Weak Assumption and Iterative Admissibility

Size: px
Start display at page:

Download "Weak Assumption and Iterative Admissibility"

Transcription

1 Weak Assumption and Iterative Admissiility Chih-Chun Yang Institute of Economics, Academia Sinica, Taipei 115, Taiwan January 2014 Astract Brandenurger, Friedenerg, and Keisler [Econometrica , ] show that rationality and common assumption of rationality RCAR is impossile in a complete and continuous type structure. We show, y inroducing an alternative notion of weak assumption, that rationality and common weak assumption of rationality RCWAR is possile in a complete and continuous type structure. This possiility result provides an epistemic characterization for iterative admissiility. Keywords: Iterative admissiility, weak assumption, common weak assumption of rationality. This paper is ased on the first chaper of the author s thesis, University of Rochester, I am indeted to my adviser, Paulo Barelli, for his valuale comments and encouragement. I thank Adam Brandenurger, Yi-Chun Chen, Amanda Friedenerg, Tai-Wei Hu, Takashi Kunimoto, Xiao Luo, Romans Pancs, Marciano Siniscalchi, Satoru Takahashi, Gáor Virág, Yongchao Zhang and participants in seminars at Academia Sinica, National Taiwan University, National University of Singapore, Shanghai University of Finance and Economics, University of Rochester, the 2010 Workshop on Epistemic Game Theory Stony Brook, and 2012 SAET conference. Financial support from the National Science Council of Taiwan is gratefully acknowledged. Fax: cyang16@econ.sinica.edu.tw.

2 1 Introduction Samuelson 1992 points out that there is a tension etween admissiility which requires the inclusion all possile strategies in a elief and knowledge/elief aout admissiility which requires the exclusion of some strategies from consideration. Brandenurger, Friedenerg and Keisler 2008a henceforth, BFK provide an epistemic analysis for iterative admissiility henceforth IA y using lexicographic proaility systems LPS s. BFK define the notion of assumption, a counterpart of knowledge, in the model constructed y LPS s. BFK show that rationality and common assumption of rationality RCAR is equivalent to the solution concept of self-admissile set, a counterpart of the rationalizale set. In particular, the IA set is a self-admissile set. However, if a player is not indifferent in every outcome, RCAR is empty in a complete and continuous type structure. BFK s impossiility result implies that RCAR fails to provide epistemic foundations of iterative admissiility. That is, if one keeps the continuity assumption, then one has to drop completeness to otain a nonempty RCAR. But then one has to provide the exact epistemic reason for the particular form of incompleteness. Keisler and Lee 2011 prove that in a finite normal form game, there is a complete ut not continuous lexicographic structure where a state elongs to RCAR if and only if the strategy part of the state is iteratively admissile. In particular, Keisler and Lee 2011 construct the type space as a Cantor space {0, 1} N, the set of all infinite sequences of 0 s and 1 s with the product topology. Their work suggests that the continuity assumption, coupled with full-support LPS s, yields a strong notion of cautious ehavior. Barelli and Galanis 2013 adapt a different approach. They propose a new notion of rationality, namely event-rationality, in a standard universal type structure. They characterize IA as the outcome of event-rationality and common validated elief of eventrationality RCvBER. The purpose of this paper is to investigate IA and RCAR y modifying the notion of assumption in a complete and continuous type structure. BFK use two axioms to define the notion of assumption: Nontriviality and Strict Determination. By keeping Strict Determination and modifying Nontriviality, we propose an alternative notion of weak assumption. The modified Nontriviality axiom comes from the idea of the consistency check in BFK This modification is designed to accommodate the consistency issue y requiring weaker conditions such that we can avoid the impossiility result: 1

3 Instead of using an open set, we opt to use a strategy cylinder as a significant event. We define rationality and common weak assumption of rationality RCWAR and prove that RCWAR is nonempty in a complete and continuous type structure. The main result of this paper is the outcome equivalence etween RCWAR and IA. Hence, we provide an epistemic foundation for iterative admissiility. The rest of this paper is organized as follows. Section 2 reviews BFK s result and defines weak assumption. Section 3 formulates the notion of RCWAR and shows the main result. The proofs of minor results are relegated to the Appendix. 2 The Model We follow the model in BFK. Let S a, S, π a, π e a two-player finite strategic form game, where S a, S are the strategy sets and π a, π are the payoff functions. Extend the payoff functions to mixed strategies in the usual way. Let X Y S a S. A strategy s a X is admissile with respect to X Y if there is no mixed strategy σ a satisfying σ a X = 1 such that for each s Y, π a σ a, s π a s a, s and for some s Y, π a σ a, s > π a s a, s. Iterative Admissiility IA: For i = a,, let S i 0 = S i and define inductively S i n+1 = { s a S i n : s i is admissile with respect to S a n S n}. A strategy s i n=1s i n s n=1s n is called iteratively admissile. Let M Ω e the space of Borel proaility measures on a separale metric space Ω. Let N Ω e the set of all finite sequences of Borel proaility measures on Ω. Let µ, µ N Ω. If the lengths of µ and µ are different, then the distance is 1. Otherwise, the distance etween µ and µ is the maximum of the Prohorov distances etween µ i and µ i for all i. Write supp µ i for the support of µ i M Ω. Say µ = µ 0,..., µ n 1 N Ω is a lexicographic proaility system LPS if µ satisfies mutual singularity: There are Borel sets U i in Ω for each i = 0,..., n 1 such that µ i U i = 1 and µ i U j = 0 if j i. Say µ N Ω is a full-support sequence if Ω is the union of the supports of µ i. We write L Ω for the set of LPS s and L + Ω for the set of full-support LPS s. 2

4 An S a, S -ased type structure is a structure S a, S, T a, T, λ a, λ, where the type spaces T a and T are nonempty Polish spaces and λ a, λ are Borel measurale functions: λ a : T a N S T and λ : T N S a T a. A type structure S a, S, T a, T, λ a, λ is complete if L + S T range λ a and L + S a T a range λ. It is continuous if λ a and λ are continuous. Each type of a player is associated with a sequence of proaility measures on the opponent s strategies and types. Let the notion L represent the lexicographic order. That is, for each a 1,..., a n, 1,..., n R n, a 1,..., a n L 1,..., n iff whenever a i < i, there is j < i such that a i > i. A strategy s a is optimal under µ = µ 0,..., µ n 1 N S T if π a s a, marg S µ i s n 1 L π a r a, marg S µ i s n 1 for all r a S a. A strategy-type pair s a, t a is rational if s a is optimal under λ a t a and λ a t a is a full-support LPS. Let R a e the set of rational strategy-type pairs s a, t a. The full support requirement says that no event is thought completely impossile y a rational player. A rational player must e so cautious that all his opponent strategies, eliefs over strategies and higher order eliefs are taken into account. This definition of rationality is designed to meet the condition of admissiility: It is well known that a strategy in X S a is admissile with respect to X Y S a S if and only if it is a est response in X to a elief with full support in Y. 2.1 BFK s result BFK define the notion of assumption, a counterpart of knowledge, in the model constructed y LPS s. Let A e the set of measurale functions from Ω to [0, 1]. For each 3

5 µ = µ 0,..., µ n 1 L + Ω, define a preference µ on A as follows. For each x, y A, define x µ y if x i ω dµ i ω ω Ω n 1 L ω Ω n 1 y i ω dµ i ω. They use two axioms on µ to define the notion of assumption. Let e a complete, transitive, reflexive inary relation, satisfying Independence axiom, on A. For each x, y A and E Ω, define x E, y Ω\E A if xe, y Ω\E = x on E and x E, y Ω\E = y on Ω\E. Let E Ω. An event E 0 is a part of E if there is an open set U Ω such E 0 = E U. An event E is assumed under if E is Borel and satisfies the following conditions: i Nontriviality: E is nonempty and for each part of E, E 0, there are acts x, y A such that x E 0 y; ii Strict Determination: For each act x, y A, x E y implies x y. Strict Determination says that the consequences in the event E are determining for. Nontriviality says that for each significant event U, an open suset of Ω, if U E, then U E is not Savage-null. Nontriviality is designed to accommodate the consistency check in BFK 2008: If a player steps into the analyst s shoes, then his elief and ehavior should e consistent with the analyst s prediction. BFK define that an event E is infinitely more likely than F under a full-support LPS µ. Say an event E is assumed under µ if E is infinitely more likely than Ω\E under µ. BFK show that an event E is assumed under µ if and only if E is assumed under µ. Define A a E {t a T a : E is assumed under λ a t a }. Define A E analogously. Let R a 1 R a and R 1 R. For each n 1, define R a n+1 and R n+1 inductively y R a n+1 R a n [ S a A a R n] and R n+1 R n [ S A R a n ]. If s a, s, t a, t n=1r a n n=1r n, then we say there is Rationality and Common Assumption of Rationality RCAR at this state. BFK show that for each integer n, proj S ar a n proj S R n = S a n S n. 4

