Comprehensive Rationalizability
|
|
- Chester Phillips
- 5 years ago
- Views:
Transcription
1 Comprehensve Ratonalzablty Avad Hefetz Martn Meer y Burkhard C. Schpper z Prelmnary & Incomplete: August 12, 2010 Abstract We present a new soluton concept for strategc games called comprehensve ratonalzablty. It embodes common con dence of cautous ratonalty based on a sound epstemc characterzaton n a unversal type space. It nether re nes nor s re ned by terated admssblty but t concdes wth terated admssblty n many applcatons n the lterature. Keywords: cautousness. Iterated admssblty, ratonalzablty, lexcographc belef systems, JEL-Class catons: C72. The Economcs and Management Department, The Open Unversty of Israel. Emal: avadhe@openu.ac.l y Insttut für Höhere Studen, Venna. Emal: meer@hs.ac.at z Department of Economcs, Unversty of Calforna, Davs. Emal: bcschpper@ucdavs.edu
2 1 Introducton We ntroduce a noton of ratonalzablty based on lexcographc conjectures (full-support mutually sngular lexcographc belef systems) about the other players strateges, comprehensve ratonalzablty. It s based on ncreasng levels of mutual assumpton of ratonalty. Ths noton nether re nes nor s re ned by terated admssblty (teratve elmnaton of weakly domnated strateges). Ths new soluton concept s provded wth an epstemc characterzaton. 2 Lexcographc conjectures A lexcographc conjecture over a Hausdor space X s a nte sequence of regular Borel probablty measures over X = ( 1 ; : : : n ) 2 ( (X)) n whch are (1) mutually sngular: `?`0 for ` 6= `0 (2) the unon of whose supports cover X : [`n supp` = X If Y and Z are two dsjont measurable subsets of X; we say that the lexcographc conjecture = ( 1 ; : : : n ) deems Y as n ntely more lkely than Z f ` (Y ) > 0 for some ` 2 f1; : : : ; ng; and whenever ` (Z) > 0 t s the case that `0 (Y ) = 0 8`0 `: If Y 0 and Z 0 are two subsets of X; not necessarly dsjont, we say that the lexcographc conjecture = ( 1 ; : : : n ) deems Y 0 as n ntely more lkely than Z 0 f t deems Y 0 n Z 0 n ntely more lkely than Z 0. We say that Y s assumed by the lexcographc conjecture = ( 1 ; : : : n ) f t deems Y as n ntely more lkely than X n Y. 1
3 3 Games, lexcographc domnaton and lexcographc best reples Consder a game G wth players 2 I and nte strategy sets (S ) 2I equpped wth the dscrete topology, and utlty functons (u ) 2I. If = ( 1; : : : n) 2 ( (S )) n s a lexcographc conjecture over the other players strategy pro les S = Q j6= Sj, and s ; ^s 2 S are two strateges of player ; we say that the strategy s s `-domnated by ^s w.r.t, denoted s ` ^s f Z u s ; Z d ì < u ^s ; d ì whle Z u s ; Z d ì 0 = u ^s ; d ì 0 for all `0 < ` (f such `0 exsts,.e. f ` 6= 1). We further say that the strategy s s lexcographcally domnated by ^s w.r.t, denoted s ^s f s ` ^s for some ` n. If s s not lexcographcally domnated w.r.t by any other strategy of player ; we say that s s a lexcographc best reply to : We denote by LBR ( ) S the set of player s lexcographc best reples to. Let M be a set of lexcographc conjectures of player on the other players strateges. If s ` ^s for some strategy ^s of player 1 but there does not exst a lexcographc conjecture 1.e., gven t s the case that s s just as good as ^s at all levels `0 < `; but s strctly worse than ^s at level `: 2
