An Introductory Review on Quantum Game Theory

Wenjie Liu1, Jingfa liu, Mengmeng Cui

School of Computer and software
Nanjing University of Information Science &
Technology
Nanjing, 210044, China
{wenjieliu, jfliu, cmmgxj}@nuist.edu.cn
Ming He
Institute of Command Automation
PLAUST
Nanjing, 211101, China
mingh@126.com


AbstractWith utilizing some quantum mechanism into the
game theory, quantum game theory provides us with the
ability to achieve results not otherwise possible. In this paper, a
brief overview of classical game theory and some game-
theoretic terms are depicted. Then, an introduction to
quantum theory is presented as well as some correlative works
in this field. Finally, some well-known quantum games are
analyzed and discussed, which shows its predominant effect on
some games problems (the Prisoners Dilemma, Parrondos
Games, etc.) .
Keywords-game theory, quantum game theory, quantum
games, quantum computation
I. INTRODUCTION
Game theory is the study of decision making of competing
agents in some conflict situation. It was originally developed
for use in economics by von Neumann and Morgenstern in
1944[1]. Subsequently, much work was carried on game
theory and it is now a mature discipline used in areas such as
the social sciences, engineering, biology, computer science
and international relations.
Computers that exploit the inherent features of quantum
mechanics, such as superposition and entanglement, are
known as quantum computer. With the recent interest in
quantum computation and quantum information, there has
been an effort to recast classical game theory using quantum
probability amplitudes, and hence study the effect of
quantum superposition, interference and entanglement on the
agents optimal strategies. In 1999, two groups firstly
presented the idea of applying quantum mechanics to game
theory, independently[2, 3]. Replacing the classical
probabilities of game theory by quantum amplitudes,
quantum game theory (QGT for short) creates the possibility
of new effects resulting from superposition or entanglement.
Up to now, QGT has concentrated on observing these new
effects amongst the traditional settings of game theory. As an
application within mathematical economics, it can be applied
to model markets and auctions, assuming traders can use
quantum protocols [4, 5]. Moreover, quantum game-theoretic
techniques could be used in quantum communication [6] or
quantum information processing[7]. For example, quantum
communication in competition with an eavesdropper can be
considered as games[8]. There has been a proposal to use
quantum game theory in a quantum teleportation[9] and in
quantum state estimation and quantum cloning[10]. And a
review of suggested applications of quantum games is given
by Guo et al [11].
In Ref. [2], the author studied a simple coin tossing game
and showed how a player utilizing quantum superposition
could win with certainty against a classical player. Using the
well known prisoners dilemma as an example, a general
protocol for two player-two strategy quantum games with
entanglement was developed In Ref. [3], and it was extended
to multiplayer games by Benjamin and Hayden [12]. Since
then, more work has been done on quantum prisoners
dilemma [13-16] and a number of other games have been
converted to the quantum realm including the battle of the
sexes [17, 18], the Monty Hall problem [19, 20], rock-
scissors-paper [21], and others[22, 23].
The rest of this paper is organized as follows. In Sec. 2,
we introduce some terminology and basic ideas of game
theory. In Sec. 3, the idea of quantum game theory and some
general quantum games will be discussed and analyzed,
along with a review of the existing work, and finally
conclusion will be drawn in Sec. 4.
II. GAME THEORY
Game theory is a tool for rational decision making in
conflict situations. It has long been commonly used in
economics, the social sciences and biology to model decision
making situations where the outcomes are contingent upon
the interacting strategies of two or more agents with
conflicting, or at best, self- interested motives. In the
following section, we will introduce some basic terms and
gives a simple example.
A. Basic Ideas and Terminology
Game theory attempts to mathematically model a
situation where agents interact. The agents in the game are
called players, their possible actions moves, and
a prescription that specifies the particular move to be made
in all possible game situations a strategy. That is, a strategy
represents a plan of action that contains all the contingencies
that can possibly arise within the rules of the game. In
response to some particular game situation, a pure strategy
consists of always playing a given move, while a strategy
that utilizes a randomizing device to select between
2010 Fourth International Conference on Genetic and Evolutionary Computing
978-0-7695-4281-2/10 $26.00 2010 IEEE
DOI 10.1109/ICGEC.2010.102
386
different moves is known as a mixed strategy. The utility to
a player of a game outcome is a numerical measure of the
desirability of that outcome for the player. A payoff matrix
gives numerical values to the players utility for all the
game outcomes. It is assumed that the players will seek to
maximize their utility within the given rules of the game.