6 Unfortunately, RCAR is empty in a natural class of games. Fix a complete and continuous type structure S a, S, T a, T, λ a, λ. If player a is not indifferent in every outcome, then there is no state at which there is RCAR. In fact, n=1r a n n=1r n =. 2.2 Weak Assumption Let Ω = S T. For each s S, let [s] {s} T. We define the notion of weak assumption as follows. Definition 1. An event E Ω is weakly assumed under if E is Borel and satisfies the following conditions: i Nontriviality : E is nonempty and for each s proj S E, there are acts x, y A such that x E [s] y; ii Strict Determination: For each act x, y A, x E y implies x y. We keep BFK s Strict Determination. Our Nontriviality is also designed to accommodate the consistency check. Our approach is ased on the idea that if the analyst s prediction, in particular, on the strategy space, is availale to a player, then the player s elief and ehavior should e consistent with this prediction. What if Ann steps into the analyst s shoes? Specifically, let us now focus on the strategies that can e played i.e., on the space S and imagine that the analyst s prediction is availale to Ann.... There, ecause Ann is confident of her prediction {L, C}, she must e more confident that Bo plays in accordance with the prediction and, in particular, plays C than that Bo violates his prediction and plays R. BFK 2008, p.1-2 Nontriviality is then modified to reflect the idea that the player is more confident of the nonempty intersection of the prediction and a strategy cylinder than its complement: Instead of using a nonempty open set, we opt to use a strategy cylinder as a significant event. Nontriviality says that for each significant event U, a strategy cylinder, if U E, then U E is not Savage-null. This modification helps us to avoid the impossiility result. 5

7 We conclude this section y the characterization of weak assumption. Lemma 1. Let µ = µ 0,..., µ n 1 L + Ω. An event E Ω is weakly assumed under µ if and only if E is Borel and there is 0 j n 1 such that 1 for each i j, µ i E = 1 and for each i j + 1, µ i E = 0; 2 proj S E = i j proj S supp µ i. If µ is a full-support LPS and an event E Ω is weakly assumed under µ, then we say E Ω is weakly assumed under µ. 3 Rationality and Common Weak Assumption of Rationality We consider the universal type spaces, T a and T, constructed in a way similar to Mertens and Zamir 1985, Brandenurger and Dekel 1993, and Heifetz Formally, ased on Heifetz 1993, we have the following result. Lemma 2. There are separale metric spaces T a and T such that T a is homeomorphic to N S T and T is homeomorphic to N S a T a. Let λ a e the homeomorphism etween T a and N S T. Likewise, define λ. Let E S T. Define W A a E {t a T a : E is weakly assumed under λ a t a }. Define W A in similar way. Let R 1 a R a and R 1 R. For each n 1, define R n+1 a and R n+1 inductively y R n+1 a R ] n a [S a W A a Rn and R n+1 R ] n [S W A Ran. By Lemma 3 in the Appendix, each R a n R n is Borel. If s a, s, t a, t n=1 R a n n=1 R n, then we say there is Rationality and Common Weak Assumption of Rationality RCWAR at this state. The main result of this paper is that RCWAR is outcome equivalent to IA in a complete and continuous type structure. Theorem 1. Let S a, S, T a, T, λ a, λ e the complete and continuous type structure defined in this section. For each n, proj s a Ra n proj s R n = S a n S n. 6

8 Moreover, proj S a Ra proj S R = n=1s a n n=1s n. In terms of strategies, IA and RCWAR are identical. RCWAR implies that a player chooses a strategy in IA. On the other hand, for each strategy s in IA, there is RCWAR at a state where s is chosen. The main idea of the proof is to show that there is a fixed s a, t a in each R a n. Since the game is finite, there is a finite m such that S a m = n=1s a n and S m = n=1s n. By Lemma 4 in the Appendix, Ω a \ R m is dense in Ω a. This fact help us to construct a full-support LPS λ a t a = µ 0, µ 1,..., µ m L + m+1 S T such that µ 0 has a finite support in R m and for each i 0, µ i has only atoms in Ω a \ R m. The notion of weak assumption is flexile enough such that for each n m, R m is weakly assumed under λ a t a. Proof of Theorem 1. We prove Theorem 1 in steps. Step 0. proj s a Ra 1 proj R s 1 = S1 a S1. This step is identical to BFK Theorem 9.1 and hence, omitted. Step 1 For each n m, proj s a Ra n proj R s n = Sn a Sn. We prove it y induction. Suppose that for each k n 1, proj R S k = S k and proj S a Ra k = Sk a. Step 1.1. Let s a R n. a We will find a type t a T a such that s a, t a R n a and hence, proj s a Ra n proj R s n Sn a Sn. By Lemma E1 in BFK, for each i {0,..., m 1}, for each s a Sn, a there is ν sa i M S such that supp ν sa i = S i and s a is a est response to ν sa i. Let U n = Ω a \ R 1, U 1 = R n 1 and for each i {2,..., n 1}, let U i = R n i\ R n i+1. By Lemma 3, each U i is Borel. By Lemma 2 separaility of T, each U i has a countale dense suset D i. By assigning positive weight to each point in D i, we may construct an LPS µ 1,..., µ n, where µ i M S T such that marg S µ i = ν sa n i, µ i U i = 1 and supp µ i is the closure of U i. By Lemma 2, there is t a T a such that λ a t a = µ 1,..., µ n. By the construction of µ, µ is a full-support LPS and s a, t a is a rational pair. Moreover, y Lemma 1 and Lemma 3, for each i {1,..., n 1}, R i is weakly assumed under µ. Therefore, s a, t a R a n. Step 1.2. Let s a, t a R a n. We will show that s a S a n and hence, proj s a Ra n proj s R n S a n S n. By hypothesis, and Lemma 1, there is j m such that proj S R n 1 = S n 1 = i j supp marg S µ i. Since s a, t a R a 1, s a is optimal under λ a t a, i.e., π a s a, marg S µ i s m L π a r a, marg S µ i s m for all ra S a. 7