4 ~ 2 M wth respect to whch s s optmal at all the levels up to and ncludng `; we say that s s `-unjust able w.r.t. M. 4 Comprehensve ratonalzablty Let C 1 = (S ) and R 1 = S. De ne nductvely C k+1 = 2 C k : R k s assumed by R k+1 = s 2 S : 9 2 C k+1 for whch s s a best reply Player s comprehensve ratonalzable strateges are Remark 1 R k R k 1 for every k 1. R 1 = Proof. Ths s clear snce C k C k 1. 1\ k=0 R k Proposton 1 C k 6= ; and R k 6= ; for any k 0. Proof. Both R 0 and C 0 are nonempty for all 2 I. Assume R k and C k are nonempty for all 2 I. Clam: There exsts a k+1 2 C k that assumes R k. Proof of the clam: To see ths, note that by the nducton hypothess k assumes R` for ` = 0; :::; k 1. Agan, by the nducton hypothess R k s nonempty and by Remark 1, R k R k 1. Fx k = ( k ;0; :::; k ;m 1) 2 N (A ) for some nteger m. Snce R k s nonempty by the nducton hypothess, and k s a lexcographc conjecture (n partcular t s full support sequence), we have a j m 1 such that k ;j(r k ) > 0. Let k ;n 1 ; :::; k ;n p be all the k ;j such that k ;j(r k ) > 0 wth n r < n r+1, for all r = 1; :::; p 1. Moreover, let k ;m 1 ; :::; k ;m q be all the k ;j such that k ;j(r k 1 n R k ) > 0. Snce k assumes R k, there exsts ` m 1 such that k ;j(r k 1 ) = 1 for all j = 0; :::; ` and k ;j(r k 1 ) = 0 for j > `. Note that we have fn 1 ; :::; n p g [ fm 1 ; :::; m q g = f0; :::; `g. 3
5 Now, de ne = k+1 ;0 ; :::; k+1 ;p+q+m 1 k+1 where k+1 ;k = 8 >< >: k ;n j+1 (jr k ) for j = 0; :::; p 1 k ;m j p+1 (jr k 1 n R k ) for j = p; :::; p + q 1 k ;j p q+`+1 for j = p + q; :::; m 2 ` + p + q By constructon k+1 assumes R k, R k 1, but also R j for j < k 1. To see ths, let j < k 1. Snce k assumes R j, there exsts `0 0 such that k ;r(r j ) = 1 for r `0 and k ;r(r j ) = 0 for all r > `0. But snce R k R j, we have `0 `. If R k = R j, there s nothng to show. If R k $ R j then, by the fact that k s a lexcographc conjecture (n partcular, a full support sequence), we must have `0 > `. Then, k+1 ;r (Rj ) = 1 for r = 0; :::; p + q 1, but also for all j = p + q; :::; p + q 1 ` + `0. And we have by constructon, k+1 ;r (Rj ) = 0 for r > p + q 1 ` + `0. So, k+1 nshes the proof of the clam. assumes also R j. Ths Snce C k+1 s nonempty and the game s nte, we must have that R k+1 s nonempty. 4.1 Comprehensve Ratonalzablty vs. Iterated Admssblty An acton a 2 A s weakly domnated wth respect to X Y A A f there exsts 2 (A ) wth (X) = 1 such that P b 2X (b )u (b ; a ) u (a ; a ) for every a 2 Y and P b 2X (b )u (b ; a ) > u (a ; a ) for some a 2 Y. Otherwse, we say that a s admssble wth respect to X Y. Let A 0 = A and de ne for k 1 A k+1 = a 2 A k : a s admssble wth respect to A k A k : (1) The set of teratvely admssble actons of player s A 1 = 1\ A k : (2) k=0 4
6 A game s solvable by terated admssblty f for every player 2 I, the payo functon u s constant on all outcomes n 2I A 1. Smlarly, a game s solvable by comprehensve ratonalzablty f for every player 2 I, the payo functon s constant on all outcomes n 2I R 1. Example 1 Surprsngly, comprehensve ratonalzablty s not a re nement of terated admssblty as the followng example demonstrates. x y z w a b c 4; 0 4; 1 0; 1 0; 0 0; 1 4; 1 3; 0 2; 2 2; 1 5; 0 0; 1 0; 2 The order of elmnaton under terated admssblty s a; w; c and then both y and z. The maxmal reducton or teratve admssble actons are f(x; b)g. For comprehensve ratonalzablty, the order of elmnaton s a and then both z and w. The set of comprehensve ratonalzable pro les s fx; yg fb; cg. Yet, there are other examples such as the next example where comprehensve ratonalzablty re nes the teratve admssble actons. Example 2 The followng example shows that comprehensve ratonalzablty may re ne terated admssblty. a b c x 4; 0 4; 1 0; 2 y 0; 0 0; 1 4; 2 z 3; 0 2; 2 2; 1 Here, only acton a can be elmnated under teratve admssblty and we reman wth fx; y; zg fb; cg as the maxmal reducton. However, under comprehensve ratonalzablty the order of elmnaton s a; z; b and x. Thus, the f(y; c)g s the unque comprehensve ratonalzable acton pro le. Every rst level admssble acton s rst level comprehensve ratonalzable and vce versa. Ths follows from Blume, Brandenburger and Dekel (1991b, Proposton 1). 5
