One Approach to Fuzzy Matrix Games Konstantin N. Kudryavtsev1 , Irina S. Stabulit2 , and Viktor I. Ukhobotov3 1 South Ural State University, Chelyabinsk, Russia kudrkn@gmail.com 2 South Ural State Agrarian University, Chelyabinsk, Russia irisku76@mail.ru 3 Chelyabinsk State University, Chelyabinsk, Russia ukh@csu.ru Abstract. The paper is concerned with a two-person zero sum matrix game with fuzzy payoffs, with saddle-point being defined with a clas- sic definition in matrix game. In order to compare fuzzy numbers some different ordering operators can be used. The original game can be asso- ciated with the bimatrix game with crisp payoffs which are the operators value on a fuzzy payoff. We propose that every player use its ordering operator. The following statement can be proposed: when the ordering operators are linear, the same equilibrium strategy profile can be used for the matrix game, with fuzzy payoffs being the same for the bimatrix crisp game. We introduce and employ the algorithm of constructing a saddle-point in a two-person zero sum matrix game with fuzzy payoffs. In the instances of matrix games we use such ordering operators and construct the saddle-point. Keywords: Fuzzy game · Matrix game · Saddle-point · Nash equilib- rium 1 Introduction Game theory is well known to take a significant part in decision making, this theory being often used to model the real world. Applied in real situations the game theory is difficult to have strict values of payoffs, because it is difficult for players to analyze some data of game. Thus, the players’ information cannot be considered complete. Besides the players can have vague targets. This uncertainty and lack of precision may be modeled as fuzzy games. Fuzzy sets are known to be initially used in non-cooperative game theory by Butnariu [3] to prove the belief of each player for strategies of other players. Since then fuzzy set theory has been applied in cooperative and non-cooperative games. The results of fuzzy games overview are given in [10]. Recently there were made Copyright c by the paper’s authors. Copying permitted for private and academic purposes. In: S. Belim et al. (eds.): OPTA-SCL 2018, Omsk, Russia, published at http://ceur-ws.org One Approach to Fuzzy Matrix Games 229 various efforts in fuzzy bi-matrix game theory namely Maeda [11], Nayak [13], Dutta [6], Seikh [14], Verma and Kumar [16]. In [9], we represented the approach which generalizes some other ideas ([4], [5],[6] at al.). 2 Fuzzy Numbers In this part, some basic definitions and results of fuzzy numbers and fuzzy arith- metic operations will be reminded. Here we will follow to [21]. A fuzzy set can be considered as a subset à of universal set X ⊆ R by its membership function µÃ (·) with a real number µÃ (x) in the interval [0, 1] and assigns to each element x ∈ R. Definition 2.1. A fuzzy subset à defined on R, is said to be a fuzzy number if its membership function µÃ (x) comply with the following conditions: (1) µÃ (x) : R → [0, 1] is upper semi-continuous; (2) µÃ (x) = 0 for ∀ x 6∈ [a, d]; (3) There exist real numbers b, c such that a 6 b 6 c 6 d and (a) x1 < x2 ⇒ µÃ (x1 ) < µÃ (x2 ) ∀ x1 , x2 ∈ [a, b]; (b) x1 < x2 ⇒ µÃ (x1 ) > µÃ (x2 ) ∀ x1 , x2 ∈ [a, b]; (c) µÃ (x) = 1, ∀x ∈ [b, c]. The α-cut of a fuzzy number à plays an important role in parametric ordering of fuzzy numbers. The α-cut or α-level set of a fuzzy number Ã, denoted by Ãα , is defined as Ãα = {x ∈ R | µÃ (x) > α} for all α ∈ (0, 1]. The support or 0-cut Ã0 is defined as the closure of the set Ã0 = {x ∈ R | µÃ (x) > 0}. Every α-cut is a closed interval Ãα = [gà (α), Gà (α)] ⊂ R, where gà (α) = inf{x ∈ R | µÃ (x) > α} and Gà (α) = sup{x ∈ R | µÃ (x) > α} for any α ∈ [0, 1]. The sets of fuzzy number is denoted as F. Then, two types of fuzzy numbers are used. Definition 2.2. Let à be a fuzzy number. If the membership function of à is given by  x−a+l  l , f or x ∈ [a − l, a], µÃ (x) = a+r−x , f or x ∈ [a, a + r],  r 0, otherwise, where, a, l and r are all real (crisp) numbers, and l, r are non-negative. Then à is called a triangular fuzzy number, denoted by à = (a, l, r). The sets of triangular fuzzy number are denoted as F3 . Definition 2.3. Let à be a fuzzy number. If the membership function of à is given by  x−a+l  l , f or x ∈ [a − l, a],  1, f or x ∈ [a, b]  µÃ (x) = b+r−x , f or x ∈ [b, b + r],  r   0, otherwise, where, a, b, l and r are all real (crisp) numbers, and l, r are non-negative. Then à is called a trapezoidal fuzzy number, denoted by à = (a, b, l, r). [a, b] is the core of Ã. 230 K. Kudryavtsev et al. The sets of trapezoidal fuzzy number are denoted as F4 . If à = (a1 , l1 , r1 ) and B̃ = (a2 , l2 , r2 ) are two triangular fuzzy numbers, arithmetic operations on à and B̃ are defined as follows: Addition: à + B̃ = C̃ = (a1 + a2 , l1 + l2 , r1 + r2 ), C̃ ∈ F3 . Scalar multiplication: ∀ k > 0, k ∈ R, k à = (ka1 , kl1 , kr1 ), k à ∈ F3 . If à = (a1 , b1 , l1 , r1 ) and B̃ = (a2 , b2 , l2 , r2 ) are two trapezoidal fuzzy num- bers, arithmetic operations on à and B̃ are defined as follows: Addition: à + B̃ = C̃ = (a1 + a2 , b1 + b2 , l1 + l2 , r1 + r2 ), C̃ ∈ F4 . Scalar multiplication: ∀ k > 0, k ∈ R, k à = (ka1 , kb1 , kl1 , kr1 ), k à ∈ F4 . Generaly, if à and B̃ are two fuzzy numbers and à + B̃ = C̃, λà = D̃ and λ = const > 0, then C̃α = [gà (α) + gB̃ (α), Gà (α) + GB̃ (α)], and D̃α = [λgà (α), λGà (α)] for any α ∈ [0, 1]. To compare fuzzy numbers is crucial issue. There are plenty of various meth- ods for comparing fuzzy numbers. For instance, fuzzy numbers can be ranked with the help of defuzzification methods.Defuzzification is the process of produc- ing a real (crisp) value which correspond to a fuzzy number, the defuzzification approach being used for ranking fuzzy numbers. Fuzzy numbers are initially defuzzified and then the obtained crisp numbers are organised using the order relation of real numbers. A function for ranking fuzzy subsets in unit interval was introduced by Yager in [17]. It was based on the integral of mean of the α-cuts. Yager index is Z1 1 Y (Ã) = (gà (α) + Gà (α)) dα. 2 0 Another methods for ordering fuzzy subsets in the unit interval were sug- gested by Jain in [8], Baldwin and Guild in [1]. The subjective approach for ranking fuzzy numbers was developed by Ibanez and Munoz in [7], the following number