,
....:
(2,): 0,
....:
(3,): 0,
....:
(4,): 0,
....:
(1, 2): 0,
....:
(1, 3): 0,
....:
(1, 4): 0,
....:
(2, 3): 0,
....:
(2, 4): 0,
....:
(3, 4): 0,
....:
(1, 2, 3): 0,
....:
(1, 2, 4): 45,
....:
(1, 3, 4): 40,
....:
(2, 3, 4): 0,
....:
(1, 2, 3, 4): 65}

sage: long_game.is_superadditive()
....:
(1,): 0,
....:
(2,): 0,
....:
(3,): 55,
....:
(4,): 0,
....:
(1, 2): 0,
....:
(1, 3): 0,
....:
(1, 4): 0,
438
....:
(2, 3): 0,
....:
(2, 4): 0,
....:
(3, 4): 0,
....:
(1, 2, 3): 0,
....:
(1, 2, 4): 45,
....:
(1, 3, 4): 40,
....:
(2, 3, 4): 0,
....:
(1, 2, 3, 4): 85}

sage: long_game.is_superadditive()
....:
(A,): 6,
....:
(B,): 12,
....:
(C,): 42,
....:
(A, B,): 12,
....:
(A, C,): 42,
....:
(B, C,): 42,
....:
(A, B, C,): 42}
sage: letter_game =
sage: letter_game.is_symmetric({A: 5, B: 14,
C: 20})
....:
(1,): 12,
....:
(2,): 12,
439
....:
(3,): 42,
....:
(1, 2,): 12,
....:
(1, 3,): 42,
....:
(2, 3,): 42,
....:
(1, 2, 3,): 42}
sage: integer_game.is_symmetric({1: 2, 2: 5, 3:
35})
....:
(1,): 0,
....:
(2,): 0,
....:
(3,): 0,
....:
(4,): 0,
....:
(1, 2): 0,
....:
(1, 3): 0,
....:
(1, 4): 0,
....:
(2, 3): 0,
....:
(2, 4): 0,
....:
(3, 4): 0,
....:
(1, 2, 3): 0,
....:
(1, 2, 4): 45,
....:
(1, 3, 4): 40,
....:
(2, 3, 4): 0,
....:
(1, 2, 3, 4): 65}
440
sage: long_game.is_symmetric({1: 20, 2: 20, 3: 5,

4: 20})
....:
(A,): 0,
....:
(B,): 12,
....:
(C,): 42,
....:
(A, B,): 12,
....:
(A, C,): 42,
....:
(B, C,): 42,
....:
(A, B, C,): 42}
sage: letter_game =
sage: letter_game.nullplayer({A: 0, B: 14, C:
14})
....:
(1,): 0,
....:
(2,): 12,
....:
(3,): 42,
....:
(1, 2,): 12,
....:
(1, 3,): 42,
....:
(2, 3,): 55,
....:
(1, 2, 3,): 55}

sage: A_game.nullplayer({1: 10, 2: 10, 3: 25})
441

....:
(1,): 0,
....:
(2,): 0,
....:
(3,): 0,
....:
(4,): 0,
....:
(1, 2): 0,
....:
(1, 3): 0,
....:
(1, 4): 0,
....:
(2, 3): 0,
....:
(2, 4): 0,
....:
(3, 4): 0,
....:
(1, 2, 3): 0,
....:
(1, 2, 4): 45,
....:
(1, 3, 4): 40,
....:
(2, 3, 4): 0,
....:
(1, 2, 3, 4): 65}

sage: long_game.nullplayer({1: 20, 2: 20, 3: 5, 4:
20})
....:
(1,): 42,
....:
(2,): 12,
....:
(3,): 0,
....:
(1, 2,): 55,
....:
(1, 3,): 42,
....:
(2, 3,): 12,
442
....:
(1, 2, 3,): 55}

sage: A_game.nullplayer({1: 10, 2: 10, 3: 25})
....:
(1,): 6,
....:
(2,): 12,
....:
(3,): 42,
....:
(1, 2,): 12,
....:
(1, 3,): 42,
....:
(2, 3,): 42,
....:
(1, 2, 3,): 42}
sage: integer_game.player_list
sage: integer_game.shapley_value()
....:
(1,): 0,
....:
(2,): 0,
....:
(3,): 0,
....:
(4,): 0,
....:
(1, 2): 0,
....:
(1, 3): 0,
....:
(1, 4): 0,
....:
(2, 3): 0,
....:
(2, 4): 0,
443
....:
(3, 4): 0,
....:
(1, 2, 3): 0,
....:
(1, 2, 4): 45,
....:
(1, 3, 4): 40,
....:
(2, 3, 4): 0,
....:
(1, 2, 3, 4): 65}