9 As in Lemma 7.1 in BFK, there is ν M S such that supp ν = Sn 1 and s a is a est response to ν. Thus, s a is admissile with respect to S a Sn 1. By hypothesis, s a Sn 1. a Therefore, s a Sn a and proj s a Ra n+1 Sn+1. a Step 2. proj S a Ra proj R S = n=1sn a n=1 Sn. Since R n a and R n are shrinking, y Step 1, proj S a Ra proj R S Sm a Sm. It remains to show that for each s a Sm, a there is a type t sa T a such that s a, t sa n=1 Ra n. Step 2.1. Let R 0 a = Ω and R 0 = Ω a. For each s a Sm, a we will find a full-support LPS µ sa 1,..., µ m sa L + Ω a such that for each i {1,..., m}, 1 s a is a est response to each marg S µ sa 2 i j=1 supp µ j is the closure of R m i, and 3 µ i R m = 0. By Lemma E1 in BFK, for each s a Sm, a there is ν sa 0,..., ν sa m M S... M S such that for each i, supp ν sa i = S m i and s a is a est response to ν sa i. Consider a suset of T a : { T a t a T a : λ a t a = µ 0, µ 1,..., µ m L m+1 S T } {, s a Sm, a k 2, s.t. s S,ν } sa 0 s = marg Ω a µ { k 0 s, t } for some t Tk 1. By Lemma 4, T a T a \ T a is dense in T a and Ω S a T a \ R m a is dense in Ω. Likewise, define T a, T a, and Ω a. Let U m = Ω a \ R 1 and for each i {1,..., m 1}, let U i = Ω a R m i \ R m i+1. By Lemma 2, T is separale and each U i has a countale dense suset D i. Similar to Step 1 1.1, y assigning positive weight to each point in D i, for each i {1,..., m}, we may construct µ sa i M S T such that marg S µ i = ν sa i, µ i U i = 1 and supp µ i is the closure of U i. Since m i=1 D i is dense in Ω a, which is dense in Ω a, for each i {1,..., m}, i j=1 supp µ j is the closure of R m i and m j=1 supp µ j = Ω a. Hence, since {D i } m 1 i=1 are disjoint, µ s a 1,..., µ sa m L + Ω a. Likewise, for each s S m, define that i, µ s 1,..., µ s m L + Ω. Step 2.2. For each s a S a m and s S m, we will find t sa T a and t s T such λ t a sa = µ s a sa = µ 0, µ sa 1,..., µ m sa L + Ω a and λ t s = µ s = µ s 0, µ s 1,..., µ s m L + Ω, 8

10 where marg S µ sa 0 = ν sa 0 s.t. µ sa 0 marg S a µ s 0 = ν s 0 s.t. µ s 0 sa µ 0, µ sa 1,..., µ sa s S m { s, µ s} = 1, and s a S a m { s a, µ sa } = 1. Let λ t a sa = µ s a = m. Let margω a k µsa = marg Ω a k µ sa 0,..., marg Ω a k µ m sa. To construct t sa T a, we will specify each of the marginal distriutions marg Ω a k µ sa i. For each i 0 and k 0, let marg Ω a k µ sa i marg Ω a k µ sa i, i.e., µ sa i = µ sa i. Let marg Ω a 1 µ sa 0 ν sa 0. Likewise define each µ s 0. Inactively, for each k 1 define marg Ω a k+1 µ sa 0 M S Tk and marg Ω k+1 µ s 0 M S a Tk a as follows: For each s Sm and s a Sm, a marg Ω a k+1 µ sa 0 marg Ω k+1 µ s 0 s, marg Ω 1 µ s,..., marg Ω k µ s = ν sa 0 s and s a, marg Ω a 1 µsa,..., margω a k µ sa = ν s 0 s a. That is, in each k + 1-th step of constructing µ sa 0, we use the set { s, marg Ω 1 µ s,..., marg Ω k µ s } : s Sm as a finite support of marg Ω a k+1 µ sa 0. By the definition of T and the construction of µ sa, µ sa L + Ω a. By Lemma 2, there is t sa T a such that λ t a sa = µ s a. Likewise } define each t s. Moreover, y the construction of µ sa 0 and Lemma 2, µ sa 0 s S {s, t m s = 1. Likewise for. Thus, each s a i, t sa Ra 1 and each s i, t s R 1. By induction, Lemma 1 and Lemma 3, for each n 2, Ra n R n is weakly assumed under λ t sa λ t a s. Therefore, each s a, t sa s, t s survives in each round of elimination. That is, s a, t sa n=1 Ra n and s i, t s R n=1 n. In defining rationality, as in BFK, we require full support, which says that no event is thought completely impossile y a rational player. If we require full support only on the marginal proailities on strategies, we still have the outcome equivalence etween IA and RCWAR. 1 Appendix Proof of Lemma 1. If part. Let E Ω such that the conditions 1 and 2 are satisfied. We show that Nontriviality and Strict Determination are satisfied. 1 This is the main difference etween this paper and Catonini I thank Amanda Friedenerg for this oservation. The proof in this paper is still valid for this weak version of rationality. 9

11 Nontriviality. By 1, E. Let s proj S E. We want to show that there are acts x, y A such that x µ E [s] y. By 2, there is i j such that s proj S supp µ i, i.e., µ i [s] > 0. By 1, µ i E = 1 and hence, µ i E [s] = µ i [s] > 0. Therefore, there are acts x, y A such that x µ E [s] y. Strict Determination. Consider acts x, y A satisfying x µ E y, i.e., n 1 n 1 x i ω dµ i ω > L y i ω dµ i ω. ω E By 2, for each i > j, x ω E i ω dµ i ω = y ω E i ω dµ i ω = 0. Hence, we have j j x i ω dµ i ω > L y i ω dµ i ω ω E By 1, for each i j, x ω E i ω dµ i ω = x ω Ω i ω dµ i ω and y ω E i ω dµ i ω = ω Ω y i ω dµ i ω. Thus, x ω Ω i ω dµ i ω j >L y ω Ω i ω dµ i ω j. Therefore, ω Ω x i ω dµ i ω n 1 >L y ω Ω i ω dµ i ω n 1, i.e., x µ y. Only If part. Let E Ω e weakly assumed under µ. We have to show that conditions 1 and 2 are satisfied. Suppose, in negation, that 1 is not satisfied. That is, either a there are i, j such that i < j, µ i E = 0 and µ j E > 0 or there is j such that µ j E 0, 1. In case a, let c = 1 and in case, let c 0, 1 µ j E µ j E ω E ω E.. For each real numer r [0, 1], let r e the constant act assigning r to each state in Ω. By mutual singularity, there are {U i } 0 i n 1 such that for each i, µ i U i = 1 and for each j i, µ i U j = 0. Now, let x c Uj E0 Ω\Uj E and y 0 E 1 Ω\E. However, we have x µ E y and y µ x, contradicting Strict Determination. Given that 1 is satisfied, we show that proj S E = i j proj S supp µ i. Suppose not. Thus, there is s proj S E\ i j proj S supp µ i. Note that for each i j, µ i [s] E µ i [s] = 0. For each i > j, y 1, µ i [s] E µ i E = 0. Therefore, [s] E is null, contradicting Nontriviality. Proof of Lemma 2. We first construct the universal type spaces: T a and T. Let Ω a 1 = S and Ω 1 = S a. We then define the first order elief spaces as T1 a = N S and T1 = N S a. For each metric space A B and each µ = µ 0,..., µ n 1 N A B, define marg A µ = marg A µ 0,..., marg A µ n 1. Inductively, for each k 1 define Ω a k+1 = S T k and T a k+1 = { δ 1,..., δ k, δ k+1 T a k N Ω a k+1 : margω a k δk+1 = δ k}. 10

12 Likewise for. Let T a and T e the projective limits of the spaces {Tk a} k=1 and { Tk}, k=1 respectively. Note that T a and T are constructed to meet coherency requirement: Different order of eliefs shouldn t contradict one another. Let { δ k} {T a k=1 k } k=1 such that δ1 = δ 1 0,..., δn 1 1 N S. For each i {0,..., n 1}, let δ i { } δ k { {δ i and k=1 a k } {T a k=1 k } k=1 : δ1 M S }. Then for each i {0,..., n 1}, δ i a. Suppose that Y, d is a separale metric space. We may have a complete metric space Y, d such that Y is a completion of Y and d preserves identical metric on Y. For the formal proof, see Kolmogorov and Fomin 1970, Theorem 4, page 62. Hence, Y, d is Polish. We refer to the metric space Y without specifying the metric d. Since Ω a 0 is a Polish, the completion T1 a is also Polish. Hence, T1 a, as a suspace of the separale metric space T1 a, is also separale. By induction, each Tk a is also a separale metric space. So is T a. Hence, each proaility measure on Ω a k is tight Parthasarathy 1967, Chapter II, Theorem 3.2 and regular Parthasarathy 1967, Chapter II, Theorem 3.1 in the sense of Heifetz Thus, y Theorem 8 in Heifetz 1993, for each i {0,..., n 1}, there exists a unique µ i M S T such that for each k 1, marg Ω a k µ i = δ k i. Therefore, µ µ1,..., µ n 1 N S T. On the other hand, for each µ = µ 0,..., µ n 1 N S T, each µ i M S T. Since every metric space is normal, y Theorem 5* of Heifetz 1993, the topology used in Definition 2 of Heifetz 1993 is weak topology Billingsley 1968, Appendix III, which can e metrizale y Prohorov metric. Hence, y Theorem 9 in Heifetz 1993, a is homeomorphic to M S T. Likewise, define, which is homeomorphic to M S a T a. By the homeomorphism etween a and M { S T, each margω a µ k i} k=1 a. Therefore, under the topology generated y Prohorov metric in this paper, T a is homeomorphic to N S T. Lemma 3. For each natural numer n, R n a is Borel in S a T a. Proof. Let N + Ω e the set of full-support sequences of proaility measures, N n Ω { µ N Ω : µ = } µ 0,..., µ n 1, Ln Ω L Ω N n Ω and L + n Ω L + Ω N n Ω. By separaility and BFK s proof for Lemma C.2 and Corollary C.1, for each natural numer n, the sets N n Ω, N n + Ω, L n Ω and L + n Ω are Borel and the sets N + Ω, L Ω and L + Ω are also Borel. Let j n e a natural numer and E Ω e a Borel set. By the proof of BFK for Lemma C.3, the set of µ = µ 0,..., µ n 1 in L + n Ω such that condition 1 in Lemma 1 holds is Borel. Note that, given condition 1, condition 2 in Lemma 1 11