7 What are su cent condtons for teratve admssble actons to be comprehensve ratonalzable? The reason for why comprehensve ratonalzablty does not re ne terated admssblty n Example 1 s that there s no full support belef of player 1 on the rst level admssble actons of player 2 such that z s a strct best reply. E.g., for player 1, playng both x or y s as good as playng z aganst the belef that puts equal probablty on both b and c. Lexcographc best reples re ne best reples. Lemma 1 If for all 2 I and any k 2, f for every a 2 A k there exsts a full support belef 2 (A k 1 ) for whch a s the strct best reply, then A k = R k for all 2 I. Consequently, A 1 = R 1 for all 2 I. Proof. For all 2 I, A 1 = R 1 follows from Blume, Brandenburger and Dekel (1991b, Proposton 1). We show by nducton that f for all 2 I, A k = R k and for every a 2 A k+1 there exsts a full support belef 2 (A k ) for whch a s the strct best reply, then A k+1 = R k+1. Consder rst the case A k+1 R k+1. We need to show that for 2 (A k ) there exst a lexcographc conjecture ( ;0 ; :::; ;n ) 2 C k+1 for some n 0 such that a s a lexcographc best reply. Consder any lexcographc conjecture ( ;0 ; :::; ;n ) wth ;0 =. Snce s full support on A k, ( ;0 ; :::; ;n ) s by de nton mutually sngular, and by the nducton hypothess A k = R k, we have that R k s assumed by ( ;0 ; :::; ;n ). Thus ( ;0 ; :::; ;n ) 2 C k+1. Snce a was the only best reply to among actons n A k and A k = R k, a s a lexcographc best reply to ( ;0 ; :::; ;n ) among actons n R k no matter what are the belefs ;`, ` = 1; :::; n, n the lexcographc conjecture. a s also a lexcographc best reply to ( ;0 ; :::; ;n ) among actons n A, snce f there exsts an acton a 0 2 A that lexcographcally domnates a wth respect to ( ;0 ; :::; ;n ), then a =2 R k. Snce by the nducton hypothess, A k = R k, a =2 A k, a contradcton. We conclude that a 2 R k+1. Consder now the case R k+1 A k+1, and suppose to the contrary that there exsts an acton a 2 R k+1 wth a =2 A k+1. Snce a 2 R k+1, there exsts a lexcographc conjecture ( ;0 ; :::; ;n ) 2 C k+1 for some n 0 such that a s a lexcographc best reply to ( ;0 ; :::; ;n ) and ( ;0 ; :::; ;n ) assumes R k. Thus, there s smallest ` 2 f1; :::; ng such that ;`0(R k ) = 0 for all `0 `. By the nducton hypothess, A k = R k. Thus by arguments analogous to Blume, Brandenburger and Dekel (1991b, Proposton 1), we construct from ( ;0 ; :::; ;`) a full support belef on A k such that a s a best reply to. Thus a 2 A k+1. 6
8 Snce there s a nte number of actons for each player, we also have A 1 = R 1 for every 2 I. In many economcally relevant examples the condton of Lemma 1 s sats ed and comprehensve ratonalzablty concdes wth terated admssblty. We dscuss some of the examples n sequel. The rst example concerns votng wth a presdent. Example 3 (Votng wth a presdent) Consder an example of majorty votng wth a presdent (Mouln, 1986, p ). Three players have to select one of three alternatves fa; b; cg. If a majorty votes for an alternatve, t wll be mplemented. Otherwse, the alternatve selected by player 1, the presdent, s selected. For smplcty, for each assgn payo s 3; 2 and 1 n the order of preferences. The strategc form s gven by the followng three matrxes where player 1 chooses matrces, player 2 rows and player 3 columns. a b c a b c a b c a b c 3; 2; 1 3; 2; 