Free Essay

Nash

In:

Submitted By mharoamorin
Words 4568
Pages 19
Lecturas Matem´ticas a Volumen 24 (2003), p´ginas 137–149 a

John Nash y la teor´ de juegos ıa
Sergio Monsalve Universidad Nacional de Colombia, Bogot´ a Al profesor y acad´mico Don Jairo Charris Casta˜ eda e n In memoriam

Abstract.

In the last twenty years, game theory has become the dominant model in economic theory and has made significant contributions to political science, biology, and international security studies. The central role of game theory in economic theory was recognized by the awarding of the Nobel Price in Economic Science in 1994 to John C. Harsanyi, John F. Nash, & Reinhard Selten. The fundamental works in game theory of John F. Nash together with a brief exposition of them are included in this article.

Key words and phrases. John Nash, History of Mathematics, Game Theory 1991 Mathematics Subject Classification. Primary 01A70. Secondary 91A12. Resumen. En los ultimos veinte a˜os, la teor´ de juegos se ha ´ n ıa convertido en el modelo dominante en la teor´ econ´mica y ha ıa o contribuido significativamente a la ciencia pol´ ıtica, a la biolog´ ıa y a estudios de seguridad nacional. El papel central de la teor´ ıa de juegos en teor´ econ´mica fue reconocido con el premio Nobel ıa o en Econom´ otorgado a John C. Harsanyi, John F. Nash & ıa Reinhard Selten en 1994. Se presentan los aportes de John Nash a la teor´ de juegos conjuntamente con una exposici´n ıa o elemental de ellos.

138

SERGIO MONSALVE

1. Introduction
La Real Academia Sueca para las Ciencias le otorg´ el premio Nobel o en Ciencias Econ´miacs del a˜ o 1994 a los economistas John C. Haro n sanyi, Reinhard Selten y al matem´tico John F. Nash, debido a su a “an´lisis pionero de equilibrios en la teor´ de juegos no cooperativos”. a ıa La Academia justifica este premio en econom´ a tres de los “grandes” ıa en teor´ de juegos con el argumento de que esta ha probado ser muy util ıa ´ en el an´lisis econ´mico. 60 a˜ os despu´s de la publicaci´n de la obra a o n e o pionera de John von Neumann y Oskar Morgenstern (Theory of Games and Economic Behavior (1944)), la teor´ de juegos hab´ reıa ıa cibido el merecido reconocimiento como herramienta fundamental del an´lisis econ´mico moderno y el aporte de John Nash fue fundamena o tal.

2. ¿Qu´ es la teor´ de juegos? e ıa
La teor´ de juegos (o teor´ de las decisiones interactivas es el estudio ıa ıa del comportamiento estrat´gico cuando dos o m´s individuos interact´an e a u y cada decisi´n individual resulta de lo que ´l (o ella) espera que los o e otros hagan. Es decir, qu´ debemos esperar que suceda a partir de las e interacciones entre individuos.

3. ¿Con qu´ estructuras estudiamos la teor´ de juegos? e ıa
Existen, fundamentalmente, dos formas distintas de aproximarnos al an´lisis de una situaci´n de interacciones entre individuos a o I) La primera (que es quiz´s la dominante dentro del ambiente de a los economistas) es la teor´ de juegos no cooperativos, en la que, b´siıa a camente, tenemos un conjunto de jugadores, cada uno con estrategias a su disposici´n, y unas asignaciones de pagos que reciben por llevar o a cabo tales estrategias. La caracter´ ıstica “no cooperativa” est´ en la a manera de c´mo eligen y en lo que saben de los otros jugadores cuando o

JOHN NASH Y LA TEOR´ DE JUEGOS IA

139

eligen: en general, se supone que los individuos toman sus decisiones independientemente unos de otros aunque conociendo sus oponentes y las posibles estrategias que estos tienen a su disposici´n. Es decir, son o individuos ego´ ıstas pero que tratan de predecir lo que los otros agentes har´n para obrar entonces en conveniencia propia. En esta estructura a de an´lisis los agentes no alcanzan ning´ n nivel de cooperaci´n. a u o Nada mejor que un ejemplo bien ilustrativo del modus operandi de este tipo de modelos. Y quiz´s el m´s elocuente de los juegos no-cooperativos a a elementales es el dilema del prisionero. La historia de este juego va como sigue: dos individuos son detenidos debido a que cometieron cierto delito. Ambos son separados en celdas diferentes y son interrogados individualmente. Ambos tienen dos alternativas: cooperar uno con otro (no-confesar) o no cooperar (confesar el delito). Ellos saben que si ninguno confiesa, cada uno ir´ a prisi´n por dos a˜os. Pero si uno de los a o n dos confiesa y el otro no, entonces al que confiesa lo dejar´n libre y al a que no confiesa lo condenar´n a 10 a˜os. Si ambos confiesan, los dos a n a o n o ir´n a prisi´n por 6 a˜os. La situaci´n se resume en la siguiente bimatriz (es decir, una matriz cuyos elementos son parejas n´ meros): u Prisionero 1 C C Prisionero 2 NC (−2, −2) (0, −10) NC (−10, 0) (−6, −6)

C = cooperar (no confesar), N C = cooperar(confesar)

La pregunta natural es: ¿qu´ har´n los detenidos? ¿cooperar´n entre e a a s´ (no confesar´n) o se traicionar´n el uno al otro (confesar´n)? Alı a a a guien desprevenido que est´ observando este juego podr´ pensar que e ıa los dos jugadores cooperar´ (no confesar´n) puesto que en ese caso ıan a ambos obtendr´ el menor castigo posible. Sin embargo, la estructura ıan no cooperativa del problema hace que este arreglo no sea cre´ ıble: si se pactara la no-confesi´n por parte de los dos, ambos tendr´ incentivos o ıan particulares para romperlo, pues dejando al otro en cumplimiento del