as the average index for fuzzy number à beind defined in [7] Z VP (Ã) = fà (α)dP (α), Y with Y is a subset of the unit interval and P is a probability distribution on Y , the definition of fà being subjective for decision maker. One Approach to Fuzzy Matrix Games 231 The ordering operator was proposed by Ukhobotov in [15] Z1 U (Ã, ν) = ((1 − ν)gà (α) + νGà (α)) dα, 0 with crisp parameter ν ∈ [0, 1], different behavior of the decision maker being corresponded with different ν. Other defuzzification operators are given in [2]. Definition 2.4. Let à and B̃ are a fuzzy numbers, T : F → R is the operator of defuzzification (T (·) = Y (·), VP (·), U (·, ν) etc.). We say that B̃ is preferable to à by the defuzzification operator T (à T B̃) if and only if T (Ã) 6 T (B̃). The defuzzification operator T is dependant on the order relation T . Then, we give the example. Example 2.1. Let T (·) = U (·, ν) and Ã, B̃, C̃ ∈ F3 , à = (40, 8, 10), B̃ = = (45, 20, 10), C̃ = (42, 6, 4). If X̃ = (a, l, r) ∈ F3 , then νr − (1 − ν)l U (X̃, ν) = a + . 2 Next, if ν = 0, then U (Ã, 0) = 36, U (B̃, 0) = 35, U (C̃, 0) = 39. If ν = 21 , then U (Ã, 12 ) = 40, 5, U (B̃, 12 ) = 42, 5, U (C̃, 21 ) = 41, 5. If ν = 1, then U (Ã, 1) = 45, U (B̃, 1) = 50, U (C̃, 1) = 44. Therefore, B̃ U (·,0) à U (·,0) C̃, à U (·, 12 ) C̃ U (·, 21 ) B̃, C̃ U (·,1) à U (·,1) B̃. Definition 2.5. If ∀ Ã, B̃ ∈ F, ∀ α, β = const T (αà + β B̃) = αT (Ã) + βT (B̃), then the defuzzification operator T (·) is the linear defuzzification operator. Clear, Yager index Y (·) and operator U (·, ν) is linear. 3 Crisp Games In this part, some basic definitions of non-cooperative game theory are presented. 232 K. Kudryavtsev et al. 3.1 Noncooperative N-Person Games Let us consider a non-cooperative N -players game in the class of pure strategies Γ = hN, {Xi }i∈N , {fi (x)}i∈N i, (1) where N = 1, ..., N is the set of players’ serial numbers; each player i chooses and applies its own pure strategy xi ∈ Xi ⊆ Rni , with no coalition with the others, which induces a strategy profile being formed Y X x = (x1 , ..., xN ) ∈ X = Xi ⊂ Rn (n = ni ); i∈N i∈N for each i ∈ N, a payoff function fi (x) is defined on the strategy profile set X, which gives the payoff of player i. fi (x) is payoff function of player i (i ∈ N). In addition, denote (xkzi ) = (x1 , ..., xi−1 , zi , xi+1 , ...xN ), f = (f1 , ..., fN ). Definition 3.1. A strategy profile xe = (xe1 , ..., xeN ) ∈ X is called a Nash equi- librium in the game (1) if max fi (xe kxi ) = fi (xe ) (i ∈ N). (2) xi ∈Xi The set of all {xe } in the game (1) will be designated by X e . 3.2 Bimatrix Games A bimatrix game defined by a pair (A, B) of real m×n matrices being considered, matrices A and B are payoffs to play I and II, respectively. M denotes the set of pure strategies of player I (matrix rows) and N stands for the set of pure strategies of player II (columns). M = (1, . . . , m), N = (1, . . . , n). The sets of mixed strategies of the two players being called X and Y , we want to write expected payoffs as matrix products xAy and xBy, for mixed strategies x and y, so that x is a row vector and y is a column vector. That is, X X = {(x1 , . . . , xm ) | xi > 0 (∀i ∈ M ), xi = 1} i∈M and X Y = {(y1 , . . . , yn ) | yj > 0 (∀j ∈ N ), yj = 1}. j∈N Definition 3.2. A pair (xe , y e ) ∈ X × Y is called a Nash equilibrium for the game (A, B) if xe Ay e > xAy e ∀x ∈ X, xe By e > xe By ∀y ∈ Y. The set of Nash equilibrium for a game (A, B) is proved to be non-empty in [12]. If matrix B = −A, a bimatrix game is considered as a zero-sum matrix game. A solve of a zero-sum matrix game being a saddle-point. One Approach to Fuzzy Matrix Games 233 3.3 Internally Instable Set of Nash Equilibrium Now, consider internal instability of X e . A subset X ∗ ⊂ Rn is internally instable, if there exist at least two strategy profiles x(j) ∈ X ∗ (j = 1, 2) such that h i h i f (x(1) ) < f (x(2) ) ⇔ fi (x(1) ) < fi (x(2) ) (i ∈ N) , (3) and internally stable otherwise. Example 3.1. Consider a bimatrix game of the form (A, B), where     2 5 1 1 0 8 A = 3 4 6, B = 9 3 5. 6 7 2 2 7 6 (i) (i) There are 3 Nash equilibrium strategy profiles x(i) = (x1 , x2 ) (i = 1, 2, 3):     (1) 1 2 4 3 x = 0, , , 0, , , 3 3 7 7     1 1 4 3 x(2) = 0, , , , 0, , x(3) = ((0, 0, 1), (0, 1, 0)). 2 2 7 7 Consequently, the set X e is internally instable in the game (A, B), as for x(3) it follows that (3) (3) (1) (1) f1 (x(3) ) = x1 Ax2 = 7 > f1 (x(1) ) = x1 Ax2 = 34 7 , (3) (3) (1) (1) f2 (x(3) ) = x1 Bx2 = 7 > f2 (x(1) ) = x1 Bx2 = 17 3 , and (3) (3) (2) (2) f1 (x(3) ) = x1 Ax2 = 7 > f1 (x(2) ) = x1 Ax2 = 30 7 , (3) (3) (3) (2) (2) (2) 11 f2 (x ) = x1 Bx2 = 7 > f2 (x ) = x1 Bx2 = 2 . We note that in the non-antagonistic setting of the game (1), the internal in- stability effect vanishes if there exists a unique Nash equilibrium strategy profile in (1). Associate the following auxiliary N -criterion problem with the game (1): Γv = hX e , {fi (x)}i∈N i, (4) where the set X e of alternatives x coincides with the set of Nash equilibrium strategy profiles xe in the game (1), and the ith criterion fi (x) is the payoff function of player i. Definition 3.3. An alternative xP ∈ X e is Pareto optimal (efficient) in the problem (4), if ∀x ∈ X e the system of inequalities fi (x) > fi (xP ) (i ∈ N), is infeasible, with at least one being a strict inequality. {xP } XP . Designate by X P the set of all {xP }. 234 K. Kudryavtsev et al. According to Definition 3.3, the set X P satisfies the inclusionX P ⊆ X e and is internally stable. The following statement is obvious: if for all x ∈ X e we have X X fi (x) 6 fi (xP ), (5) i∈N i∈N then xP gives the Pareto optimal alternative in the problem (4). Definition 3.4. [18] A strategy profile x∗ ∈ X is called a Pareto-optimal Nash equilibrium (P ON E) for the game (1) if x∗ is a) a Nash equilibrium in (1) (Definition 3.1), and b) a Pareto optimum in (5) (Definition 3.3). 