sage: long_game.shapley_value()
444
:Matching Games
N Suitors N
Reviewers Rank Preferences
.
:
N S R N
S R : Preference List
g: R SN
and
f : S RN
} S = { J , K , L, M
}R = { A, B, C, D

if s = J ,
if s = K ,
if s = L,
if s = M ,
) ( A, D, C, B
) ( A, B, C, D
) ( B, D, C, A
) ( C, A, B, D
( s ) =
( L, J , K , M ) if s = A,
( J , M , L, K ) if s = B,
( K , M , L, J ) if s = C,
( M , K , J , L ) if = s = D.
445
g ( s ) =
Sage
sage: suitr_pref = {J: (A, D, C, B),
K: (A, B, C, D),
....:
L: (B, D, C, A),
....:
})M: (C, A, B, D
....:
sage: reviewr_pref = {A: (L, J, K, M),

B: (J, M, L, K),
....:
C: (K, M, L, J),
....:
})D: (M, K, J, L
....:
)]sage: m = MatchingGame([suitr_pref, reviewr_pref

sage: m
A matching game with 4 suitors and 4 reviewers
)(sage: m.suitors
)(K, J, M, L
)(sage: m.reviewers
)(A, C, B, D
:
M Bijection S . R s S r R
M . M ( s ) = r
Stable
.
:
Blocking Pairs
) ( s, r M ( s ) r s r ) M(r r s ).M-1(r
446
: Sage
sage: m.solve()
{J: A, K: C, L: D, M: B}
Bipartite Graphs
sage: plot(m)
Graphics object consisting of 13 graphics
primitives
:
sage: left_dict = {a: (A, B, C),
....:
b: (B, C, A),
....:
c: (B, A, C)}
sage: right_dict = {A: (b, c, a),

....:
B: (a, c, b),
....:
C: (a, b, c)}
447
sage: quick_game = MatchingGame([left_dict,

right_dict])
sage: quick_game.solve()
{a: A, b: C, c: B}
sage: quick_game.solve(invert=True)
{A: c, B: a, C: b}
sage: suitr_pref = {J: (A, D, C, B),

....:
K: (A, B, C, D),
....:
L: (B, D, C, A),
....:
M: (C, A, B, D)}
sage: reviewr_pref = {A: (L, J, K, M),

....:
B: (J, M, L, K),
....:
C: (K, M, L, J),
....:
D: (M, K, J, L)}
sage: m = MatchingGame([suitr_pref, reviewr_pref])

sage: m._suitors
[K, J, M, L]
sage: m._reviewers
[A, C, B, D]
sage: suit = {0: (3, 4),
....:
1: (3, 4)}
sage: revr = {3: (0, 1),

....:
4: (1, 0)}
sage: g = MatchingGame([suit, revr])
448
sage: g = MatchingGame(3)
sage: g
sage: for s in g.suitors():
....:
s, s.pref
(1, [])
(2, [])
(3, [])
sage: for r in g.reviewers():

....:
r, r.pref
(-1, [])
(-2, [])
(-3, [])
sage: g.solve()
sage: for s in g.suitors():
....:
s.pref = (-1, -2, -3)
sage: for r in g.reviewers():

....:
r.pref = (1, 2, 3)
sage: g.solve()
{1: -1, 2: -2, 3: -3}
449

Combinatorial Games two-
person .

.
. .
.
:
-1 Impartial Games
.
-2 Partizan Games
.
: :A Simple Take-Away Game
:
-1 A .B
-2 21.
-3
.
-4 A.
-5 )
(.
450
.
: Backward Induction

.
4
.
4 . 5 6
7 4 .
8 5 6 7
.
0 4 8 12 16 ...
Target Positions .
21 . 21 4
. 20
.
:
:
-1 .
451
-2 .
-3
.
.
-4 .
-5 .
-6 .
-7 .
-8
.
P :P-positions, N-positions N
0 4 8 12 16 ...
) Previous player ( 1 2 3
5 6 7 9 10 11 ... Next player
.
P N P
4
.
P N Labelling
) Terminal Position (.
P :N
:1 .P
452
:2 P .N
:3 N
.P
:4 P 3 . 2
P
P N P
P .
P N :
P N :
-1 .P
-2 N .P
-3 P .N
:Subtraction Games

.
S . S
:
n .
s s S .

}. S = {1, 2,3
453
} S = {1,3, 4
.P
.0 1 3 4 N
.0 2 P
2 1 N 5 6 N
.2 7 P
7 6 4 3 .N
8 10 11 N 9 P 12 13
N 14 P P
} P = {0, 2,7,9,14,16,...
0 2 .7 N
} N = {1,3, 4,5,6,8,10,11,12,13,15,... :
9 10 11 12 13 14...
position P
P ...
P N P N N N N 7 .
100
P 0 2 .(Modulus 7) 7
100 2 7 100 P
.
:
454
-1 Misere Version
. .
-2 }. S = {1, 2,3, 4,5,6
- .
- P
- 31
-3 m n ) ( m, n
m > 0 . n > 0 .