140

SERGIO MONSALVE

pacto de no confesar y ´ste confesando, el que rompe el pacto obtiene e la libertad mientras al otro lo condenar´n a 10 a˜os. Y, similarmena n te, estudiando las otras tres posibilidades del juego (es decir, (C, N C), (N C, N C), (N C, C)) observamos que el unico acuerdo cre´ (que sig´ ıble nifica que ninguno de los dos querr´ romper el pacto unilateralmente ıa porque perder´ es (N C, N C). En definitiva, la predicci´n de lo que ıa) o ocurrir´ en el juego es que ambos confesar´n y permanecer´n en la carcel a a a 6 a˜os. n La conclusi´n en situaciones similares a ´sta (que son comunes en la o e vida diaria) es que la competencia ego´ puede conducir a estados que ısta son inferiores (en t´rminos de beneficio personal y social) a los estados e cooperativos, pero que estos ultimos no podr´n implementarse a menos ´ a que existan reforzmientos externos (contratos firmados por ley, con verificaci´n, etc.) que obliguen a las partes a cumplir con el acuerdo de o cooperaci´n. o Esta es la idea esencial de Nash al definir el concepto de equilibrio de Nash en su tesis doctoral en Matem´ticas en la Universidad de Princeton a (Non-cooperative Games (1950)): un equilibrio de Nash de un juego es un acuerdo que ninguna de las partes puede romper a discreci´n sin perder. o Es decir, si alguien quiere romper el pacto y lo hace unilateralmente, se arriesga a ganar por debajo de lo que hubiese ganado dentro del pacto. Sin embargo, como queda claro en el juego del dilema del prisionero, esto puede no ser lo mejor socialmente para los jugadores. Uno de los resultados que hacen del equilibrio de Nash un punto a de referencia para casi todo an´lisis en el que las interacciones entre individuos est´n involucradas es que e todo juego finito (es decir, finitos jugadores y finitas estrategias de cada jugador) tiene al menos un equilibrio de Nash, aunque involucre ciertas probabilidades objetivas de juego de las estrategias por parte de los jugadores. Este resultado es del mismo Nash. En un art´ ıculo previo a su tesis de doctorado, y publicado tambi´n en 1950 (Equilibrium points in e n−person games) ´l prueba, utilizando un teorema de punto fijo (el coe nocido teorema de punto fijo de Brouwer), que un equilibrio de Nash es

JOHN NASH Y LA TEOR´ DE JUEGOS IA

141

un “punto fijo”: las expectativas de los gentes con respecto a lo que los dem´s har´n, coinciden, todas, en el equilibrio de Nash. a a II) La segunda estructura fundamental para el estudio de la teor´ ıa de juegos para desde all´ predecir resultados de la interacci´n, es la ı o teor´ de juegos cooperativos o coalicionales. Aqu´ todav´ tenemos los ıa ı ıa mismos agentes ego´ ıstas, pero ahora se asume que, si pueden obtener alg´n beneficio de la cooperaci´n, no dudar´n en formar coaliciones que u o a son cre´ ıbles. Por supuesto, bajo una estructura como la de juegos no cooperativos, un acuerdo de cooperaci´n puede no ser la “soluci´n”, o o de manera que los agentes deben tener una estructura de informaci´n o diferente si queremos un comportamiento acorde. En una estructura cooperativa tenemos el mismo conjunto de jugadores ego´ ıstas, solo que ahora tienen informaci´n sobre cierta valoraci´n o o a priori de las coaliciones. Es decir, se reconoce cu´les coaliciones son a las m´s “valiosas” y cu´les las “menos valiosas”. Para concretar ideas, a a vamos a presentar un modelo que muestra bien el poder predictivo que puede tener este tipo de estructura y las soluciones asociadas. El modelo del peque˜o mercado muestra una estructura de mercado en la que n hay un vendedor (jugador 1) de cierto bien que no podemos dividir en partes (pi´nsese, por ejemplo, en un autom´vil como bien de uso), y dos e o compradores (jugador 2 y jugador 3) que desean comprar ese bien. Las a valoraciones que a priori se le asignan a las coaliciones ser´n, en este caso, un reflejo del ´xito o fracaso de la negociaci´n entre el vendedor e o y el comprador dependiendo de c´mo se emparejen, y no una predico ci´n sobre qui´n de los dos obtendr´ el bien (este problema tiene otra o e a valoraci´n). En este caso, asignamos la valoraci´n a todas las posibles o o coaliciones de la siguiente forma: V ({1, 2, 3}) = V ({1, 2}) = V ({1, 3}) = 1
(“si hay vendedor y comprador, el negocio se lleva a cabo”)

V ({1}) = V ({2}) = V ({2, 3}) = 0
(“si solo hay compradores o vendedor, no se realiza el negocio”)

Existen varias “soluciones” a ese tipo de juegos en forma cooperativa. Por supuesto, una “soluci´n” debe significar una repartici´n de la riqueza o o

142

SERGIO MONSALVE

o valoraci´n total de todo el grupo de jugadores como gran coalici´n, de o o tal manera que a cada jugador le corresponda su “aporte” a ella. C´mo o determinar el nivel de este aporte es algo que analizamos con las dos m´s importantes soluciones a juegos cooperativos: el n´cleo y el valor a u de Shapley. a) El n´cleo del juego cooperativo (Gillies (1953)) u La idea de repartici´n detr´s de la soluci´n de n´cleo es que si esta le o a o u asigna x1 al vendedor, x2 al comprador-jugador 2, y x3 al compradorjugador 3, entonces debemos, por lo menos, tener que: x1 + x2 + x3 = 1 (ef iciencia) (i)