4 Game with Fuzzy Payoffs Further, a non-cooperative N -person game is considered Γe = hN, {Xi }i∈N , {fei (x)}i∈N i. (6) A game (6) differs from (1) according to payoffs functions. In (6), a payoff func- tion of player i is fei (x) : X → F. In addition, Xi involves only a finite number of elements, Γe being a finite game with fuzzy payoffs. A game(6) is a bimatrix game with fuzzy payoffs when N = {1, 2}. Determining the concept of optimality, we have to compare payoffs. We can used some defuzzification operator T (T (·) = Y (·), VP (·), U (·, ν) etc.). In [9] we proposed the following definition. Definition 4.1. [9] A strategy profile xe = (xe1 , ..., xeN ) ∈ X is called a T (·)-Nash equilibrium in the game (6) if fi (xe kxi ) T fi (xe ) (i ∈ N). We note that the solutions, which defined in [11], [5] and [6], are particular cases of Definition 4.1. Next, we consider the associated crisp game for (6) Γec = hN, {Xi }i∈N , {T (fei (x))}i∈N i. (7) Theorem 4.1. [9] Let xe is a Nash equilibrium in (7) and T (·) is a linear defuzzification operator, then xe is T (·)-Nash equilibrium in a game (6). For example, we consider one bimatrix game with a triangular fuzzy payoffs. Example 4.1. Consider a bimatrix game with fuzzy payoffs Γe of the form (A, e B), e where A e and Be are the triangular fuzzy matrixes:   (30, 6, 12) (10, 8, 6) (15, 10, 5) Ae =  (20, 10, 5) (22, 6, 10) (30, 5, 10)  (10, 8, 12) (30, 20, 4) (20, 8, 16) One Approach to Fuzzy Matrix Games 235 and   (10, 4, 6) (15, 10, 5) (20, 15, 4) e =  (20, 10, 10) (10, 6, 14) (15, 10, 5)  . B (12, 6, 10) (15, 10, 10) (10, 5, 15) The operator U (·, ν) is used. As a result, we obtain the associated crisp game (4). If ν = 0, then the game (4) given as follows     27 6 10 8 10 12, 5 A =  15 19 27, 5  , B =  15 7 10  . 6 20 16 9 10 7, 5 There are 3 pure and mixed U (·, 0)-Nash equilibrium strategy profiles. It is xe = (xe1 , xe2 ), where 10 ) xe1 = (0, 0, 1), xe2 = (0, 1, 0), 10 9 20 ) xe1 = ( 19 , 19 , 0), xe2 = ( 35 24 59 , 0, 59 ), 0 e 1 8 e 1 9 3 ) x1 = (0, 9 , 9 ), x2 = ( 10 , 10 , 0). If ν = 12 , then the associated crisp game (4) given as follows     31, 5 9, 5 13, 75 10, 5 13, 75 17, 25 A =  18, 75 23 31, 25  , B =  20 12 13, 75  . 11 26 22 13 15 12, 5 There are 3 U (·, 21 )-Nash equilibrium strategy profiles. It is xe = (xe1 , xe2 ), where 10 ) xe1 = (0, 0, 1), xe2 = (0, 1, 0), 25 27 70 51 20 ) xe1 = ( 52 , 52 , 0), xe2 = ( 121 , 0, 121 ), 0 e 1 4 e 12 31 3 ) x1 = (0, 5 , 5 ), x2 = ( 43 , 43 , 0). If ν = 1, then the game (4) given as follows     36 13 17, 5 13 17, 5 22 A =  22, 5 27 35  , B =  25 17 17, 5  . 16 32 28 17 20 17, 5 There are 3 U (·, 1)-Nash equilibrium strategy profiles. It is xe = (xe1 , xe2 ), where 10 ) xe1 = (0, 0, 1), xe2 = (0, 1, 0), 5 6 20 ) xe1 = ( 11 , 11 , 0), xe2 = ( 35 27 62 , 0, 62 ), 0 e 3 8 e 10 13 3 ) x1 = (0, 11 , 11 ), x2 = ( 23 , 23 , 0). Another example: one zero-sum matrix game with a trapezoidal fuzzy payoffs is considered . Example 4.2. Let A e be the trapezoidal fuzzy payoff matrixes of the fuzzy zero- sum matrix game Γ , which is given as follows: e   Ae = (20, 30, 12, 8) (1, 5, 8, 4) . (5, 9, 20, 4) (10, 26, 8, 12) The operator Y (·) is used. As a result, we obtain the associated crisp game (4)   24 2 A= . 3 19 236 K. Kudryavtsev et al. 8 11 The mixed Y (·)-Nash equilibrium is xe = (xe1 , xe2 ), where xe1 = ( 19 , 19 ), xe2 = 17 21 ( 38 , 38 ). 