. ) . (1,1
. .P
-4 P :
)( }. S = {1,3,5,7
)( }. S = {1,3,6
)( 100
:The Game of Nim

:
-1 3 x1 x2 x3 )
5 7 9 (.
-2 .
455
-3 .

.
-4 .
:
http://www.chlond.demon.co.uk/Nim.html
http://www.dotsphinx.com/nim/
:
) ( 0,0,0 ) . P
: ( .
) ( 0,0, x x > 0
. N P
) ( 0,1,1 ) ( 0, 2, 2

.
.
) (1,1,1 ) (1,1, 2 ) (1,1,3 ) (1, 2, 2 N
) (1,1,0 ) ( 0, 2, 2 ) (1, 2,3
P N .
P ) (1, 4,5 ) ( 2, 4,6
) ( 5,7,9 P ) (15, 23,30 P
456

.
:Nim Sum
without carry
.2
: x 2
x = xm 2m + xm1 2m1 + + x1 21 + x0
.m xi .1
( xm xm1...x1 x0 )2 x .
22 = 1 16 + 0 8 + 1 4 + 1 2 + 0 1 = (10110 )2
2
2 :
: ( xm x0 )2 ( ym y0 )2 ( zm z0 )2
( xm x0 )2 ( ym y0 )2 = ( zm z0 )2
k
) zk = xk + yk ( mod 2
1, xk + yk = 1
zk =
0, otherwise
:
(10110 )2 (110011)2 = (100101)2

22 51 = 37
457
22 = 0 1 0 1 1 02
51 = 1 1 0 0 1 12
nim sum = 1 0 0 1 0 12 = 37
:
2
0 0 1 1
0 1 0 1
sum 0 1 1 0
) ( x ( xor ) y x y
:
) ( x1 , x2 , x3 P
. x1 x2 x3 = 0
) ( x1 , x2 , x3 ) = (13,12,8 P

13 12 : 8
13 = 1 1 0 12
12 = 1 1 0 02
1 0 0 02
= 9
nim sum = 1 0 0 12
N .
458
=8
P
1 . 9 13 4
:
1 0 02
=4
12 = 1 1 0 02
1 0 0 02
= 0
=8
nim sum = 0 0 0 02
7 12 5 .

:
3
4 .

) ( x1 , x2 , x3 , x4 P . x1 x2 x3 x4 = 0
)
(.
: N )
(1 ) .(1
.
:Gambit
459
. 5 .
.
.
:
-1 27 17
-2 38 x 25 x
-3 :
( 3 12 19 .27
460
( 4 13 17 19 .23
( )( )(
: Misere Version
.
-4 :
http://www.chlond.demon.co.uk/Coins.html
http://www.chlond.demon.co.uk/Northcott.html
http://www.math.ucla.edu/~tom/Games/Moore.htm
461
:Utility Theory
) (
a1 a2 .
"
" .

.

.
Preferences Consistent .
""
Utility

.
P P1 , P2 ,... . P
:
Preference relation P P
)( :
) -1 (Linearity P1 P2 P P1 P2 ) P2 P1
(.
462
) -2 (Transitivity P1 P2 P3 P P1 P2 P2 P3
. P1 P3 P1 P2 P2 P1 P1 P2
Equivalent . P1 P2
P
. P1 P2 P1 P2
P2 P1 . P1 P2 P1 P2
) Indifferent ( P1 P2 P1 P2 P2
P1 .
P2 P1
P2 . P1
)( Lottery .
:
Lottery . P
*. P
P1 P2 P3 p P1 1/2
P2 1/4 P3 1/4 . p1 p2
p3 *. P
p1 p2 0 1 p1 + (1 ) p2
.
P
* P.
463
* P Utility
.Function
:
Utility Function . P . u : P
) u ( P Domain u * P
) u ( p P* p
* p P P1 , P2 ,..., Pk 1 , 2 ,..., k
i 0 = 1
)(1
) u ( p ) = i u ( Pi
i =1
.p u
* P :
p1 p2

) u ( p1 ) u ( p2
)( 2
) p1 p2 u ( p1 ) u ( p2
.
* P u P ) (2
:
:A1 p1 p2 q * P 0 1
)( 3
p1 p2 p1 + (1 ) q p2 + (1 ) q
464
:A2 p1 p2 q *P
)( 4
p1 p2 > 0 : p1 q + (1 ) p2
)( 5
p1 p2 > 0 : q + (1 ) p1 p2
A1 :
q p1 . p2 p2
. p1 p2
.
A2 Continuity Axiom ) (4
p1 q + (1 ) p2 = 0
. A2
.
:
* P A1 A2
u P ) .(2 u unique
location .scale
) u ( P ) (2 a b > 0
) u ( P ) = a + b u ( P
) .(2 u
.
* P
1 2 P
465
.