(los tres agentes se reparten el “poder” del mercado que est´, a priori, a en la uni´n de los tres). o V ({1, 2}) ≤ x1 + x2 , V ({1, 3}) ≤ x1 + x3 , V ({2, 3}) ≤ x2 + x3 , V ({1}) ≤ x1 , V ({2}) ≤ x2 , V ({3}) ≤ x3 (“lo que los agentes reciben a trav´s de la asignaci´n de n´cleo es mejor e o u que lo que recibir´ antes de la asignaci´n a priori V , y esto sucede en ıan o cualquier coalici´n que formen”). o Obs´rvese que la asignaci´n de n´cleo es un acuerdo muy b´sico. Es e o u a una “invitaci´n” a formar coaliciones: si usted forma alguna coalici´n o o ganar´ m´s de lo que gana en cualquier coalici´n del status quo detera a o o minado por la asignaci´n a priori V . Un poco de aritm´tica nos muestra que la unica asignaci´n de n´cleo e ´ o u en nuestro peque˜o mercado es x1 = 1, x2 = x3 = 0. En otras palabras, el n n´cleo est´ pronosticando que la importancia total del mercado la tiene u a el vendedor y que los compradores no tienen niguna. Pobre predicci´n: o no existe casi ning´n mercado en el que los compradores no tengan u ninguna importancia estrat´gica. e (ii)

JOHN NASH Y LA TEOR´ DE JUEGOS IA

143

b) El valor de Shapley del juego cooperativo (Shapley (1953) Otra forma de distribuci´n con consideraciones m´s sut´ que las de o a ıles n´cleo, es el valor de Shapley. Podr´ decirse que el valor de Shapley es u ıa a los juegos cooperativos, lo que el equilibrio de Nash es a los juegos no cooperativos. Como ya hab´ ıamos descrito con el ejemplo anterior, un juego cooperativo consiste en un conjunto de jugadores N (= N = n) y una asignaci´n o monetaria V (S) para cada subcoalici´n S ⊆ N. El problema que se ino tenta resolver es: ¿C´mo distribuir la riqueza total V (N ) entre todos o los participantes? El valor de Shapley busca solucionarlo imponiendo ciertas condiciones a la distribuci´n: o o ıa o Si xi , i ∈ N , es la asignaci´n que recibir´ en la distribuci´n de Shapley el jugador i, entonces a) (Eficiencia) x1 + x2 + ...xn = V (N ) b) (Jugador “dummy” o fantasma) Si para alg´n i ∈ N , V (S ∪ {I}) = u V (S) para toda coalici´n S, entonces xi = 0 o c) (Simetr´ Si las valoraciones de las coaliciones no cambian cuando ıa) se reemplaza un jugador por cualquier otro, entonces, todos reciben lo mismo. Es decir, x1 = ... = xn . d) (Aditividad) Si V y W son dos valoraciones distintas sobre el mismo conjunto N de jugadores, entonces la asignaci´n de cualquier o jugador para la valoraci´n V y para la valoraci´n W es aditiva. Es o o decir, para todo i ∈ N , xi (V + W ) = xi (V ) + xi (W ) Solo el valor de Shapley satisface estas cuatro condicions. Y m´s a a´n, existe una f´rmula expl´ u o ıcita sobre c´mo calcularlo, basada en las o “contribuciones marginales” (es decir, una medida de cu´nto un jugador a le aporta a las coaliciones en donde est´): a (∆i (Si (r))) (∗) n! donde es el conjunto de los n! ordenes de N , Si (r) = conjunto de ´ jugadores que preceden a i en el orden r, ∆i (S) = V (S ∪ {i}) V (S) = contribuci´n marginal del agente i a la coalici´n S. o o xi = r∈ 144

SERGIO MONSALVE

Como un ejemplo de la “mejor” distribuci´n total de la riqueza del o juego entre los participantes que el valor de Shapley realiza, es f´cil a calcular (con la f´rmula (*) de arriba) que en el juego del Peque˜o o n Mercado que analiz´bamos antes, el valor de Shapley es a x1 = 2/3, x2 = 1/6, x3 = 1/6. A diferencia del n´cleo, cuya asignaci´n es (1, 0, 0), el valor de Shapley u o s´ est´ asign´ndole importancia a los compradores, aunque reconoce un ı a a poder mayor al vendedor. IIa) Una muy importante categor´ en el estudio de la teor´ de juegos ıa ıa cl´sica es, realmente, una subdivisi´n de la teor´ de juegos cooperatia o ıa vos: los modelos de negociaci´n. En estos juegos, dos o m´s jugadores o a buscan ganar a trav´s de la cooperaci´n, pero deben negociar el procedie o miento y la forma en que se dividir´n las ganancias de esta cooperaci´n. a o Un modelo de negociaci´n espec´ o ıfica: c´mo y cu´ndo se alcanzan los o a acuerdos y c´mo se dividir´n las ganancias, dependiendo de las reglas o a de negociaci´n y de las caracter´ o ısticas de los negociadores. John Nash realiz´ contribuciones fundacionales a la teor´ de jueo ıa gos de negociaci´n. En su art´ o ıculo de 1950, The bargaining problem, se aparta radicalmente de la teor´ econ´mica ortodoxa que consideraba inıa o determinados los problemas de negociaci´n. En contraste, Nash asume o que la negociaci´n entre agentes racionales conduce a un unico reultado, o ´ y lo determina imponi´ndole al modelo ciertas “propiedades deseables”. e La formulaci´n de Nash del problema de negociaci´n y su soluci´n (la o o o o o ıa soluci´n de negociaci´n de Nash) constituyen el fundamento de la teor´ moderna de la negociaci´n. o Para aclarar un tanto lo dicho arriba con una versi´n muy elemental o del problema de negociaci´n de Nash, supongamos que tenemos dos juo ıstas que buscan dividirse una cantidad de dinero M y que gadores ego´ no est´n de acuerdo en que los dos no obtengan nada de la negociaci´n, a o pero que si no llegan aun acuerdo, entonces s´ les corresponder´ “irse ı a con las manos vac´ ıas”. John Nash mostr´ que bajo ciertas condicioo nes plausibles, los dos jugadores acordar´ la repartici´n (x1 , x2 ) que ıan o maximiza, en este caso, el producto de las dos. Es decir, resuelven M´x x1 x2 , a sujeto a la condici´n x1 + x2 = M. o