Games in which the choices of the players are known as
soon as they are made are called games of perfect
information. These are the main ones that are of interest to
us here.
A dominant strategy is one that does at least as well as
any competing strategy against any possible moves by the
other player(s). The Nash equilibrium (NE) is the most
important of the possible equilibriums in game theory. It is
the combination of strategies from which no player can
improve his/her payoff by a unilateral change of strategy. A
Pareto optimal outcome is one from which no player can
obtain a higher utility without reducing the utility of another.
Strategy A is evolutionary stable against B if, for all
sufficiently small, positive c , A performs better than
B against the mixed strategy (1 ) A B c c + . An evolutionary
stable strategy (ESS) [24] is one that is evolutionary stable
against all other strategies. The set of all strategies that are
ESS is a subset of the NE of the game. A two player zero-
sum game is one where the interests of the players are
diametrically opposed. That is, the sum of the payoffs for
any game result is zero. In such a game a saddle point is an
entry in the payoff matrix for (say) the row player that is
both the minimum of its row and the maximum of its
column.
B. An Example: the Prisoners Dilemma
A famous 2 x 2 game that has attracted much attention is
the Prisoners Dilemma [25]. Here, the players moves are
known as cooperation C or defection D, for reasons that
shall become clear. The payoff matrix is such that there is a
conflict between the NE and the PO outcome. The payoff
matrix can be written as
Bob: C Bob: D
Alice: C (3, 3) (0, 5)
Alice: C (5, 0) (1, 1)
(1)
where the numbers in parentheses represent the row (Alice)
and column (Bob) players payoffs, respectively. The game
is symmetric and there is a dominant strategy, that of always
defecting, since it gives a better payoff if the other player
cooperates (five instead of three) or if the other player
defects (one instead of zero). Where both players have a
dominant strategy, this combination is the NE.
The NE outcome {D, D} is not such a good one for the
players, however, since if they had both cooperated they
would have both received a payoff of three, the PO result. In
the absence of communication or negotiation we have a
dilemma between the personal and the mutual good, some
form of which is responsible for much of the misery and
conflict through out the world. Game theory does not have a
solution. In a one-off Prisoners Dilemma the rational player
postulated by the theory should defect. In the real world the
opportunity to play the game repeatedly and the ability to
negotiate helps to foster some degree of cooperation even in
pure prisoners dilemma situations[26]. There is extensive
literature on the prisoners dilemma and it is mentioned in
any introductory text on game theory (for example, see Ref.
[27] ).
III. QUANTUM GAME THEORY
A. Quantum Penny Flip
A two state system, such as a coin, is one of the simplest
gaming devices. If we have a player that can utilize quantum
moves, we can demonstrate how the expanded space of
possible strategies can be turned to advantage. Meyer, in his
seminal work on quantum game theory [2], considered the
simple game Penny Flip that consists of the following: Alice
prepares a coin in the heads state and places it in a box
where neither player can see it. Bob can choose to either flip
the coin or leave its state unaltered, and Alice, without
knowing Bobs action, can do likewise. Finally, Bob has a
second turn at the coin. The coin is now examined and Bob
wins if it shows heads. A classical coin clearly gives both
players an equal probability of success unless they utilize
knowledge of the others psychological bias, and such
knowledge is beyond analysis by game theory.
To quantize this game, we should replace the coin by a
two state quantum system such as a spin one-half particle.
Suppose Bob is given the power to make quantum moves
while Alice is restricted to classical ones. Let 0 represent
the heads state and 1 the tails state. Alice initially prepares
the system in the 0 state. Bob can proceed by first
applying the Hadamard Operator,
1
(( 0 1 ) 0 ( 0 1 ) 1 )
2
H = + + , (2)
putting the system into the equal superposition of the two
states : ( 0 1 ) / 2 + Now Alice can leave the "coin
alone or interchange the states 0 and 1 . If the coherence
of the system is maintained either action will leave the
system unaltered, a fact that can be exploited by Bob. In his
second move he applies the Hadamard operator again
resulting in the pure state 0 , thus winning the game with
certainty. Bob exploits his ability to apply any unitary
operation and the possibility of a superposition to make
Alices strategy irrelevant. In this case it is just the increased
possibilities available to the quantum player that proved
decisive. And in some situation, Quantum entanglement is
often exploited by the quantum player.
B. Eiserts Quantum PrisonersDilemma
Eisert et al [3, 13] gave a physical model of the prisoners
dilemma and suggested that the players can escape the
dilemma if they both resort to quantum strategies. In this
model, Unitary quantum strategies are any