.
) : (Utility Theory
:
(1) u ( p1 ) u ( p2 ) p1 p2
) (2) u p1 + (1 ) p2 = u ( p1 ) + (1 ) u ( p2
) u ( P ) = a + b u ( P ) (1 ).(2
: Prisoners Dilemma

. .
.
. :
.
:
Payoffs
prisoner 2
deny confess
prisoner1 deny
1, 1 10,0
confess 0, 10 5, 5
)
(:
Payoffs: -10, -5, -1, 0
466
10 ) 0 : (Satisfaction
5 )(5 :
1 )(10 :
0 )(15 :
:
Payoff Utility (Satisfaction) Utility Function
0
5
10
15
0
0.33
0.66
1
- 10
-5
-1
0
Utility
prisoner 2
deny confess
prisoner1 deny 10,10
0,15
confess 15,0
5,5
:
Utility Function
prisoner 2
deny
confess
prisoner1 deny 0.66,0.66
0,1
confess
1,0
0.33,0.33
:
467
Utility Function
1
0
10
Prison Time
:Gambit
:
468
Satisfaction
0.5
469
470
471
472
473
474
475
:
. A
A1 A2 . A3 B
B1 B2 B3
. B4
. :A
476
Payoff
A1
A2
A3
B1
B2
B3
B4
8 2 9 3
6 5 6 8
2 4 9 5
:
Payoffs Scale
-9
-3
-2
4
5
6
8
9
Utility
A1
A2
A3
B1
B2
0
1
2
3
4
5
6
7
B3
Utility
0
0.142857
0.285714
0.428571
0.571429
0.714286
0.857143
1
B4
1
0.143
0.857 0.286
0.714 0.571 0.714 0.857
0.286 0.429
0
0.571
477
utility
1
0.8
0.6
0.4
0.2
0
9
:
-1 .
-2 .Gambit
-3 .SageMath
:
.
478
-3
-6
-9
:
:
Excel Solver .Gambit .SageMath(1

.
:A
B4
B3
B2
B1
)(a
A1
A2
A3
B2
B1
B4
B3
4 4 5
3 4 9 2
A2
3 9
A4
(2
. :A
479
A1
A3
8 9
5
)(b
B4
B3
B2
B1
)(a
A1
A2
5 2 10
A3
A4
B3
B2
B1
)(a
A1
10
A2
A3
A4
4 2 5
B4
B3
B2
B1
)(c
A1
A2
A3
B4
B3
B2
B1
) (d
A1
A2
A3
(3
. %50
.
.
.
480
. %50
%30 %20
.

.
(4
.

300 %50
2
%30 ) 300(.
.
(5
A :
B2
B1
B3
50
50
A1
0.1
A2
10
10
A3
5
1
A , 0, B
6
6
49 5
, .
, 0
54 54
(6
.
481
B1
B2
B3
B4
A1
3 2
A2
3 3 0
A3
2 2 2
A4
1 2
1 2
4 1
(7
482
(8
483
(9
484
(10
(11
485
(12
(13
486
(14
(15
487
(16
(17
488
(18
(19
489
(20
(21
490
(22
(23
491
(24
(25
492
(26
493
(27
494
(28
495
(29
496
(30
497
(31
498

Applied Management Science, By J.A. Lawrence and B.A.
Pasternack, 2nd ed. Wiley
Operations Research, an introduction, By Hamdy Taha, 8th
ed. Prentice-Hall
A Guide to Game Theory,By Fiona Carmichael, 1st. ed.
Prentice-Hall
Game Theory, Thomas S. Ferguson, Publisher: UCLA 2008
Spreadsheet Modeling & Decision Analysis,Cliff T.
Ragsdale,5e. Thomson, South-Western
Treeplan Manual, http://www.treeplan.com.
Gambit: Software Tools for Game Theory, online Manual,
http://www.gambit-project.org/doc/index.html.
Sage Reference Manual: Game Throry Release 6.8, Jan
29,2015.
. . .
. . .
499
:LINGO :
For lingo
maximize : v
st :
v 2 x1 4 x2 0
v 2 x1 3x2 0
v 3 x1 2 x2 0
v + x1 6 x2 0
x1 + x2 = 1
x1 0, x1 0, v unrestricted
MIN = v;
-v - 2*x1 - 4*x2 <= 0;
-v - 2*x1 -3*x2 <=0;
-v - 3*x1 -2*x2 <= 0;
-v + x1 - 6*x2 <= 0;
x1 + x2 =1;
@FREE(v);
Objective value:
Infeasibilities:
-2.500000
0.000000
3
Variable
V
X1
X2
Row
1
2
3
4
5
6
Value
-2.500000
0.5000000
0.5000000
Slack or Surplus
-2.500000
0.5000000
0.000000
0.000000
0.000000
0.000000
500
Reduced Cost
0.000000
0.000000
0.000000
Dual Price
-1.000000
0.000000
0.000000
0.8750000
0.1250000
2.500000
maximize v
st
v 3x1 + 2 x2 + 5 x3 0
v x1 4 x2 + 6 x3 0
v + 3x1 + x2 2 x3 0
x1 + x2 + x3 = 1
x1 , x2 , x3 0
v unrestricted
MIN = v;
-v - 3*x1
-v - x1 -v + 3*x1
x1 + x2 +
@FREE(v);
+ 2*x2 + 5*x3 <= 0;