JOHN NASH Y LA TEOR´ DE JUEGOS IA

145

La soluci´n a esto es, como se ve f´cilmente, (x1 )∗ = (x2 )∗ = 1 M . Esta o a 2 es la soluci´n de negociaci´n de Nash a este simple problema: distrio o buci´n equitativa de la cantidad de dinero disponible. Esta soluci´n se o o ilustra en el diagrama de abajo. Por supuesto, problemas de negociaci´n o m´s complicados dan or´ a ıgen a soluciones de Nash menos obvias. grafico!!!! Aquellas “propiedades deseables” mencionadas arriba, que satisface la soluci´n de negociaci´n de Nash son las siguientes: o o a) Eficiencia: las soluciones de negociaci´n de Nash agotan todas las o oportunidades de mejorar las ganancias de ambos jugadores. b) Simetr´ si el cambio de un jugador por otro no cambia el probleıa: ma de negociaci´n, en la soluci´n ellos deben obtener lo mismo. o o c) Invarianza: la soluci´n no puede depender de las unidades arbitrao rias en las que se midan los beneficios de la negociaci´n. o d) Independencia de alternativas irrelevantes: es la hip´tesis m´s suo a til. Dice que el resultado de una negociaci´n no puede depender de o alternativas de negociaci´n que los negociadores no escogen aunque o pudieran hacerlo. Estas cuatro hip´tesis conducen al conoocido teorema de negociaci´n o o de Nash: La soluci´n de negociaci´n de Nash es la unica soluci´n o o ´ o que satisface las propiedades a), b), c), d) en un modelo de negociaci´n. o

3. El programa Nash: ¿Es mejor cooperaci´n que o competencia?
Nash aseguraba que la teor´ de juegos cooperativa y no cooperativa ıa eran complementarias, que cada una ayudaba a justificar y clarificar la otra. Si una soluci´n cooperativa pod´ ser obtenida a partir de un o ıa conjunto convincente de hip´tesis, esto indicaba que tambi´n pod´ ser o e ıa apropiada en una variedad de situaciones m´s amplia que las encontradas a en un modelo no cooperativo simple. Y de otro lado, si reformulamos

146

SERGIO MONSALVE

juegos cooperativos como no cooperativos y buscamos los equilibrios de Nash de estos ultimos, la discusi´n abstracta sobre lo razonables ´ o que puedan ser los principios o los resultados se reemplazan por una discusi´n m´s mundana acerca de lo apropiadas que son las reglas del o a juego. El programa Nash buscaba la posibilidad de la unificaci´n te´rica, y o o algunos logros ya se tienen en este sentido. Uno de ellos es el famoso teorema de Aumann (1975) que afirma, en t´rminos elementales, y a la e vez un tanto vagos, lo siguiente: En un grupo conformado por muchos agentes que tratan de disputarse cierta cantidad de dinero, las distribuciones de equilibrio de Nash, de n´cleo, de valor de Shapley y u o de negociaci´n de Nash coinciden. Aqu´ la palabra “muchos” intenta capturar la idea de que cada jugador, ı, por s´ mismo, tiene un poder estrat´gico pr´cticamente nulo. Solo tienen ı e a poder, para la asignaci´n, grupos verdaderamente “grandes”, no indivio duos aislados. Una implicaci´n directa de este teorema es que cuando o las interacciones individuales tienden a anularse, ¡¡la competencia y la cooperaci´n conducen a los mismos resultados!! o Otro modelo b´sico muy importante en la direcci´n del programa a o Nash es el de Rubinstein (1982) que muestra que cualquier resultado de negociaci´n Nash entre pocos agentes, puede aproximarse mediante o equilibrios de Nash de juegos no cooperativos secuenciales. Pero, por encima de esto, el gran impulso del programa Nash se recibi´ en 1994 o en el seminario Nobel sobre el trabajo de John Nash en la Universidad de Princeton (publicado en Les Prix Nobel 1994). Despu´s de estos, e se han obtenido resultados que involucran otros valores diferentes al de a Shapley en estructuras de juegos m´s sofisticadas que muestran que la afirmaci´n de Shapley en el sentido de que ambas teor´ (coopeo ıas rativa y no cooperativa) son, no solo complemento una de otra, sino que realmente son una sola teor´ vista con dos lentes diferentes, estaba ıa en el camino correcto (v´ase Hart & Mas Colell (1996), Hart & e Monsalve (2001)).

JOHN NASH Y LA TEOR´ DE JUEGOS IA

147

5. Otras contribuciones de Nash a la teor´ de juegos ıa
Se reconocen muy poco las contribuciones de John Nash a la primera literatura en econom´ experimental (v´ase Roth (1993)). Entre ıa e otras, Nash hace algunos comentarios importantes respecto a la cooperaci´n observada cuando repetimos infinitamente el dilema del prisioneo ro. Adem´s, el trabajo pionero de Nash, conjuntamente con Kalish, a Milnor y Nering sobre experimentos en juegos cooperativos (Kalish et al. (1954)) fue un importante est´ ımulo para Reinhard Selten en sus experimentos (Selten (1993)) que han conducido al florecimiento de la escuela alemana en econom´ experimental. ıa