13 is equivalent to the following statement: For each s a proj S ae, there is i j such that marg µ i {s a } T a > 0. Thus, y fact ii in BFK s Appendix C, the set of µ = µ 0,..., µ n 1 in L + n Ω such that E is weakly assumed under µ is Borel. Similar to the proof of BFK for Lemma C.4, R n a is Borel in S a T a. Lemma 4. Let µ 1,..., µ m L m S T and ν sa 0 M S. Consider a suset of T a : { T a t a T a : λ a t a = µ 0, µ 1,..., µ m L S T } {, s a Sm, a k 2, s.t. s S,ν } sa 0 s = marg Ω a µ { k 0 s, t } for some t Tk 1. Then T a T a \ T a is dense in T a. Also, Ω \ R a m is dense in Ω. Thus, Ω S a T a \ R a m is dense in Ω. Proof. To prove this, y Lemma 2, it remains to show that for each t a T a, we may find a sequence {t n } T a such that lim n t n = t a. Suppose that λ a t a = µ 0, µ 1,..., µ m. Let X S T e the set of atoms of µ 0,..., µ m. Since T is uncountale and X is countale, S T \X. By Lemma 2, there is t T a such that λ a t = µ 0, µ 1,..., µ m and µ 0 µ 0 is a degenerate p.m. with mass at a point in S T \X. Then the sequence of {t n } n=1 can e found y letting λ a t n = 1 1 µ n 0 + 1n µ 0, µ 1..., µ n 1. Also, Ω \ R m a is dense in Ω. To prove this, y Lemma 2, it remains to show that for each s a, t a R m, a we may find a sequence {t n } T a such that lim n t n = t a and each s a, t n Ω \ R m. a Suppose that λ a t a = µ 0, µ 1,..., µ k L + k Ωa. Let X S T e the set of atoms of µ 0,..., µ k. Since Ω a \ R 1 is uncountale and X is countale, Ω a \ R 1 \X. By Lemma 2, there is t T a such that λ a t = µ 0, µ 1,..., µ k and µ 0 is a degenerate p.m. with mass at a point in can e found y letting λ a t n = Ω a \ R µ n 0 + 1n µ 0, µ 1..., µ k. \X. Then the sequence of {t n } n=1 Each s a, t n Ω \ R m a ecause, y Lemma 1, R 1 is not weakly assumed under λ a t n. Therefore, Ω S a T a \ R m a is dense in Ω. 12

14 References [1] Barelli, P., and S. Galanis, Admissiility and Event-Rationality, Games and Economic Behavior, , [2] Billingsley, P., Convergence of Proaility Measures. New York: Wiley [3] Brandenurger, A., and E. Dekel, Hierarchies of Beliefs and Common Knowledge, Journal of Economic Theory, , [4] Brandenurger, A., A. Friedenerg and H.J. Keisler, Admissiility in Games, Econometrica a, [5] Brandenurger, A., A. Friedenerg and H.J. Keisler, Supplement to Admissiility In Games: Appendices, Econometrica Supplementary Material, [6] Catonini, E., Common assumption of cautious rationality and iterated admissiility, Unpulished Manuscript [7] Heifetz, A., The Bayesian Formulation of Incomplete Information The Non-Compact Case, International Journal of Game Theory, , [8] H. Jerome Keisler and Byung Soo Lee, Common Assumption of Rationality, Unpulished Manuscript [9] Kolmogorov, A.N., and S.V. Fomin, Introductory Real Analysis, New York: Dover [10] Mertens, J.-F., and S. Zamir, Formulation of Bayesian Analysis for Games with Incomplete Information, International Journal of Game Theory, , [11] Parthasarathy, K. R., Proaility Measures on Metric Spaces, San Diego: Academic [12] Samuelson, L., Dominated Strategies and Common Knowledge, Games and Economic Behavior, ,

Lexicographic Beliefs and Assumption

Lexicographic Beliefs and Assumption Lexicographic Beliefs and Assumption Eddie Dekel Amanda Friedenberg Marciano Siniscalchi May 8, 2014 1 Introduction Lexicographic beliefs (henceforth l-beliefs) have become a relatively standard tool,

More information

Lexicographic Beliefs and Assumption

Lexicographic Beliefs and Assumption Lexicographic Beliefs and Assumption Eddie Dekel Amanda Friedenberg Marciano Siniscalchi May 5, 2016 Abstract Foundations for iterated admissibility (i.e., the iterated removal of weakly dominated strategies)

More information

Iterated Dominance Revisited

Iterated Dominance Revisited Iterated Dominance Revisited Amanda Friedenberg H. Jerome Keisler August 14, 2011 1 Introduction Iterated deletion of strongly dominated strategies has a long tradition in game theory, going back at least

More information

Forward induction reasoning revisited

Forward induction reasoning revisited Theoretical Economics 7 (2012), 57 98 1555-7561/20120057 Forward induction reasoning revisited Pierpaolo Battigalli Department of Economics, Bocconi University Amanda Friedenerg Department of Economics,

More information

Fixed Points in Epistemic Game Theory

Fixed Points in Epistemic Game Theory Fixed Points in Epistemic Game Theory Adam Brandenburger, Amanda Friedenberg, and H. Jerome Keisler Abstract. The epistemic conditions of rationality and common belief of rationality and rationality and

More information

Weak Robust (Virtual) Implementation

Weak Robust (Virtual) Implementation Weak Robust (Virtual) Implementation Chih-Chun Yang Institute of Economics, Academia Sinica, Taipei 115, Taiwan April 2016 Abstract We provide a characterization of (virtual) implementation in iterated

More information

Levels of Knowledge and Belief Computational Social Choice Seminar

Levels of Knowledge and Belief Computational Social Choice Seminar Levels of Knowledge and Belief Computational Social Choice Seminar Eric Pacuit Tilburg University ai.stanford.edu/~epacuit November 13, 2009 Eric Pacuit 1 Introduction and Motivation Informal Definition:

More information

Comprehensive Rationalizability

Comprehensive Rationalizability Comprehensive Rationalizability Aviad Heifetz Martin Meier Burhard C. Schipper September 6, 2018 Abstract We present a new solution concept for strategic games called comprehensive rationalizability that

More information

COMMON ASSUMPTION OF RATIONALITY

COMMON ASSUMPTION OF RATIONALITY COMMON ASSUMPTION OF RATIONALITY H. Jerome Keisler Byung Soo Lee 2011-08-22 Abstract In this paper, we provide an epistemic characterization of iterated admissibility (IA), i.e., iterated elimination of

More information

HIERARCHIES OF BELIEF AND INTERIM RATIONALIZABILITY

HIERARCHIES OF BELIEF AND INTERIM RATIONALIZABILITY HIERARCHIES OF BELIEF AND INTERIM RATIONALIZABILITY JEFFREY C. ELY AND MARCIN PESKI Abstract. In games with incomplete information, conventional hierarchies of belief are incomplete as descriptions of

More information

CRITICAL TYPES. 1. Introduction

CRITICAL TYPES. 1. Introduction CRITICAL TYPES JEFFREY C. ELY AND MARCIN PESKI Abstract. Economic models employ assumptions about agents infinite hierarchies of belief. We might hope to achieve reasonable approximations by specifying

More information

Epistemic Game Theory: Language and Observation

Epistemic Game Theory: Language and Observation Epistemic Game Theory: Language and Observation Adam Brandenburger NYU Stern School of Business NYU Polytechnic School of Engineering NYU Shanghai October 4, 2015 Theory of Mind in Tasks Gallagher, H.