4*x2 + 6*x3 <=0;
+ x2 - 2*x3 <= 0;
x3 =1;

Objective value:
Infeasibilities:
0.8350515
0.000000
4
Variable
V
X1
X2
X3
Row
1
2
3
4
5
Value
0.8350515
0.4432990
0.2061856
0.3505155
Slack or Surplus
0.8350515
0.000000
0.000000
0.000000
0.000000
501
Reduced Cost
0.000000
0.000000
0.000000
0.000000
Dual Price
-1.000000
0.2989691
0.9278351E-01
0.6082474
-0.8350515
minimize v
st
v 3 y1 + y2 + 3 y3 0
v + 2 y1 4 y2 + y3 0
v + 5 y1 + 6 y2 2 y3 0
y1 + y2 + y3 = 1
y1 , y2 , y3 0
v unrestricted
max v
st
v 2 x1 + 2 x2 2 x3 0
v 2 x1
7 x3 0
v + 8 x1 6 x2 x3 0
v 6 x1 + 4 x2 + 3 x3 0
x1 + x2 + x3 = 1
x1 , x2 , x3 0, v unrestrected
MIN = v;
-v - 2*x1
-v - 2*x1
-v + 8*x1
-v - 6*x1
x1 + x2 +
@FREE(v);
+ 2*x2
- 0*x2
- 6*x2
+ 4*x2
x3 =1;
2*x3 <= 0;
7*x3 <=0;
x3 <= 0;
3*x3 <= 0;

Objective value:
Infeasibilities:
0.4440892E-15
0.000000
4
Variable
V
X1
X2
Value
0.000000
0.3888889
0.5000000
502
Reduced Cost
0.000000
0.000000
0.000000
X3
Row
1
2
3
4
5
6
0.1111111
Slack or Surplus
0.000000
0.000000
1.555556
0.000000
0.000000
0.000000
503
0.000000
Dual Price
-1.000000
0.3333333
0.000000
0.3333333
0.3333333
0.000000
: SageMath :
open-source
Python R .
Matlab Mathematica Maple.
http://www.sagemath.org/
) (Linux Virtual
Box SageMathCloud
https://cloud.sagemath.com SageMathCell
https://sagecell.sagemath.org
http://sagemath.wikispaces.com
SageMath

.
.
Notebook Command Line
.
):
(sage:
SageMath : Python
http://www.abarry.ws/Statistics with Python by Example.pdf
504
:
sage:
sage:
5
sage:
True
sage:
False
sage:
True
sage:
True
a = 5
a
2 == 2
2 == 3
2 < 3
a == 5
sage: 2**3
# ** means exponent
8
sage: 2^3
# ^ is a synonym for **
(unlike in Python)
8
sage: 10 % 3 # for integer arguments, %
means mod, i.e., remainder
1
sage: 10/4
5/2
sage: 10//4
# for integer arguments, //
returns the integer quotient
2
sage: 4 * (10 // 4) + 10 % 4 == 10
True
sage: 3^2*4 + 2%5
38
sage: sqrt(3.4)
1.84390889145858
sage: sin(5.135)
-0.912021158525540
505
sage: sin(pi/3)
1/2*sqrt(3)
sage: exp(2)
e^2
sage: n(exp(2))
7.38905609893065
sage: sqrt(pi).numerical_approx()
1.77245385090552
sage: sin(10).n(digits=5)
-0.54402
sage: N(sin(10),digits=10)
-0.5440211109
sage: numerical_approx(pi, prec=200)
3.141592653589793238462643383279502884197169
3993751058209749
sage: a = 5
# a is an integer
sage: type(a)
<type 'sage.rings.integer.Integer'>
sage: a = 5/3 # now a is a rational number
sage: type(a)
<type 'sage.rings.rational.Rational'>
sage: a = 'hello' # now a is a string
sage: type(a)
<type 'str'>
sage: 011
9
sage: 8 + 1
9
sage: n = 011
sage: n.str(8)
# string representation of
n in base 8
'11'
:
506
sage: tan?
Type:
<class
'sage.calculus.calculus.Function_tan'>
Definition: tan( [noargspec] )
Docstring:
The tangent function
EXAMPLES:
sage: tan(pi)
0
sage: tan(3.1415)
-0.0000926535900581913
sage: tan(3.1415/4)
0.999953674278156
sage: tan(pi/4)
1
sage: tan(1/2)
tan(1/2)
sage: RR(tan(1/2))
0.546302489843790
sage: log2?
Type:
<class
'sage.functions.constants.Log2'>
Definition: log2( [noargspec] )
Docstring:
The natural logarithm of the real number
2.
EXAMPLES:
sage: log2
log2
sage: float(log2)
0.69314718055994529
507
sage: RR(log2)
0.693147180559945
sage: R = RealField(200); R
Real Field with 200 bits of precision
sage: R(log2)
0.693147180559945309417232121458176568075500
13436025525412068
sage: l = (1-log2)/(1+log2); l
(1 - log(2))/(log(2) + 1)
sage: R(l)
0.181232218299282499487613818646503114233306
09774776013488056
sage: maxima(log2)
log(2)
sage: maxima(log2).float()
.6931471805599453
sage: gp(log2)
0.6931471805599453094172321215
# 32-bit
0.69314718055994530941723212145817656807
# 64-bit
sage: sudoku?
File:
sage/local/lib/python2.5/sitepackages/sage/games/sudoku.py
Type:
<type 'function'>
Definition: sudoku(A)
Docstring:
Solve the 9x9 Sudoku puzzle defined by the
matrix A.
EXAMPLE:
508
sage: A = matrix(ZZ,9,[5,0,0, 0,8,0, 0,4,9,