6. Final
El par´grafo final de los peri´dicos anunciando el premio Nobel en a o Econom´ de 1994 dice: “a trav´s de sus contribuciones al an´lisis de ıa e a equilibrios en teor´ de juegos no cooperativos, los tres laureados consıa tituyen una combinaci´n natural: Nash dio los fundamentos para el o an´lisis, Selten los desarroll´ con respecto a la din´mica, y Harsanyi a o a con respecto a la informaci´n incompleta”. Esta es la unidad “natural” o de este premio. Sin embargo, n´tese que la teor´ de juegos cooperativos o ıa fue completamente ignorada e, indirectamente, tambi´n los aportes de e John Nash a esta teor´ incluyendo los resultados del Programa Nash. ıa, El razonamiento de la teor´ de juegos es ahora el fundamento de ıa importantes areas de la teor´ econ´mica, y est´ r´pidamente entrando ´ ıa o a a en disciplinas aparentemente dis´ ımiles como finanzas, ciencia pol´ ıtica, sociolog´ derecho y biolog´ Las contribuciones de John Nash (junıa, ıa. to con Harsanyi, Selten, Aumann & Shapley) constituyen importantes piedras angulares en el desarrollo de la teor´ de juegos y en el ıa establecimiento de una metodolog´ com´n para analizar la interacci´n ıa u o e estrat´gica dentro de todas las ciencias sociales, e incluso (este es el reto) en otras ciencias.

148

SERGIO MONSALVE

Referencias
[1] Aumann, R. J. (1975) Values of markets with a continuum of traders, Econometrica 43:611–646 [2] Aumann, R. J. (1977) Game Theory. In The New Palgrave: A Dictionary of Economics Vol. 2 ed J. Eatwell, M. Milgate, and P. Newman, 460–482. London: Macmillan. [3] Aumann, R. J. (1988) On the state of the art in game theory: and interviewwith Robert Aumann, Games and Economic Behavior 24:181–210. [4] Gillies, D. B. (1953) Some theorems on n−person games. Ph. D. Dissertation, Department of Mathematics, Princeton University. [5] Hart, S. & Mas Colell, A. (1996) Bargaining and Value, Econometrica 64: 357–380. [6] Hart, S. & Monsalve, S. (2001) The Asymptotic approach of the MaschlerOwen values Journal of Game Theory (forthcoming). [7] Kalish, G. K., Milnor, J. W., Nash, J. F., Nering, E. D. (1954) Some experimental n person games, Cap´ ıtulo 21: 301–327 En: Thrall R. M., Coombs, C. H., Davids, R. L. (eds) Decision Processes. Wiley, New York. [8] Mayberry, J. P., Nash, J. F. y Shubik, M. (1953) A comparison of treatments of a duopoly situation, Econometrica 21:141–155. [9] Monsalve, S. (1999) Introducci´n a los conceptos de equilibrio en econom´a. o ı Editorial Unibiblos, Santaf´ de Bogot´. e a [10] Monsalve, S. (2002) Teor´a de Juegos: ¿Hacia d´nde vamos? (60 a˜os ı o n despu´s de von Neumann y Morgenstern) Revista de Econom´ Institucional e ıa, No. 7 Universidad Externado de Colombia, Bogot´ (Pr´xima Publicaci´n). a o o [11] Myerson, R. (1991). Game Theory: Analysis of Conflict. Cambridge, MA: Harvard University Press. [12] Nash, J. F. (1950) Equilibrium points in n person games, Procedings from the National Academy of Sciences, USA 36:48–49. [13] Nash, J. F. (1950) The Bargaining Problem, Econometrica 18:155–162. [14] Nash, J. F. (1950) Non cooperative Games, Ph. D. Dissertation Princeton University. [15] Nash, J. F. (1953) Two person cooperative games, Econometrica 21:128– 140. [16] Roth, A. (1993) The Early history of experimental economics, Journal of History of Economic Thought 15: 184–209. [17] Rubinstein, A. (1982) Perfect equilibrium in a bargaining model, Econometrica 50:97–109 [18] Selten, R. (1993) In search of a a better understanding of economic behavior. En : Heertje A. (ed.) The Makers of modern economics I:115–139. Harvester Wheatsheaf Hertfordshire.

JOHN NASH Y LA TEOR´ DE JUEGOS IA

149

[19] Shapley, L. S. (1953) A value for n-person games. En: Contributions to the Theory of Games, vol 2, H. W. Kuhn y A. W. Tucker, eds. [20] Von Neumann, J. & Morgenstern, O. (1944) Theory of Games and Economic Behavior. Princeton, NJ: Princeton University Press.

(Recibido en junio de 2002; la versi´n revisada agosto de 2003) o

Sergio Monsalve e-mail: email: monsalvesergio@yahoo.com ´ Departamento de Matematicas, Universidad Nacional de Colombia ´ Bogota, Colombia

Similar Documents

Premium Essay

Nash

...Draft chapter from An introduction to game theory by Martin J. Osborne. Version: 2002/7/23. Martin.Osborne@utoronto.ca http://www.economics.utoronto.ca/osborne Copyright © 1995–2002 by Martin J. Osborne. All rights reserved. No part of this book may be reproduced by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from Oxford University Press, except that one copy of up to six chapters may be made by any individual for private study. 2 Nash Equilibrium: Theory 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 Strategic games 11 Example: the Prisoner’s Dilemma 12 Example: Bach or Stravinsky? 16 Example: Matching Pennies 17 Example: the Stag Hunt 18 Nash equilibrium 19 Examples of Nash equilibrium 24 Best response functions 33 Dominated actions 43 Equilibrium in a single population: symmetric games and symmetric equilibria 49 Prerequisite: Chapter 1. 2.1 Strategic games is a model of interacting decision-makers. In recognition of the interaction, we refer to the decision-makers as players. Each player has a set of possible actions. The model captures interaction between the players by allowing each player to be affected by the actions of all players, not only her own action. Specifically, each player has preferences about the action profile—the list of all the players’ actions. (See Section 17.4, in the mathematical appendix, for a discussion of profiles.) More precisely, a strategic game...