5 Fuzzy Matrix Game with Different Preferences In this section, we consider a two-person zero sum game Γea = h{1, 2}, {Xi }i=1,2 , fe1 (x)i, (8) where {1, 2} is the set of players’ serial numbers; each player i chooses and applies his own pure strategy xi ∈ Xi ⊆ Rni (i = 1, 2), a strategy profile is x = (x1 , x2 ) ∈ X = X1 × X2 ⊂ Rn (n = n1 + n2 ); a payoff function of player 1 is fe1 (x) : X → F. A payoff function of player 2 is fe2 (x) = −fe1 (x). In addition, let Xi contains only a finite number of elements. In this case, Γea is a two-person zero sum matrix game with fuzzy payoffs. This game is determined by a fuzzy matrix A. e In the last section, we considered that the players prefer the same defuzzi- fication operator. However, the players can have the different preferences. For example, it can be caused by the various attitude to the risk. In this case, the players can use the different defuzzification operators. The main idea of this paper is following: Suppose that the player 1 has decided to use a defuzzification operator T1 (·). And the player 2 chose to use a defuzzification operator T2 (·) (T1 (·) 6= T2 (·)). Definition 5.1. A strategy profile x∗ = (x∗1 , x∗2 ) ∈ X is called a T1 (·)T2 (·)- saddle-point in the game (8) if fe1 (x1 , x∗2 ) T1 fe1 (x∗1 , x∗2 ) T2 fe1 (x∗1 , x2 ) ∀ x1 ∈ X1 , x2 ∈ X2 . Next, we consider the associated crisp game for (8) Γa = h{1, 2}, {Xi }i=1,2 , {T1 (fe1 (x)), −T2 (fe1 (x))}i. (9) In contrast to (8), a crisp game (9) is a bimatrix game. Usually, the solution of crisp game (9) is Nash equilibrium. But, a set of Nash equilibriums X e is internally instable. We will use PONE in this game. Theorem 5.1. Let x∗ is a Pareto-optimal Nash equilibrium in (9) and T1 (·), T2 (·) are a linear defuzzification operators, then x∗ is T1 (·)T2 (·)-saddle-point in a game (8). For example, we consider zero-sum matrix game with a trapezoidal fuzzy payoffs from Example 4.2.. Example 5.1. Let A e be the trapezoidal fuzzy payoff matrixes of the fuzzy zero- sum matrix game Γ , which is given as follows: e   (20, 30, 12, 8) (1, 5, 8, 4) . (5, 9, 20, 4) (10, 26, 8, 12) The operator Y (·) is used for player 1, and the operator U (·, 0) is used for player 2. One Approach to Fuzzy Matrix Games 237 As a result, we obtain the associated crisp bimatrix game (9) given as follows     24 2 −14 3 A= , B= . 3 19 5 −6 The Y (·)U (·, 0)-saddle-point is x∗ = (x∗1 , x∗2 ), where x∗1 = ( 11 17 ∗ 17 21 28 , 28 ), x2 = ( 38 , 38 ). 6 Conclusion In this paper we proposed a method for formalizing and constructing equilib- rium in fuzzy matrix games to generalize some already known methods. In the future, we will apply it for formalizing a Berge equilibrium [19] and a coalition equilibrium [20] in n-person games with fuzzy payoffs. The case of continuous game is also if great interest. In the case, when we construct equilibrium in pure strategies, the linearity condition of a defuzzification operator is not required. We plan to study a continuous games case. Acknowledgement. The work was supported by Act 211 Government of the Russian Federation, contract N 02.A03.21.0011 and by Grant of the Foundation for perspective scientific researches of Chelyabinsk State University (2018). References 1. Baldwin, J.F., Guild, N.C.F.: Comparison of fuzzy sets on the same decision space. Fuzzy sets and Systems 2(3), 213–231 (1979) 