0,0,0, 5,0,0,0,3,0, 0,6,7, 3,0,0, 0,0,1,
1,5,0, 0,0,0, 0,0,0, 0,0,0, 2,0,8, 0,0,0,
0,0,0, 0,0,0, 0,1,8, 7,0,0, 0,0,4, 1,5,0,
0,3,0, 0,0,2,0,0,0, 4,9,0, 0,5,0, 0,0,3])
sage: A
[5 0 0 0 8 0 0 4 9]
[0 0 0 5 0 0 0 3 0]
[0 6 7 3 0 0 0 0 1]
[1 5 0 0 0 0 0 0 0]
[0 0 0 2 0 8 0 0 0]
[0 0 0 0 0 0 0 1 8]
[7 0 0 0 0 4 1 5 0]
[0 3 0 0 0 2 0 0 0]
[4 9 0 0 5 0 0 0 3]
sage: sudoku(A)
[5 1 3 6 8 7 2 4 9]
[8 4 9 5 2 1 6 3 7]
[2 6 7 3 4 9 5 8 1]
[1 5 8 4 6 3 9 7 2]
[9 7 4 2 1 8 3 6 5]
[3 2 6 7 9 5 4 1 8]
[7 8 2 9 3 4 1 5 6]
[6 3 5 1 7 2 8 9 4]
[4 9 1 8 5 6 7 2 3]
:
sage: def is_even(n):
....:
return n%2 == 0
sage: is_even(2)
True
sage: is_even(3)
False
sage: def is_divisible_by(number,
divisor=2):
509
....:
return number%divisor == 0
sage: is_divisible_by(6,2)
True
sage: is_divisible_by(6)
True
sage: is_divisible_by(6, 5)
False
sage: is_divisible_by(6, divisor=5)
False
sage: is_divisible_by(divisor=2, number=6)
True
sage: def even(n):
....:
v = []
....:
for i in range(3,n):
....:
if i % 2 == 0:
....:
v.append(i)
....:
return v
Syntax Error:
return v
sage: def even(n):
....:
v = []
....:
for i in range(3,n):
....:
if i % 2 == 0:
....:
v.append(i)
....:
return v
sage: even(10)
[4, 6, 8]
sage: a = 5; b = a + 3; c = b^2; c
64
510
sage: 2 + \
....:
3
5
sage: for i in range(3):
....:
print i
0
1
2
sage: for i in range(2,5):
....:
print i
2
3
4
sage: for i in range(1,6,2):
....:
print i
1
3
5
sage: for i in range(5):
....:
print '%6s %6s %6s'%(i, i^2, i^3)
0
0
0
1
1
1
2
4
8
3
9
27
4
16
64
sage: range(2,10)
[2, 3, 4, 5, 6, 7, 8, 9]
511
sage: v = [1, "hello", 2/3, sin(x^3)]

sage: v
[1, 'hello', 2/3, sin(x^3)]
sage: v[0]
1
sage: v[3]
sin(x^3)
sage: len(v)
4
sage: v.append(1.5)
sage: v
[1, 'hello', 2/3, sin(x^3),
1.50000000000000]
sage: del v[1]
sage: v
[1, 2/3, sin(x^3), 1.50000000000000]
sage: d = {'hi':-2,
sage: d['hi']
-2
sage: d[e]
pi
3/8:pi,
e:pi}
sage: class Evens(list):