Words: 19938 - Pages: 80

Premium Essay

Nash Equilibrium

...In game theory, the Nash equilibrium is a solution concept of a non-cooperative game involving two or more players, in which each player is assumed to know the equilibrium strategies of the other players, and no player has anything to gain by changing only their own strategy.[1] If each player has chosen a strategy and no player can benefit by changing strategies while the other players keep theirs unchanged, then the current set of strategy choices and the corresponding payoffs constitute a Nash equilibrium. Stated simply, Amy and Will are in Nash equilibrium if Amy is making the best decision she can, taking into account Will's decision, and Will is making the best decision he can, taking into account Amy's decision. Likewise, a group of players are in Nash equilibrium if each one is making the best decision that he or she can, taking into account the decisions of the others in the game. Contents   [hide]  * 1 Applications * 2 History * 3 Definitions * 3.1 Informal definition * 3.2 Formal definition * 3.3 Nash's Existence Theorem * 4 Examples * 4.1 Coordination game * 4.2 Prisoner's dilemma * 4.3 Network traffic * 4.4 Competition game * 4.5 Nash equilibria in a payoff matrix * 5 Stability * 6 Occurrence * 6.1 Where the conditions are not met * 6.2 Where the conditions are met * 7 NE and non-credible threats * 8 Proof of existence * 8.1 Proof using the Kakutani fixed...

Words: 1390 - Pages: 6

Premium Essay

Nash Equilibrium

...A Nash equilibrium is a pair of strategies, one for each player, in which each strategy is a best response to other. These represent the likely outcome of the game. According to Roger B. Myerson, “If we can predict the behavior of all the players in such a game, then our prediction must be a Nash equilibrium, or else it would violate this assumption of intelligent rational individual behavior. That is, if our predicted behavior does not satisfy the conditions for Nash equilibrium, then there must be at least one individual whose expected welfare could be improved simply by re-educating him to more effectively pursue his own best interests, without any other social change.” The above argument does not prove that Nash equilibrium should be the only methodological basis for analysis of social institutions. But it does explain why studying Nash equilibria should be a fruitful part of the critical analysis of almost any kind of social institution. The prisoners dilemma is one of the best examples of Nash Equilibrium. | | Jack | | | | C | NC | Tom | C | -10,-10 | 0,-20 | | NC | -20,0 | -5,-5 | | | | | *Numbers represent years in prison If every player in a game plays his dominant pure strategy (assuming every player has a dominant pure strategy), then the outcome will be a Nash equilibrium.According to the above game both players know that 10 years is better than 20 and 0 years is better than 5; therefore, C is their dominant strategy and they will both...

Words: 405 - Pages: 2

Premium Essay

Nash Equilibrium

...this rebate into account, your marginal cost of a midsized automobile is $11,000. What price should you charge for a midsized automobile if you expect to maintain your record sales? 2) In a two-player, one-shot simultaneous-move game each player can choose strategy A or strategy B. If both players choose strategy A, each earns a payoff of $500. If both players choose strategy B, each earns a payoff of $100. If player 1 chooses strategy A and player 2 chooses strategy B, then player 1 earns $0 and player 2 earns $650. If player 1 chooses strategy B and player 2 chooses strategy A, then player 1 earns $650 and player 2 earns $0. a. Write the above game in normal form. b. Find each player's dominant strategy, if it exists. c. Find the Nash equilibrium (or equilibria) of this game. d. Rank strategy pairs by aggregate payoff (highest to lowest). e. Can the outcome with the highest aggregate payoff be sustained in equilibrium? Why or why not? a. Write the above game in normal form. | |Player 1 chooses A |Player 1 chooses B...

Words: 1598 - Pages: 7

Premium Essay

Nash Equilibrium Exercises

...Information k = units of Capital l = units of Labour rK = Cost of Capital in dollars = $20 wL = Labour cost in dollars = $10 Output: Q=K^(1/4) √L Solution Min: C_((K,L))=r_K .k+ w_L .l=20k+10l Constraint: Q=k^(1/4) √l Using Lagrangian method L_((k,l,λ))=20k+10l- λ(k^(1/4) .l^(1/2)-Q) dL/dk=20- 1/4 λk^(-3/4) .l^(1/2)=0 dL/dl=10- 1/2 λk^(1/4) .l^(-1/2)=0 dL/dλ=- 1/4 λk^(1/4) .l^(1/2)-Q=0 Equating 20- 1/4 λk^(-3/4) .l^(1/2)= 0 20= 1/4 λk^(-3/4) .l^(1/2) λ= (80k^(3/4))/l^(1/2) 10- 1/2 λk^(1/4) .l^(-1/2)=0 10= 1/2 λk^(1/4) .l^(-1/2) λ= (20l^(1/2))/k^(1/4) (80k^(3/4))/l^(1/2) = (20l^(1/2))/k^(1/4) 80k=20l 4k=l Substituting Lagrangian relation into Output equation Q=K^(1/4) √L Q=k^(1/4) 〖(4k)〗^(1/2) Q=k^(1/4) 〖2k〗^(1/2) Q/2=k^(3/4) k= Q^(4/3)/2^(4/3) = Q^(4/3)/∛16= Q^(4/3)/(2∛2) Solving for l 4k=l 2^2 (Q^(4/3)/2^(4/3) )=l l= Q^(4/3)/2^(2/3) = Q^(4/3)/∛4 Minimum total expenditure on capital and labour in terms of Q C_((K,L))=20k+10l C_((K,L) )=2^2.5.(Q^(4/3)/2^(4/3) )+10(Q^(4/3)/2^(2/3) ) C_((K,L) )= (5Q^(4/3)+10Q^(4/3))/2^(2/3) = 〖15Q〗^(4/3)/2^(2/3) = 〖2^2.5.Q〗^(4/3)/2^(2/3) = (5Q^(4/3))/∛2 Information x denotes units of good X y denotes units of good Y Cx denotes the unit cost of good X in dollars = $1 Cy denotes the unit cost of good Y in dollars = $1 0 < β < 1 M denotes the maximum...