More information

WEAKLY DOMINATED STRATEGIES: A MYSTERY CRACKED

WEAKLY DOMINATED STRATEGIES: A MYSTERY CRACKED WEAKLY DOMINATED STRATEGIES: A MYSTERY CRACKED DOV SAMET Abstract. An informal argument shows that common knowledge of rationality implies the iterative elimination of strongly dominated strategies. Rationality

More information

Consistent Beliefs in Extensive Form Games

Consistent Beliefs in Extensive Form Games Games 2010, 1, 415-421; doi:10.3390/g1040415 OPEN ACCESS games ISSN 2073-4336 www.mdpi.com/journal/games Article Consistent Beliefs in Extensive Form Games Paulo Barelli 1,2 1 Department of Economics,

More information

Bayesian coalitional rationality and rationalizability

Bayesian coalitional rationality and rationalizability Bayesian coalitional rationality and rationalizability Xiao Luo a, Yongchuan Qiao a, Chih-Chun Yang b a Department of Economics, National University of Singapore, Singapore 117570 b Institute of Economics,

More information

WHY SATURATED PROBABILITY SPACES ARE NECESSARY

WHY SATURATED PROBABILITY SPACES ARE NECESSARY WHY SATURATED PROBABILITY SPACES ARE NECESSARY H. JEROME KEISLER AND YENENG SUN Abstract. An atomless probability space (Ω, A, P ) is said to have the saturation property for a probability measure µ on

More information

Events Concerning Knowledge

Events Concerning Knowledge Events Concerning Knowledge Edward J. Green Department of Economics The Pennsylvania State University University Park, PA 16802, USA eug2@psu.edu DRAFT: 2010.04.05 Abstract Aumann s knowledge operator

More information

Higher Order Beliefs in Dynamic Environments

Higher Order Beliefs in Dynamic Environments University of Pennsylvania Department of Economics June 22, 2008 Introduction: Higher Order Beliefs Global Games (Carlsson and Van Damme, 1993): A B A 0, 0 0, θ 2 B θ 2, 0 θ, θ Dominance Regions: A if

More information

Part III. 10 Topological Space Basics. Topological Spaces

Part III. 10 Topological Space Basics. Topological Spaces Part III 10 Topological Space Basics Topological Spaces Using the metric space results above as motivation we will axiomatize the notion of being an open set to more general settings. Definition 10.1.

More information

Qualitative analysis of common belief of rationality in strategic-form games

Qualitative analysis of common belief of rationality in strategic-form games Qualitative analysis of common belief of rationality in strategic-form games Giacomo Bonanno University of California, Davis Elias Tsakas Maastricht University October 2, 2017 Abstract In this paper we

More information

Algorithms for cautious reasoning in games

Algorithms for cautious reasoning in games Algorithms for cautious reasoning in games Geir B. Asheim a Andrés Perea b October 16, 2017 Abstract We provide comparable algorithms for the Dekel-Fudenberg procedure, iterated admissibility, proper rationalizability

More information

Epistemic game theory

Epistemic game theory Epistemic game theory Eddie Dekel and Marciano Siniscalchi Tel Aviv and Northwestern University, and Northwestern University January 14, 2014 Contents 1 Introduction and Motivation 2 1.1 Philosophy/Methodology.................................

More information

Merging and splitting endowments in object assignment problems

Merging and splitting endowments in object assignment problems Merging and splitting endowments in oject assignment prolems Nanyang Bu, Siwei Chen, and William Thomson April 26, 2012 1 Introduction We consider a group of agents, each endowed with a set of indivisile

More information

The Role of Monotonicity in the Epistemic Analysis of Strategic Games

The Role of Monotonicity in the Epistemic Analysis of Strategic Games Games 2010, 1, 381-394; doi:10.3390/g1040381 OPEN ACCESS games ISSN 2073-4336 www.mdpi.com/journal/games Article The Role of Monotonicity in the Epistemic Analysis of Strategic Games Krzysztof R. Apt 1,

More information

Rationalizability in general situations

Rationalizability in general situations Rationalizability in general situations Yi-Chun Chen a, Xiao Luo a,y, Chen Qu b a Department of Economics, National University of Singapore, Singapore 117570 b Department of Economics, BI Norwegian Business

More information

Weak Dominance and Never Best Responses

Weak Dominance and Never Best Responses Chapter 4 Weak Dominance and Never Best Responses Let us return now to our analysis of an arbitrary strategic game G := (S 1,...,S n, p 1,...,p n ). Let s i, s i be strategies of player i. We say that

More information

Subjective expected utility in games

Subjective expected utility in games Theoretical Economics 3 (2008), 287 323 1555-7561/20080287 Subjective expected utility in games ALFREDO DI TILLIO Department of Economics and IGIER, Università Bocconi This paper extends Savage s subjective

More information

The Effect of Changes in Risk Attitude on Strategic Behavior

The Effect of Changes in Risk Attitude on Strategic Behavior The Effect of Changes in Risk Attitude on Strategic Behavior Jonathan Weinstein Washington University in St. Louis First draft: October 2013; This version: November 2015 Abstract We study families of normal-form

More information

HW 4 SOLUTIONS. , x + x x 1 ) 2

HW 4 SOLUTIONS. , x + x x 1 ) 2 HW 4 SOLUTIONS The Way of Analysis p. 98: 1.) Suppose that A is open. Show that A minus a finite set is still open. This follows by induction as long as A minus one point x is still open. To see that A

More information

Speculative trade under unawareness: the infinite case

Speculative trade under unawareness: the infinite case Econ Theory Bull (04) :47 60 DOI 0.007/s40505-04-0040-x RESEARCH ARTICLE Speculative trade under unawareness: the infinite case Martin Meier Burkhard C. Schipper Received: 6 August 03 / Accepted: 7 April

More information

Bounded Reasoning and Higher-Order Uncertainty

Bounded Reasoning and Higher-Order Uncertainty Bounded Reasoning and Higher-Order Uncertainty Willemien Kets This version: June 24, 2012 First version: November 2009 Abstract The standard framework for analyzing games with incomplete information models

More information

Thema Working Paper n Université de Cergy Pontoise, France. Hölder Continuous Implementation. Oury Marion

Thema Working Paper n Université de Cergy Pontoise, France. Hölder Continuous Implementation. Oury Marion Thema Working Paper n 2010-06 Université de Cergy Pontoise, France Hölder Continuous Implementation Oury Marion November, 2010 Hölder Continuous Implementation Marion Oury November 2010 Abstract Building

More information

Game Theory and Rationality

Game Theory and Rationality April 6, 2015 Notation for Strategic Form Games Definition A strategic form game (or normal form game) is defined by 1 The set of players i = {1,..., N} 2 The (usually finite) set of actions A i for each

More information

Sequential Equilibria of Multi-Stage Games with Infinite Sets of Types and Actions

Sequential Equilibria of Multi-Stage Games with Infinite Sets of Types and Actions Sequential Equilibria of Multi-Stage Games with Infinite Sets of Types and Actions by Roger B. Myerson and Philip J. Reny* Draft notes October 2011 http://home.uchicago.edu/~preny/papers/bigseqm.pdf Abstract:

More information

Hierarchies of Conditional Beliefs and Interactive Epistemology in Dynamic Games 1