1 3; 2; 1 a 3; 2; 1 2; 1; 3 2; 1; 3 a 3; 2; 1 1; 3; 2 1; 3; 2 3; 2; 1 2; 1; 3 3; 2; 1 b 2; 1; 3 2; 1; 3 2; 1; 3 b 1; 3; 2 2; 1; 3 1; 3; 2 3; 2; 1 3; 2; 1 1; 3; 2 c 2; 1; 3 2; 1; 3 1; 3; 2 c 1; 3; 2 1; 3; 2 1; 3; 2 a b c At the rst round, the only admssble acton s a. For players 1 t s the set fa; cg and for player 2 t s fb; cg. Hence the game s reduced to a c b c 3; 2; 1 3; 2; 1 3; 2; 1 1; 3; 2 a Ths game can be further reduced by weak domnance to (a; c; c), the only teratve admssble pro le. Interestngly, n ths pro le the presdent faces hs lowest ranked alternatve. Note that for player 2 a s the strct best reply to a full support belef concentrated on for nstance the pro le of other players actons (c; a) and c s the strct best reply to a full support belef concentrated on the pro le of other players actons (b; c). Smlarly, for player 3. Thus Lemma 1 apples and the unque comprehensve ratonalzable acton pro le s (a; c; c) as well. 7
9 Example 4 (Dvdng Money (Brams, Klgour, and Davs, 1993, Osborne, 2004, p. 38)) Two players use the followng procedure to dvde $10 between themselves. Each person names a number of dollars (a nonnegatve nteger), at most equal to $10. If the sum s at most $10, then each person receves the amount of money she named and the remander s burned. If the sum exceeds $10 and the players named d erent amounts, then the person who named the smaller amount receves that amount and the other player receves the remanng money. If the sum exceeds $10 and the amounts named by the players are the same, then each player receves $5. In the rst round, every amount weakly lower than $5 s weakly domnated by $6. If the opponent names more than $6, then the sum exceeds $10 and the player receves $6. If the opponent names exactly $6, then the sum also exceeds $10, and both players receve $5. If the opponent names any amount strctly than $5, then the sum at most $10, and the player receves $6. Any other amount a 2 f$6; :::; $9g s a strct best response to a+1, and $10 s a strct best response to $0. In the followng rounds, the hghest remanng amount s weakly domnated by $6 and all other amounts are a strct best response to that amount plus 1. Thus, the maxmal reducton under terated admssblty s ($6; $6) yeldng a payo of $5 to each player. Note that snce every amount not elmnated at the prevous round s a strct best reply to some full support belef on the remanng amounts, Lemma 1 apples and the game s solvable by comprehensve ratonalzablty. 4.2 Prce Competton Consder a symmetrc Bertrand duopoly n whch each rm s restrcted to choose nteger prces. Let p be the prce and c the cost (also n ntegers), and D(p) = ( p f p 0 f p > and assume c + 1 <. The pro t functon of rm 6= j s gven by 8 >< (p ; p j ) = >: (p c)d(p ) f p > p j 1 2 c)d(p ) f p = p j 0 f p < p j Assume that the monopoly prce s unque. 8
10 Proposton 2 In the Bertrand duopoly, p = c + 1 s the unque comprehensve ratonalzable prce. It also the unque teratve admssble prce. Proof. We consder teratvely admssblty and then apply Lemma 1 to show that t concdes wth comprehensve ratonalzablty. In the rst round every prce n excess of the monopoly prce s weakly domnated by the monopoly prce. If the opponent sets a prce weakly hgher than the monopoly prce, then the monopoly prce s strctly better than any prce strctly hgher than the monopoly prce. If the opponent sets a prce strctly below the monopoly prce, the monopoly prce s as good as any prce strctly hgher than the monopoly prce. In fact, there s no belef about the opponent s prces for whch a prce strctly hgher than the monopoly prce s a best