Words: 1054 - Pages: 5

Free Essay

Ball Like Steve Nash

...Ball Like Steve Nash Standing at 6’3 and weighing in at 178 lbs most people know the infamous Steve Nash from the past fifteen years that he has dedicated his life to professional basketball. Before pursuing the American dream of becoming a professional athlete, Nash was born in South Africa but grew up in Victoria, Canada where he played soccer till his late teens. He later discovered a passion for basketball where he hooped at St. Michaels University School with his younger brother Martin. He comes from a family of athletes including his father John Nash who played minor professional soccer in South Africa. His sister Joann Nash held the title as captain of the soccer team for The University of Victoria for three years. It was only right for him to keep on the family legacy. Nash took his talent to Santa Clara University in California where he led his team to the NCAA tournament as a freshman. He established a name for himself by his sophomore year with his quick feet, fundamentals, and amazing skills. Nba scouts were all over him and at no surprise he was drafted in the first round by the Phoenix Suns in 1996. He played two seasons with the Suns before being traded to the Dallas Mavericks. There Nash made a name for himself in the league with Dirk Nowitzki and Michael Finley by his side. They formed an unstoppable trio that carried the team to the Western Conference Finals. Although they didn’t get much farther Nash participated in the NBA All-Star Game and was recognized...

Words: 640 - Pages: 3

Premium Essay

John Nash Research Paper

...MGF1106 “John Nash Biography” “I would not dare to say that there is a direct relation between mathematics and madness, but there is no doubt that great mathematicians suffer from maniacal characteristics, delirium and symptoms of schizophrenia”. – John F. Nash Jr. John Nash Jr. was born on June 13, 1928 in Bluefield, West Virginia. Mr. Nash was the older of the two children that his father, John Nash Sr. and mother Martha Martin had. John and his sister grew up in the Great Depression yet they were fortunate that their father was able to keep his job as an electrical engineer with Appalachian Power Company and they never suffered through that time. They actually lived in a beautiful house not very far from a country club, so the family never suffered in any way,...

Words: 813 - Pages: 4

Premium Essay

Case Study of John Forbes Nash, Jr.

...Running head: CASE STUDY OF JOHN FORBES NASH, JR. 1 Case Study of John Forbes Nash, Jr. Lauren Shipp PSY410 May 26, 2014 Kidd Colt, Ed. D., LMHC CASE STUDY OF JOHN FORBES NASH, JR. 2 Case Study of John Forbes Nash, Jr. John Forbes Nash Jr. is a renowned and accomplished mathematician. He received his Ph.D. from Princeton University and taught at MIT and Princeton. He wrote The Equilibrium Point, later becoming known as the Nash Equilibrium, which revolutionized economics. In 1994, he received the Nobel Peace Prize in Economic Science for his pioneering work in game theory. He is one of the most brilliant mathematicians of modern time, but most of his life he suffered from schizophrenia (Meyer, et al., 2009). The following is a brief account of a case study depicting his struggle with schizophrenia. Overview Early in Nash’s life he showed signs of abnormal behavior. He was extremely intelligent and could read by age 4, but was unsociable and had problems with concentrating and following simple directions. As he grew older, his behavior became more bizarre. He would do such things as eat grass, torture animals, and use explosives in chemical experiments. He still showed sign of unsocial behavior (Meyer, et al., 2009). When he entered Carnegie Institute of Technology to study chemical engineering, his abnormal behavior continued. He acted childish, and would do such things as repeatedly hit a single key on a piano for hours. After receiving his Ph.D. from Princeton...

Words: 1072 - Pages: 5

Premium Essay

Definitions and Illustrations of the Concepts of Pareo and Nash Equilibria

...Definitions and illustrations of the concepts of Pareo and Nash equilibria. PARETO OPTIMUM/EQUILIBIRUM/EFFICIENT The idea of pareto optimum runs through all aspects of economics, for example, comparing different tax rules, and not just in game theory. It gained currency as a test to be applied in selecting economic policy as a result of a contradiction in the theory of the market deriving from the philosophy of utilitarianism. Utilitarianism could be construed as a justification for the oppression of the individual in the name of the greater good of society. This approach sits uncomfortably with the libertarian ideals of market economics, which put the individual above church and the (feudal) state, and by implication above the monarch in direct challenge to feudal rule. (see, eg, http://www.google.co.uk/#hl=en&source=hp&biw=1126&bih=425&q=Shanti+Chakravarty+neoliberal&btnG=Google+Search&aq=f&aqi=&aql=&oq=Shanti+Chakravarty+neoliberal&fp=45f26fda8f9185dd). To get around the difficulty posed by classical utilitarianism, market efficiency theorems came to rely on the idea of ordinal utility which does not allow for inter-personal comparison of utility. The paretian criterion explicitly rejects inter-personal comparison in arriving at economic policy. The paretian criterion is focused entirely on the individual. A pareto optimum is a state of affairs whereby NO ONE, no individual, can be made better off without making someone worse off. Since no one is above anyone in the...

Words: 2188 - Pages: 9

Free Essay

Nash

...but as her “boarder.” With her taking him back as a boarder it made it to where he did not become homeless. He returned to Princeton and went back to his wicked ways, roaming the halls leaving cryptic messages all around the school’s math and physics buildings, on black board’s, wall and windows. He occasionally left numerology treatises. The students became wary of him and gave him the nick name “Phantom Of Fine Hall” Then in 1970, his delusions cleared up, and he became more aware of the world surrounding him. Although it was unclear to how he became better. In 1994, at the ripe age of 66, John Nash received The Nobel Prize in economics, for his work on game theory. Thirty eight years after his and Alicia’s divorce the decided to remarried. Nash got a job at Princeton, where he continues to study mathematics. In pure mathematics he is know best for what is sometimes called “Nash Embedding Theorm.” It is very long and confusing process. You can get a collage degree in math and not even skim over what he accomplished. It is very hard and long thought out process, which requires a lot of mental and physical ability....