(2) : U SU e
387
cos( / 2) sin( / 2)

( , , )
sin( / 2) cos( / 2)
i i
i i
e ie
U
ie e
o |
| o
u u
u o |
u u

| |
=
|
\ .
(3)
Where | | 0, u t e and | | , , o | t t e . In quantum prisoners
dilemma, a player with access to quantum strategies can
always do at least as well as a classical player. If
cooperation is associated with the 0 state and defection
with the 1 state, then the strategy always cooperate

(0) C U I = and the strategy always defect is

( ) C U F t = . Against a classical Alice playing

( ) U u , a
quantum Bob can play Eiserts miracle move:
1 1

( , , 0)
2 2
1 1
2
i
M U
t t
| |
= =
|

\ .
(4)
that yields a payoff of
B
$ 3 2sinu = + for Bob while
leaving Alice with only
A
$ (1 sin ) / 2 u = .
In this case, the dilemma is removed in favor of the
quantum player. In the partially entangled case, there is a
critical value of the entanglement parameter
arcsin(1/ 5) = , below which the quantum player should
revert to the classical dominant strategy D to ensure a
maximal payoff [3]. At the critical level of entanglement
there is a phase change-like transition between the quantum
and classical domains of the game (see Refs. [14, 28].
C. Quantum Parrondos Games
A Parrondos game is an interesting problem in game
theory. Two games that are losing when played individually
can be combined to produce a winning game[29]. The game
can be put into the form of a gambling utilizing a set of
biased coins.
There have been recent attempts to create quantum
versions of Parrondos games. Flitney and Abbott [30, 31]
studied a quantum version of the Parrondos game where the
rotation operators representing the toss of a classical biased
coin are replaced by general SU(2) operators to transform the
game into the quantum domain. They found that
superposition of qubits can couple the two games and
produce interference leading to different payoffs than in the
classical case. Meyer and Blumer[32] uses a quantum lattice
gas automaton to construct a Parrondo game involving a
single particle in an unbiased random walk between lattice
sites. Ratcheting in one direction is achieved by
multiplication by a position dependent phase factor and the
resulting quantum interference.

IV. CONCLUSION
Game theory is the mathematical theory of decision
making in competitive situations. The new field of quantum
game theory is the extension of game theory into the
quantum realm. In this article, A protocol for two player-two
strategy quantum games has been discussed, which can be
extended to more players and larger strategic spaces. Some
examples of the various quantum game-theoretic
investigations have been given.
ACKNOWLEDGMENT
This work is supported by the Research Foundation of
NUIST (20080298) and National Natural Science
Foundation of China (60873101).