Hierarchies of Conditional Beliefs and Interactive Epistemology in Dynamic Games 1 Hierarchies of Conditional Beliefs and Interactive Epistemology in Dynamic Games 1 Pierpaolo Battigalli Princeton University and European University Institute Marciano Siniscalchi Princeton University

More information

On the Converse Law of Large Numbers

On the Converse Law of Large Numbers On the Converse Law of Large Numbers H. Jerome Keisler Yeneng Sun This version: March 15, 2018 Abstract Given a triangular array of random variables and a growth rate without a full upper asymptotic density,

More information

EQUIVALENCE OF THE INFORMATION STRUCTURE WITH UNAWARENESS TO THE LOGIC OF AWARENESS. 1. Introduction

EQUIVALENCE OF THE INFORMATION STRUCTURE WITH UNAWARENESS TO THE LOGIC OF AWARENESS. 1. Introduction EQUIVALENCE OF THE INFORMATION STRUCTURE WITH UNAWARENESS TO THE LOGIC OF AWARENESS SANDER HEINSALU Abstract. Here it is shown that the unawareness structure in Li (29) is equivalent to a single-agent

More information

Maths 212: Homework Solutions

Maths 212: Homework Solutions Maths 212: Homework Solutions 1. The definition of A ensures that x π for all x A, so π is an upper bound of A. To show it is the least upper bound, suppose x < π and consider two cases. If x < 1, then

More information

Epistemic Foundations for Set-algebraic Representations of Knowledge

Epistemic Foundations for Set-algebraic Representations of Knowledge Epistemic Foundations for Set-algebraic Representations of Knowledge Satoshi Fukuda September 10, 2018 Abstract This paper formalizes an informal idea that an agent s knowledge is characterized by a collection

More information

INTERDEPENDENT PREFERENCES AND STRATEGIC DISTINGUISHABILITY. Dirk Bergemann, Stephen Morris and Satoru Takahashi. September 2010 Revised August 2016

INTERDEPENDENT PREFERENCES AND STRATEGIC DISTINGUISHABILITY. Dirk Bergemann, Stephen Morris and Satoru Takahashi. September 2010 Revised August 2016 INTERDEPENDENT PREFERENCES AND STRATEGIC DISTINGUISHABILITY By Dirk Bergemann, Stephen Morris and Satoru Takahashi September 2010 Revised August 2016 COWLES FOUNDATION DISCUSSION PAPER NO. 1772R3 COWLES

More information

On the Consistency among Prior, Posteriors, and Information Sets

On the Consistency among Prior, Posteriors, and Information Sets On the Consistency among Prior, Posteriors, and Information Sets Satoshi Fukuda September 23, 2018 Abstract This paper studies implications of the consistency conditions among prior, posteriors, and information

More information

Interactive epistemology in games with payoff uncertainty

Interactive epistemology in games with payoff uncertainty Research in Economics 61 (2007) 165 184 www.elsevier.com/locate/rie Interactive epistemology in games with payoff uncertainty Pierpaolo Battigalli a,, Marciano Siniscalchi b,1 a Università Bocconi, IEP

More information

Topological vectorspaces

Topological vectorspaces (July 25, 2011) Topological vectorspaces Paul Garrett garrett@math.umn.edu http://www.math.umn.edu/ garrett/ Natural non-fréchet spaces Topological vector spaces Quotients and linear maps More topological

More information

Perfect Conditional -Equilibria of Multi-Stage Games with Infinite Sets of Signals and Actions (Preliminary and Incomplete)

Perfect Conditional -Equilibria of Multi-Stage Games with Infinite Sets of Signals and Actions (Preliminary and Incomplete) Perfect Conditional -Equilibria of Multi-Stage Games with Infinite Sets of Signals and Actions (Preliminary and Incomplete) Roger B. Myerson and Philip J. Reny Department of Economics University of Chicago

More information

Foundations for Structural Preferences

Foundations for Structural Preferences Foundations for Structural Preferences Marciano Siniscalchi April 28, 2016 Abstract The analysis of key game-theoretic concepts such as sequential rationality or backwardand forward-induction hinges on

More information

Online Appendix: Optimal Retrospective Voting

Online Appendix: Optimal Retrospective Voting Online Appendix: Optimal Retrospective Voting Ethan Bueno de Mesquita 1 Amanda Friedenberg 2 The notation and setup will be as in the main text, with the following exceptions: Let x l : Ω R be a random

More information

Measure and Integration: Concepts, Examples and Exercises. INDER K. RANA Indian Institute of Technology Bombay India

Measure and Integration: Concepts, Examples and Exercises. INDER K. RANA Indian Institute of Technology Bombay India Measure and Integration: Concepts, Examples and Exercises INDER K. RANA Indian Institute of Technology Bombay India Department of Mathematics, Indian Institute of Technology, Bombay, Powai, Mumbai 400076,

More information

int cl int cl A = int cl A.

int cl int cl A = int cl A. BAIRE CATEGORY CHRISTIAN ROSENDAL 1. THE BAIRE CATEGORY THEOREM Theorem 1 (The Baire category theorem. Let (D n n N be a countable family of dense open subsets of a Polish space X. Then n N D n is dense

More information

Strategies and Interactive Beliefs in Dynamic Games

Strategies and Interactive Beliefs in Dynamic Games Strategies and Interactive Beliefs in Dynamic Games Pierpaolo Battigalli 1 Alfredo Di Tillio 2 Dov Samet 3 1 Department of Decision Sciences Bocconi University 2 Department of Economics Bocconi University

More information

Open Sequential Equilibria of Multi-Stage Games with Infinite Sets of Types and Actions

Open Sequential Equilibria of Multi-Stage Games with Infinite Sets of Types and Actions Open Sequential Equilibria of Multi-Stage Games with Infinite Sets of Types and Actions By Roger B. Myerson and Philip J. Reny Department of Economics University of Chicago Paper can be found at https://sites.google.com/site/philipjreny/home/research

More information

ENDOGENOUS REPUTATION IN REPEATED GAMES

ENDOGENOUS REPUTATION IN REPEATED GAMES ENDOGENOUS REPUTATION IN REPEATED GAMES PRISCILLA T. Y. MAN Abstract. Reputation is often modelled by a small but positive prior probability that a player is a behavioral type in repeated games. This paper

More information

INFORMATIONAL ROBUSTNESS AND SOLUTION CONCEPTS. Dirk Bergemann and Stephen Morris. December 2014 COWLES FOUNDATION DISCUSSION PAPER NO.

INFORMATIONAL ROBUSTNESS AND SOLUTION CONCEPTS. Dirk Bergemann and Stephen Morris. December 2014 COWLES FOUNDATION DISCUSSION PAPER NO. INFORMATIONAL ROBUSTNESS AND SOLUTION CONCEPTS By Dirk Bergemann and Stephen Morris December 2014 COWLES FOUNDATION DISCUSSION PAPER NO. 1973 COWLES FOUNDATION FOR RESEARCH IN ECONOMICS YALE UNIVERSITY

More information

The Folk Theorem for Finitely Repeated Games with Mixed Strategies

The Folk Theorem for Finitely Repeated Games with Mixed Strategies The Folk Theorem for Finitely Repeated Games with Mixed Strategies Olivier Gossner February 1994 Revised Version Abstract This paper proves a Folk Theorem for finitely repeated games with mixed strategies.