Words: 412 - Pages: 2

Premium Essay

Nash Equalibrium

...team 1 and team 2 would invest money until they reach Nash equilibrium and their profits are maximized. Since this model is a symmetric model, both teams have the same incentive to win and therefore at equilibrium it can be assumed that win% for team 1 will equal the win% for team 2. Knowing these givens and the equation given, profits for team 1 can be found as followed: π1 = Vw1 – t1 → V(2-g)/4 π1 = (500)(2-.5)/4 π1 = 187.5 The profit can then be plugged back into the initial equation to determine team 1’s investment: π1 = Vw1 – t1 The win percentages for both teams will be .5 because .5 + .5 = 100% in a symmetric model 187.5 = 500(.5) – t1 t1= 62.5 Since this is a symmetric model, it is assumed that the investments and profits for team 1 will equal team 2’s investments and profits. π1 = π2 187.5 = π2 t1 = t2 62.5 = t2 2. For model 1, profits are larger when g=1/2 because the investments made by both teams are lower and therefore π1 = Vw1 – t1 will result in higher profits. The reason team 1 and team 2’s investments will be lower is because there is less incentive to invest when the advantage for investing goes down, as it does with a lower sensitivity parameter. A lower g (sensitivity parameter) means that teams need to investment less to yield a higher win%. The following graph shows the difference in profits for g=1 and g= ½: 3. To calculate the Nash equilibrium for an asymmetric model win % and team...

Words: 2149 - Pages: 9

Free Essay

David Nash Art

...1. David Nash was born in Esher, Surrey in 1945. He studied at Kingston College of Art (1963-64), Brighton College of Art (1964-67) and Chelsea School of Art (1969-70). After finishing school at Brighton, Nash moved to North Wales before returning to Chelsea for one year in 1969. In Wales he purchased a chapel which has remained both his studio and home ever since. Wales was ideal for Nash, because he was surrounded by abundant resources and plenty of time to further develop his wood works. 2. The most interesting thing I discovered about David Nash was that he uses a chainsaw and axe as his primary tools as well as using a blowtorch in many of his works.David Nash uses wood as a medium for all his works. His interest in working with wood began as a child, when Nash helped clear and replant a forest his father owned. He also worked for the Commercial Forestry Group, where he learned about many kinds of wood. He is known for doing land art involving wood that remains in nature, as well as displaying his wood sculpture in studios. He carves wood from fallen trees as well as creates sculptures from growing plants. 3. Nash first decided to move to Wales after graduating from Brighton College due to the extremely low cost of living. In these years he experimented with making tower like sculptures and some very abstract works. He used paint to give more detail to these sculptures as well, which was unique to this period for Nash. He continued experimenting with this tower theme...

Words: 938 - Pages: 4

Free Essay

John Nash Jr

...JOHN FORBES NASH JR. John Forbes Nash Jr. was born June 13, 1928 in Bluefield, West Virginia. Mr. Nash Jr. is an American mathematician who won the 1994 Nobel Prize for his works in the late 1980’s on game theory. Game theory is the study of strategic decision making or more formally known as the mathematical models of conflict and cooperation between intelligent and rational decision makers. Game theory is mainly used in economics, political science, and psychology, as well as logic and biology. Mr. Nash Jr. has also contributed numerous publications involving differential geometry, and partial differential equation (PDE). Differential geometry is a mathematical discipline that uses differential calculus and integral calculus, linear algebra and multi linear algebra to study geometry problems. Partial differential equation is a differential equation that contains unknown multivariable functions and their partial derivatives. These are used to formulate problems involving functions of several variables. Mr. Nash Jr. used all of these skills and is known for developing the Nash embedding theorem. The Nash embedding theorem stated that every Riemannian manifold ( a real smooth manifold equipped with an inner product on each tangent space that varies smoothly from point to point) can be isometrically embedded into some Euclidean space ( a three dimensional space of Euclidean geometry, distinguishes these spaces from the curved spaces...

Words: 528 - Pages: 3

Premium Essay

Kate Nash Thesis

...It is no secret that during the post-Cold War era there was a wave of skepticism that arose and questions about moving forward. After the Cold War there was a period of stillness where people wondered if there was going to be another global up rise. With WWI, WWII, then the Cold War, many people were left uncertain about the future. One of those people was active scholar Kate Nash, a professor of Sociology at Goldsmiths College, University of London and faculty of the Center for Cultural Sociology at Yale University. In The Cultural Politics of Human Rights Nash talks about why she is skeptical of the ability of global solidarity movements to address rights issues including: poverty, gender inequity, and structural violence. Nash also gives...

Words: 328 - Pages: 2

Premium Essay

Essentials of Game Theory

...Essentials of game theory 1. Introduction Game theory is the study of strategic decision making. More formally, it is "the study of mathematical models of conflict and cooperation between intelligent rational decision-makers."[1] An alternative term suggested "as a more descriptive name for the discipline" is interactive decision theory.[2] Game theory is mainly used in economics, political science, and psychology, as well as logic and biology. The subject first addressed zero-sum games, such that one person's gains exactly equal net losses of the other participant(s). Today, however, game theory applies to a wide range of class relations, and has developed into an umbrella term for the logical side of science, to include both human and non-humans, like computers. Classic uses include a sense of balance in numerous games, where each person has found or developed a tactic that cannot successfully better his results, given the other approach. Modern game theory began with the idea regarding the existence of mixed-strategy equilibria in two-person zero-sum games and its proof by John von Neumann. Von Neumann's original proof used Brouwer's fixed-point theorem on continuous mappings into compact convex sets, which became a standard method in game theory and mathematical economics. His paper was followed by his 1944 book Theory of Games and Economic Behavior, with Oskar Morgenstern, which considered cooperative games of several players. The second edition of this book provided...

Words: 4437 - Pages: 18