....:
def __init__(self, n):
....:
self.n = n
....:
list.__init__(self, range(2,
n+1, 2))
....:
def __repr__(self):
512
....:
up to n."
return "Even positive numbers
sage: e = Evens(10)
sage: e
Even positive numbers up to n.
sage: list(e)
[2, 4, 6, 8, 10]
sage: e.n
10
sage: e[2]
6
:
sage: u = var('u')
sage: diff(sin(u), u)
cos(u)
sage: x = var('x')
sage: solve(x^2 + 3*x + 2, x)
[x == -2, x == -1]
sage:
sage:
[x ==
1/2*b
x, b, c = var('x b c')
solve([x^2 + b*x + c == 0],x)
-1/2*b - 1/2*sqrt(b^2 - 4*c), x == + 1/2*sqrt(b^2 - 4*c)]
sage: x, y = var('x, y')

sage: solve([x+y==6, x-y==4], x, y)
[[x == 5, y == 1]]
513
sage: var('x y p q')

(x, y, p, q)
sage: eq1 = p+q==9
sage: eq2 = q*y+p*x==-6
sage: eq3 = q*y^2+p*x^2==24
sage: solve([eq1,eq2,eq3,p==1],p,q,x,y)
[[p == 1, q == 8, x == -4/3*sqrt(10) - 2/3,
y == 1/6*sqrt(5)*sqrt(2) - 2/3],
[p == 1, q == 8, x == 4/3*sqrt(10) - 2/3, y
== -1/6*sqrt(5)*sqrt(2) - 2/3]]
sage: solns =
solve([eq1,eq2,eq3,p==1],p,q,x,y,
solution_dict=True)
sage: [[s[p].n(30), s[q].n(30), s[x].n(30),
s[y].n(30)] for s in solns]
[[1.0000000, 8.0000000, -4.8830369, 0.13962039],
[1.0000000, 8.0000000, 3.5497035, 1.1937129]]
sage: theta = var('theta')
sage: solve(cos(theta)==sin(theta), theta)
[sin(theta) == cos(theta)]
sage: phi = var('phi')
sage: find_root(cos(phi)==sin(phi),0,pi/2)
0.785398163397448...
sage: u = var('u')
sage: diff(sin(u), u)
514
cos(u)
sage: diff(sin(x^2), x, 4)
16*x^4*sin(x^2) - 48*x^2*cos(x^2) 12*sin(x^2)
sage:
sage:
sage:
2*x
sage:
34*y
x, y = var('x,y')
f = x^2 + 17*y^2
f.diff(x)
f.diff(y)
sage: integral(x*sin(x^2), x)
-1/2*cos(x^2)
sage: integral(x/(x^2+1), x, 0, 1)
1/2*log(2)
sage: f = 1/((1+x)*(x-1))
sage: f.partial_fraction(x)
-1/2/(x + 1) + 1/2/(x - 1)
sage: t = var('t')
# define a variable t
sage: x = function('x',t)
# define x to be
a function of that variable
sage: DE = diff(x, t) + x - 1
sage: desolve(DE, [x,t])
(_C + e^t)*e^(-t)
sage: s = var("s")
sage: t = var("t")
sage: f = t^2*exp(t) - sin(t)
sage: f.laplace(t,s)
-1/(s^2 + 1) + 2/(s - 1)^3
515

http://doc.sagemath.org/html/en/tutorial/
516
:
:
)1
sage.game_theory.catalog_normal_form_games.AntiCoor
)dinationGame(A=3, a=3,B=5, b=1,C=1, c=5,D=0, d=0

.
:
D
c
a
B =
b
A=
< and < , < , <
)2
sage.game_theory.catalog_normal_form_games.B
)(attleOfTheSexes
.
.

517
0 2
2 1
B=
0 3
A=
3
-1
-2
-3 %75 %75 .
)3
(sage.game_theory.catalog_normal_form_games.Chicken
)A=0,a=0, B=1, b=-1,C=-1, c=1,D=-10, d=-10
2 .
. .

D
c
A=
B =
b d
< and < , < , < > >

0
1 10
0
1
B=
1 10
A=
518
3
-1 .
-2 .
-3 1 10 .
)4
sage.game_theory.catalog_normal_form_games.Coordina
tionGame
)(A=10, a=5, B=0, b=0, C=0, c=0,D=5, d=10
D
c
a
B =
b
A=
> , > > > ,

10
0 5
5 0
B=
0 10
A=
)5
sage.game_theory.catalog_normal_form_games.HawkDove
)(v=2, c=3
519
.
) ( )(.
-1
.
-2 .
-3 ) (
c .c > v

v
2c v
0
v 2
v 2c
0
B=
v
v
2
A=
= 2 . = 3
2
0
1
2 0
B=
2
A=
anti coordination game 3

:
-1 .
-2 .
)6
sage.game_theory.catalog_normal_form_games.Matching
)(Pennies
520

. 1 .
:
A=

.
)7
)(sage.game_theory.catalog_normal_form_games.Pigs
. . .
6

. 3/4
.
. 1/3
.
)
.(A
521
6 0
1
4
B=
1 0
A=
.
)8
sage.game_theory.catalog_normal_form_games.Prisoner
sDilemma
)(R=-2, P=-4, S=-5,T=0

.
.
.
.