More information

Measure and Category. Marianna Csörnyei. ucahmcs

Measure and Category. Marianna Csörnyei.   ucahmcs Measure and Category Marianna Csörnyei mari@math.ucl.ac.uk http:/www.ucl.ac.uk/ ucahmcs 1 / 96 A (very short) Introduction to Cardinals The cardinality of a set A is equal to the cardinality of a set B,

More information

Online Appendices for Large Matching Markets: Risk, Unraveling, and Conflation

Online Appendices for Large Matching Markets: Risk, Unraveling, and Conflation Online Appendices for Large Matching Markets: Risk, Unraveling, and Conflation Aaron L. Bodoh-Creed - Cornell University A Online Appendix: Strategic Convergence In section 4 we described the matching

More information

Problem Set 2: Solutions Math 201A: Fall 2016

Problem Set 2: Solutions Math 201A: Fall 2016 Problem Set 2: s Math 201A: Fall 2016 Problem 1. (a) Prove that a closed subset of a complete metric space is complete. (b) Prove that a closed subset of a compact metric space is compact. (c) Prove that

More information

Uncertain Rationality, Depth of Reasoning and Robustness in Games with Incomplete Information

Uncertain Rationality, Depth of Reasoning and Robustness in Games with Incomplete Information Uncertain Rationality, Depth of Reasoning and Robustness in Games with Incomplete Information Fabrizio Germano Jonathan Weinstein Peio Zuazo-Garin July 2, 2017 Abstract Predictions under common knowledge

More information

Robust Knowledge and Rationality

Robust Knowledge and Rationality Robust Knowledge and Rationality Sergei Artemov The CUNY Graduate Center 365 Fifth Avenue, 4319 New York City, NY 10016, USA sartemov@gc.cuny.edu November 22, 2010 Abstract In 1995, Aumann proved that

More information

Ergodic Theory. Constantine Caramanis. May 6, 1999

Ergodic Theory. Constantine Caramanis. May 6, 1999 Ergodic Theory Constantine Caramanis ay 6, 1999 1 Introduction Ergodic theory involves the study of transformations on measure spaces. Interchanging the words measurable function and probability density

More information

The Kolmogorov extension theorem

The Kolmogorov extension theorem The Kolmogorov extension theorem Jordan Bell jordan.bell@gmail.com Department of Mathematics, University of Toronto June 21, 2014 1 σ-algebras and semirings If X is a nonempty set, an algebra of sets on

More information

Bargaining Under Strategic Uncertainty

Bargaining Under Strategic Uncertainty Bargaining Under Strategic Uncertainty Amanda Friedenberg September 2, 2013 Extremely Preliminary Abstract This paper provides a novel understanding of delays in reaching agreements based on the idea of

More information

g 2 (x) (1/3)M 1 = (1/3)(2/3)M.

g 2 (x) (1/3)M 1 = (1/3)(2/3)M. COMPACTNESS If C R n is closed and bounded, then by B-W it is sequentially compact: any sequence of points in C has a subsequence converging to a point in C Conversely, any sequentially compact C R n is

More information

WEYL-HEISENBERG FRAMES FOR SUBSPACES OF L 2 (R)

WEYL-HEISENBERG FRAMES FOR SUBSPACES OF L 2 (R) PROCEEDINGS OF THE AMERICAN MATHEMATICAL SOCIETY Volume 129, Numer 1, Pages 145 154 S 0002-9939(00)05731-2 Article electronically pulished on July 27, 2000 WEYL-HEISENBERG FRAMES FOR SUBSPACES OF L 2 (R)

More information

SVETLANA KATOK AND ILIE UGARCOVICI (Communicated by Jens Marklof)

SVETLANA KATOK AND ILIE UGARCOVICI (Communicated by Jens Marklof) JOURNAL OF MODERN DYNAMICS VOLUME 4, NO. 4, 010, 637 691 doi: 10.3934/jmd.010.4.637 STRUCTURE OF ATTRACTORS FOR (a, )-CONTINUED FRACTION TRANSFORMATIONS SVETLANA KATOK AND ILIE UGARCOVICI (Communicated

More information

Rationalizability in General Situations

Rationalizability in General Situations Rationalizability in General Situations Yi-Chun Chen a, Xiao Luo a,y, Chen Qu b a Department of Economics, National University of Singapore, Singapore 117570 b Department of Economics, BI Norwegian Business

More information

Columbia University. Department of Economics Discussion Paper Series

Columbia University. Department of Economics Discussion Paper Series Columbia University Department of Economics Discussion Paper Series Equivalence of Public Mixed-Strategies and Private Behavior Strategies in Games with Public Monitoring Massimiliano Amarante Discussion

More information

4 Countability axioms

4 Countability axioms 4 COUNTABILITY AXIOMS 4 Countability axioms Definition 4.1. Let X be a topological space X is said to be first countable if for any x X, there is a countable basis for the neighborhoods of x. X is said

More information

Rationalizability in General Situations

Rationalizability in General Situations Rationalizability in General Situations Yi-Chun Chen a, Xiao Luo a,y, Chen Qu b a Department of Economics, National University of Singapore, Singapore 117570 b Department of Economics, BI Norwegian Business

More information

HILBERT SPACES AND THE RADON-NIKODYM THEOREM. where the bar in the first equation denotes complex conjugation. In either case, for any x V define

HILBERT SPACES AND THE RADON-NIKODYM THEOREM. where the bar in the first equation denotes complex conjugation. In either case, for any x V define HILBERT SPACES AND THE RADON-NIKODYM THEOREM STEVEN P. LALLEY 1. DEFINITIONS Definition 1. A real inner product space is a real vector space V together with a symmetric, bilinear, positive-definite mapping,

More information

Tijmen Daniëls Universiteit van Amsterdam. Abstract

Tijmen Daniëls Universiteit van Amsterdam. Abstract Pure strategy dominance with quasiconcave utility functions Tijmen Daniëls Universiteit van Amsterdam Abstract By a result of Pearce (1984), in a finite strategic form game, the set of a player's serially

More information

DO FIVE OUT OF SIX ON EACH SET PROBLEM SET

DO FIVE OUT OF SIX ON EACH SET PROBLEM SET DO FIVE OUT OF SIX ON EACH SET PROBLEM SET 1. THE AXIOM OF FOUNDATION Early on in the book (page 6) it is indicated that throughout the formal development set is going to mean pure set, or set whose elements,

More information

Rationalization and Incomplete Information

Rationalization and Incomplete Information Rationalization and Incomplete Information Pierpaolo Battigalli Bocconi University and IGIER pierpaolo.battigalli@uni-bocconi.it Marciano Siniscalchi Northwestern University and Princeton University marciano@northwestern.edu

More information

Conservative Belief and Rationality

Conservative Belief and Rationality Conservative Belief and Rationality Joseph Y. Halpern and Rafael Pass Department of Computer Science Cornell University Ithaca, NY, 14853, U.S.A. e-mail: halpern@cs.cornell.edu, rafael@cs.cornell.edu January

More information

PAIRWISE EPISTEMIC CONDITIONS FOR NASH EQUILIBRIUM 1

PAIRWISE EPISTEMIC CONDITIONS FOR NASH EQUILIBRIUM 1 PAIRWISE EPISTEMIC CONDITIONS FOR NASH EQUILIBRIUM 1 Christian W. Bach and Elias Tsakas We introduce a framework for modeling pairwise interactive beliefs and provide an epistemic foundation for Nash equilibrium

More information

Essential equilibria in normal-form games

Essential equilibria in normal-form games Journal of Economic Theory 145 (2010) 421 431 www.elsevier.com/locate/jet Note Essential equilibria in normal-form games Oriol Carbonell-Nicolau 1 Department of Economics, Rutgers University, 75 Hamilton

More information

NAME: Mathematics 205A, Fall 2008, Final Examination. Answer Key

NAME: Mathematics 205A, Fall 2008, Final Examination. Answer Key NAME: Mathematics 205A, Fall 2008, Final Examination Answer Key 1 1. [25 points] Let X be a set with 2 or more elements. Show that there are topologies U and V on X such that the identity map J : (X, U)

More information

ONLINE APPENDICES FOR INCENTIVES IN EXPERIMENTS: A THEORETICAL INVESTIGATION BY AZRIELI, CHAMBERS & HEALY

ONLINE APPENDICES FOR INCENTIVES IN EXPERIMENTS: A THEORETICAL INVESTIGATION BY AZRIELI, CHAMBERS & HEALY ONLINE APPENDICES FOR INCENTIVES IN EXPERIMENTS: A THEORETICAL INVESTIGATION BY AZRIELI, CHAMBERS & HEALY Appendix B. Modeling Games as Decisions In this appendix we describe how one can move seamlessly