REFERENCES

[1] J. V. Neumann, and O. Morgenstern, The Theory of Games and
Economic Behavior: Princeton University Press, 1944.
[2] D. A. Meyer, Quantum Strategies, Physical Review Letters, vol. 82,
no. 5, pp. 1052, 1999.
[3] J. Eisert, M. Wilkens, and M. Lewenstein, Quantum Games and
Quantum Strategies, Physical Review Letters, vol. 83, no. 15, pp.
3077, 1999.
[4] E. W. Piotrowski, and J. Sladkowski, Quantum market games,
Physica A: Statistical Mechanics and its Applications, vol. 312, no. 1-
2, pp. 208-216, 2002.
[5] E. W. Piotrowski, J. Sladkowski, and J. Syska, Interference of
quantum market strategies, Physica A: Statistical Mechanics and its
Applications, vol. 318, no. 3-4, pp. 516-528, 2003.
[6] H. E. Brandt, Qubit devices and the issue of quantum decoherence,
Progress in Quantum Electronics, vol. 22, no. 5-6, pp. 257-370, 1999.
[7] C. F. Lee, and N. F. Johnson, Exploiting randomness in quantum
information processing, Physics Letters A, vol. 301, no. 5-6, pp. 343-
349, 2002.
[8] L. Goldenberg, L. Vaidman, and S. Wiesner, Quantum Gambling,
Physical Review Letters, vol. 82, no. 16, pp. 3356, 1999.
[9] S. Pirandola, A quantum teleportation game, International Journal
of Quantum Information, vol. 3, no. 1, pp. 239-243, 2005.
[10] C. F. Lee, and N. F. Johnson, Game-theoretic discussion of quantum
state estimation and cloning, Physics Letters A, vol. 319, no. 5-6, pp.
429-433, 2003.
[11] H. Guo, J. Zhang, and G. J. Koehler, A survey of quantum games,
Decision Support Systems, vol. 46, no. 1, pp. 318-332, 2008.
[12] S. C. Benjamin, and P. M. Hayden, Multiplayer quantum games,
Physical Review A, vol. 64, no. 3, pp. 030301, 2001.
[13] J. Eisert, and M. Wilkens, Quantum games, Journal of Modern
Optics, vol. 47, no. 14, pp. 2543 - 2556, 2000.
[14] J. Du, X. Xu, H. Li et al., Entanglement playing a dominating role in
quantum games, Physics Letters A, vol. 289, no. 1-2, pp. 9-15, 2001.
[15] J. Du, H. Li, X. Xu et al., Entanglement enhanced multiplayer
quantum games, Physics Letters A, vol. 302, no. 5-6, pp. 229-233,
2002.
[16] A. Iqbal, and A. H. Toor, Quantum repeated games, Physics Letters
A, vol. 300, no. 6, pp. 541-546, 2002.
[17] L. Marinatto, and T. Weber, A quantum approach to static games of
complete information, Physics Letters A, vol. 272, no. 5-6, pp. 291-
303, 2000.
[18] A. Iqbal, and A. H. Toor, Evolutionarily stable strategies in quantum
games, Physics Letters A, vol. 280, no. 5-6, pp. 249-256, 2001.
[19] C.-F. Li, Y.-S. Zhang, Y.-F. Huang et al., Quantum strategies of
quantum measurements, Physics Letters A, vol. 280, no. 5-6, pp.
257-260, 2001.
[20] A. P. Flitney, and D. Abbott, Quantum version of the Monty Hall
problem, Physical Review A, vol. 65, no. 6, pp. 062318, 2002.
[21] A. Iqbal, and A. H. Toor, Quantum mechanics gives stability to a
Nash equilibrium, Physical Review A, vol. 65, no. 2, pp. 022306,
2002.
388
[22] N. F. Johnson, Playing a quantum game with a corrupted source,
Physical Review A, vol. 63, no. 2, pp. 020302, 2001.
[23] K. Roland, and et al., Evolutionary quantum game, Journal of
Physics A: Mathematical and General, vol. 34, no. 41, pp. L547,
2001.
[24] J. M. Smith, and G. R. Price, The Logic of Animal Conflict, Nature,
vol. 246, no. 5427, pp. 15-18, 1973.
[25] A. Rapoport, and A. M. Chammah, Prisoner's dilemma: a study in
conflict and cooperation, Anatol Rapoport,Albert M. Chammah:
Anatol Rapoport,Albert M. Chammah, 1965.
[26] R. Axelrod, and W. Hamilton, The evolution of cooperation,
Science, vol. 211, no. 4489, pp. 1390-1396, March 27, 1981, 1981.
[27] M. J. Osborne, An introduction to game theory, Cambridge, UK:
Cambridge University Press, 2004.
[28] J. Du, and et al., Phase-transition-like behaviour of quantum games,
Journal of Physics A: Mathematical and General, vol. 36, no. 23, pp.
6551, 2003.
[29] P. V. E. McClintock, Random fluctuations: Unsolved problems of
noise, Nature, vol. 401, no. 6748, pp. 23-25, 1999.
[30] A. P. Flitney, J. Ng, and D. Abbott, Quantum Parrondo's games,
Physica A: Statistical Mechanics and its Applications, vol. 314, no. 1-
4, pp. 35-42, 2002.
[31] A. P. Flitney, and D. Abbott, Quantum models of Parrondo's
games, Physica A: Statistical Mechanics and its Applications, vol.
324, no. 1-2, pp. 152-156, 2003.
[32] D. A. Meyer, and H. Blumer, Parrondo Games as Lattice Gas
Automata, Journal of Statistical Physics, vol. 107, no. 1, pp. 225-
239, 2002.


389