:
R
T P
R T
B=
S P
A=
> > > .

R -1 .
S -2 .
522
P -3 .
T -4 .
:
2
0
4
2
0
B=
5 4
A=
.
)9
)(sage.game_theory.catalog_normal_form_games.RPS
- -
. 3
.
) (
) ( .
) ( .
.

523
1
0
1
A= 1
)10
)(sage.game_theory.catalog_normal_form_games.RPSLS
.
-1 .
-2 .
-3 .
-4 ) (.
-5 .
-6 .
-7 .
-8 .
-9 .
-10 .
:
1
1
0
1 1
1 1 0
1 1 1
A = 1
)11
sage.game_theory.catalog_normal_form_games.StagHunt
)(
524
. .
.
.
.
:
5
4 2
5 4
B=
0 2
A=
. 3 :
-1 .
-2 .
-3 2/3.
)12
sage.game_theory.catalog_normal_form_games.Travelle
)rsDilemma(max_value=10
.
.
10

2 10 .
525

.
2
2 .

:
6 5 4 3 2 1 0
6 5 4 3 2 1
0
0
0
0
0
10 8 5 4 3 2 1
9 7 4 3 2 1
8 8 6 3 2 1
7 7 7 5 2 1
6 6 6 6 4 1
5 5 5 5 5 3
4 4 4 4 4 4
5
4
11 10 9 8 7 6 5 4
9 10 9 8 7 6 5 4
9 8 7 6 5 4
2 2 2 4 5 4
1 1 1 1 3 4
0 0 0 0 0 2
1
0
1
0
7 8 7 6 5 4
4 6 7 6 5 4
3 3 5 6 5 4
526
10
11
10
A = 8
4
10
B = 4
527

Decision and Games

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Decision and Games

Încărcat de

Drepturi de autor:

Formate disponibile

Excel Solver Treeplan

SilverDecisions Gambit SageMath

Two Persons Zero Sum Games

Maximizing the Utility Function

Payoff Table Analysis

Decision Making Criteria

States of Nature - The Maximum Criterion

States of Nature - The Payoff Table

States of Nature - The Maximum Criterion

States of Nature - The Payoff Table

500 - (-100) = 600

States of Nature - The Regret Table

States of Nature - The Maximax Criterion

- Decision Making Under

Decision Making Under Risk

Treeplan Excel Treeplan

Decision Tree Designer

1 100 = 20 60= 0.4 X100 +

4 150 = 60 65= 0.1 X50 + ) 100

1,200 + 200 - 1,300 = 100.

Number of Cars Ordered

lij = max rmj rij

Expected Monetary Value

s1: both bids rejected

Additional Cost Per Kilo

Probability 0.25 0.08 0.10 0.12 0.15 0.12 0.10

Cooperative :Non-Cooperative Games

Best of the Worst

Maximin value value of the game Minimax value

a11 a12 a1n

A's expected payoff

B's pure strategy

B' expected payoff

A's pure strategy

max min ai1 xi , ai 2 xi ,..., ain xi

v = min ai1 xi , ai 2 xi ,..., ain xi

min max a1 j y j , a2 j y j ,..., amj y j

AMi = probability A makes move i, i = a, b, or c.

Global optimal solution found.

Strictly Dominant Strategy

}{raise price, raise price }price, lower price

} {raise price, raise price

Compute Nash equlibria

:Sage Normal Form

sage: A = matrix([[3, 1], [0, 2]])

print "Utility for {}: ".format(ne)

print vector(ne[0]) * A * vector(ne[1]),

Utility for [(0, 1), (0, 1)]:

A = matrix([[1, -1], [-1, 1]])

sage: [vector([1/2, 1/2]) * M * vector([1/2, 1/2])

[1, 0, -1, -1, 1],

[-1, 1, -1, 0, 1],

[1, -1, 1, -1, 0]])

[15, 10, 105],

[19, 204, 55],

[20, 200, 590]])

sage: B = matrix([[2, 1, 2],

sage: g=NormalFormGame([A, B])

sage: g = NormalFormGame([A, B])

sage: A = matrix([[160, 205, 44],

[175, 180, 45],

[201, 204, 50],

[120, 207, 49]])

sage: B = matrix([[2, 2, 2],