2. Basiura, B., Duda, J., Gawel, B., Opila, J., Pelech-Pilichowski, T., Rebiasz, B., Skalna, I.: Ordering of fuzzy numbers. In: Advances in Fuzzy Decision Making. pp. 27-48. Springer International Publishing (2015) 3. Butnariu, D.: Fuzzy games: a description of the concept. Fuzzy sets and systems 1(3), 181–192 (1978) 4. Campos, L.: Fuzzy linear programming models to solve fuzzy matrix games. Fuzzy Sets and Systems 32(3), 275–289 (1989) 5. Cunlin, L., Qiang, Z.: Nash equilibrium strategy for fuzzy non-cooperative games. Fuzzy Sets and Systems 176(1), 46–55 (2011) 6. Dutta, B., Gupta, S.K.: On Nash equilibrium strategy of two-person zero-sum games with trapezoidal fuzzy payoffs. Fuzzy Information and Engineering 6(3), 299–314 (2014) 7. Ibanez, L. M. C., Munoz A.G.: A subjective approach for ranking fuzzy numbers. Fuzzy sets and systems 29(2), 145–153 (1989) 8. Jain, R.: A procedure for multiple-aspect decision making using fuzzy sets. Inter- national Journal of systems science 8(1), 1–7 (1977) 9. Kudryavtsev, K.N., Stabulit, I.S., Ukhobotov, V.I.: A Bimatrix game with fuzzy payoffs and crisp game. In: CEUR Workshop Proceedings. OPTIMA-2017. vol. 1987, pp. 343–349 (2017) 10. Larbani, M.: Non cooperative fuzzy games in normal form: A survey. Fuzzy Sets and Systems 160(22), 3184–3210 (2009) 238 K. Kudryavtsev et al. 11. Maeda, T.: On characterization of equilibrium strategy of two-person zero-sum games with fuzzy payoffs. Fuzzy Sets and Systems 139(2), 283–296 (2003) 12. Nash, J.F.: Equilibrium points in N-Person games. Proceedings of the National Academy of Sciences U.S.A. 36(1), 48–49 (1950) 13. Nayak, P.K., Pal, M.: The bi-matrix games with interval payoffs and its Nash equilibrium strategy. Journal of Fuzzy Mathematics 17(2), 421-435 (2009) 14. Seikh, M.R., Nayak, P.K., Pal, M.: Solving bi-matrix games with pay-offs of trian- gular intuitionistic fuzzy numbers. European Journal of Pure and Applied Mathe- matics 8(2), 153–171 (2015) 15. Ukhobotov, V.I., Mikhailova, E.S.: Comparison of fuzzy numbers in decision- making problems. Vestnik Udmurtskogo Universiteta. Matematika. Mekhanika. Komp’yuternye Nauki 26(1), 87–94 (2016), (in Russian) 16. Verma, T., Kumar, A.: Ambika methods for solving matrix games with Atanassov’s intuitionistic fuzzy payoffs. IEEE Transactions on Fuzzy Systems 26(1), 270–283 (2018) 17. Yager, R. R.: A procedure for ordering fuzzy subsets of the unit interval. Informa- tion sciences 24(2), 143-161 (1981) 18. Zhukovskiy, V.I., Kudryavtsev, K.N.: Pareto-optimal Nash equilibrium: Sufficient conditions and existence in mixed strategies. Automation and Remote Control 77(8), 1500-1510 (2016) 19. Zhukovskiy, V.I., Kudryavtsev, K.N.: Mathematical Foundations of the Golden Rule. I. Static Case. Automation and Remote Control 78(10), 1920-1940 (2017) 20. Zhukovskiy, V.I., Kudryavtsev, K.N.: Coalition equilibrium in a three-person game. In: 2017 Constructive Nonsmooth Analysis and Related Topics (dedicated to the memory of V.F. Demyanov) (CNSA), St. Petersburg, 2017. pp. 1-4. (2017). https: //doi.org/10.1109/CNSA.2017.7974037 21. Zimmermann, H.J.: Fuzzy Set Theory – and its Applications. Kluwer Academic Publishers (1996)