More information

Statistics for Financial Engineering Session 2: Basic Set Theory March 19 th, 2006

Statistics for Financial Engineering Session 2: Basic Set Theory March 19 th, 2006 Statistics for Financial Engineering Session 2: Basic Set Theory March 19 th, 2006 Topics What is a set? Notations for sets Empty set Inclusion/containment and subsets Sample spaces and events Operations

More information

TWO TO ONE IMAGES AND PFA

TWO TO ONE IMAGES AND PFA TWO TO ONE IMAGES AND PFA ALAN DOW Astract. We prove that all maps on N that are exactly two to one are trivial if PFA is assumed. 1. Introduction A map f : X K is precisely two to one if for each k K,

More information

Bayesian Persuasion Online Appendix

Bayesian Persuasion Online Appendix Bayesian Persuasion Online Appendix Emir Kamenica and Matthew Gentzkow University of Chicago June 2010 1 Persuasion mechanisms In this paper we study a particular game where Sender chooses a signal π whose

More information

Monotone equilibria in nonatomic supermodular games. A comment

Monotone equilibria in nonatomic supermodular games. A comment Monotone equilibria in nonatomic supermodular games. A comment Lukasz Balbus Kevin Reffett Lukasz Woźny April 2014 Abstract Recently Yang and Qi (2013) stated an interesting theorem on existence of complete

More information

An axiomatization of minimal curb sets. 1. Introduction. Mark Voorneveld,,1, Willemien Kets, and Henk Norde

An axiomatization of minimal curb sets. 1. Introduction. Mark Voorneveld,,1, Willemien Kets, and Henk Norde An axiomatization of minimal curb sets Mark Voorneveld,,1, Willemien Kets, and Henk Norde Department of Econometrics and Operations Research, Tilburg University, The Netherlands Department of Economics,

More information

STRONGLY CONNECTED SPACES

STRONGLY CONNECTED SPACES Undergraduate Research Opportunity Programme in Science STRONGLY CONNECTED SPACES Submitted by Dai Bo Supervised by Dr. Wong Yan-loi Department of Mathematics National University of Singapore Academic

More information

The Context of the Game

The Context of the Game The Context of the Game Amanda Friedenberg Martin Meier First Draft: October 2007 This Draft: June 2015 Abstract Here, we study games of incomplete information and argue that it is important to correctly

More information

Topologies on Types. Drew Fudenberg Harvard University. First Draft: April 2004 This Draft: January Abstract

Topologies on Types. Drew Fudenberg Harvard University. First Draft: April 2004 This Draft: January Abstract Topologies on Types Eddie Dekel Northwestern University and Tel Aviv University Drew Fudenberg Harvard University Stephen Morris Princeton University First Draft: April 2004 This Draft: January 2006 Abstract

More information

#A50 INTEGERS 14 (2014) ON RATS SEQUENCES IN GENERAL BASES

#A50 INTEGERS 14 (2014) ON RATS SEQUENCES IN GENERAL BASES #A50 INTEGERS 14 (014) ON RATS SEQUENCES IN GENERAL BASES Johann Thiel Dept. of Mathematics, New York City College of Technology, Brooklyn, New York jthiel@citytech.cuny.edu Received: 6/11/13, Revised:

More information

arxiv: v1 [math.pr] 26 Mar 2008

arxiv: v1 [math.pr] 26 Mar 2008 arxiv:0803.3679v1 [math.pr] 26 Mar 2008 The game-theoretic martingales behind the zero-one laws Akimichi Takemura 1 takemura@stat.t.u-tokyo.ac.jp, http://www.e.u-tokyo.ac.jp/ takemura Vladimir Vovk 2 vovk@cs.rhul.ac.uk,

More information

Order on Types based on Monotone Comparative Statics

Order on Types based on Monotone Comparative Statics Order on Types based on Monotone Comparative Statics Takashi Kunimoto Takuro Yamashita July 10, 2015 Monotone comparative statics Comparative statics is important in Economics E.g., Wealth Consumption

More information

II - REAL ANALYSIS. This property gives us a way to extend the notion of content to finite unions of rectangles: we define

II - REAL ANALYSIS. This property gives us a way to extend the notion of content to finite unions of rectangles: we define 1 Measures 1.1 Jordan content in R N II - REAL ANALYSIS Let I be an interval in R. Then its 1-content is defined as c 1 (I) := b a if I is bounded with endpoints a, b. If I is unbounded, we define c 1

More information

Payoff Continuity in Incomplete Information Games

Payoff Continuity in Incomplete Information Games journal of economic theory 82, 267276 (1998) article no. ET982418 Payoff Continuity in Incomplete Information Games Atsushi Kajii* Institute of Policy and Planning Sciences, University of Tsukuba, 1-1-1

More information

A True Expert Knows which Question Should be Asked.

A True Expert Knows which Question Should be Asked. A True Expert Knows which Question Should be Asked. Eddie Dekel and Yossi Feinberg June 2004 Abstract We suggest a test for discovering whether a potential expert is informed of the distribution of a stochastic

More information

Bounded Reasoning and Higher-Order Uncertainty

Bounded Reasoning and Higher-Order Uncertainty Bounded Reasoning and Higher-Order Uncertainty Willemien Kets JOB MARKET PAPER November 15, 2010 Abstract Harsanyi type structures, the device traditionally used to model players beliefs in games, generate

More information

The Index of Nash Equilibria

The Index of Nash Equilibria Equilibria in Games, Santiago, Chile January 10, 2017 Finite Normal-Form Games We consider in these lectures, the set of finite games with fixed strategy sets and parametrized by the payoff functions.

More information

Topologies on Types. Drew Fudenberg Harvard University. First Draft: April 2004 This Draft: August Abstract

Topologies on Types. Drew Fudenberg Harvard University. First Draft: April 2004 This Draft: August Abstract Topologies on Types Eddie Dekel Northwestern University and Tel Aviv University Drew Fudenberg Harvard University Stephen Morris Princeton University First Draft: April 2004 This Draft: August 2005 Abstract

More information

Important Concepts Read Chapter 2. Experiments. Phenomena. Probability Models. Unpredictable in detail. Examples

Important Concepts Read Chapter 2. Experiments. Phenomena. Probability Models. Unpredictable in detail. Examples Probability Models Important Concepts Read Chapter 2 Probability Models Examples - The Classical Model - Discrete Spaces Elementary Consequences of the Axioms The Inclusion Exclusion Formulas Some Indiscrete

More information

Chapter 2 Metric Spaces

Chapter 2 Metric Spaces Chapter 2 Metric Spaces The purpose of this chapter is to present a summary of some basic properties of metric and topological spaces that play an important role in the main body of the book. 2.1 Metrics

More information

Division of the Humanities and Social Sciences. Sums of sets, etc.

Division of the Humanities and Social Sciences. Sums of sets, etc. Division of the Humanities and Social Sciences Sums of sets, etc. KC Border September 2002 Rev. November 2012 Rev. September 2013 If E and F are subsets of R m, define the sum E + F = {x + y : x E; y F

More information

ESSAYS IN GAME THEORY

ESSAYS IN GAME THEORY The Pennsylvania State University The Graduate School College of the Liberal Arts ESSAYS IN GAME THEORY A Dissertation in Economics by Byung Soo Lee c 2011 Byung Soo Lee Submitted in Partial Fulfillment

More information

INVERSE LIMITS AND PROFINITE GROUPS

INVERSE LIMITS AND PROFINITE GROUPS INVERSE LIMITS AND PROFINITE GROUPS BRIAN OSSERMAN We discuss the inverse limit construction, and consider the special case of inverse limits of finite groups, which should best be considered as topological

More information

The Local Best Response Criterion: An Epistemic Approach to Equilibrium Refinement. Herbert Gintis

The Local Best Response Criterion: An Epistemic Approach to Equilibrium Refinement. Herbert Gintis The Local Best Response Criterion: An Epistemic Approach to Equilirium Refinement Herert Gintis May 2, 2008 Astract The standard refinement criteria for extensive form games, including sugame perfect,

More information