reply. Also n the rst round, every prce equal to at most c s weakly domnated by the prce c + 1. If the opponent sets a prce weakly larger than c+1, then a prce equal to c+1 s strctly better than a prce equal to at most c. If the opponent sets a prce strctly below c + 1, then a prce equal to c+1 s as good as a prce equal to c and strctly better than a prce strctly below c. That s, c s a best reply to the belef that the opponent sets a prce of at most c but t s not the unque best reply. c can never be a best reply to a full support belef. Every other prce p s a strct best response to p+1, so no other prce s weakly domnated. To see ths, note that for any p c+1 t s strctly better to obtan all the demand at the prce p than to obtan half of the demand at the prce p+1. That s, consder any p c+2. We need to show that 1 (p c)( p) < (p 1 c)( p+1) = (p 1 c)( p)+p 1 c. 2 We have 1 (p c) p 1 c and p 1 c > 0 because p c + 2. By the same argument, 2 at each subsequent round of teratve elmnaton, the hghest remanng prce s weakly domnated by the next hghest prce. The par of prces that remans s (c + 1; c + 1). Let p` be the hghest prce that remans after the `-th round of elmnaton of weakly domnated prces. Lemma 1 mples that at each level ` = 0; 1; :::, the set of admssble prces concde wth the set of comprehensve ratonalzable prces snce every prce n fc + 1; :::; p`g s a strct best reply to a full support belef over prces remanng from the prevous round. E.g., p 2 fc + 1; :::; p`g s a strct best reply to the belef that s concentrated on p Second Prce Common Value Auctons Consder a second prce common value aucton n whch each of the n bdders receves a prvately observed sgnal about the value of the object to be auctoned o. Ths sgnal x s ndependently and dentcally dstrbuted over some nte set of sgnals X N 9
11 (ntegers) and assume that every sgnal may be drawn wth postve probablty. Let x max denote the realzaton of the hghest of n sgnals. The common value of the object to each bdder s x max. Each bdder submts a bd n a sealed envelop. The hghest bdder wns and pays the second hghest bd. In case of a te, each hghest bdder obtans the object wth equal probablty. Let b : X! R denote the bd functon of player. Bddng your value s the unque bddng functon that s obtaned after two rounds of terated admssblty (see Harstad and Levn, 1985). Comprehensve ratonalzablty yelds the same outcome. Proposton 3 In the second prce common value sealed bd aucton, R k = fb(x ) = x g for all k 2 and 2 I. Proof. For each bdder 2 I, any bd strctly above max X and strctly below x s weakly domnated. Thus, by Blume, Brandenburger and Dekel (1991b, Proposton 1) R 1 = fb : For all x 2 X; x b(x ) max Xg. At the second level, suppose a bdder bds strctly above hs sgnal. If hs bd s below the hghest bd, then he does not obtan the object. Bddng hs sgnal would be a lexcographc best reply n ths case as he receves nothng and pays nothng. If the second hghest bd s below hs sgnal, then he stll receves the object. Bddng hs sgnal would be a lexcographc best reply n ths case snce he would stll obtan the object and hs payment would reman unchanged. Consder now the case n whch the second hghest bd s between hs bd and hs sgnal. By hs second level lexcographc conjectures he assumes that all bdders j 6= select a bd n fx j ; max Xg. Thus, n ths case x max > x. Ths means he would pay more than the value of the object. A lexcographc best reply s to bd hs sgnal nstead, n whch case he pays nothng and obtans nothng. In fact, t s the only lexcographc best reply snce opponents may have drawn any sgnals from X. Thus we have shown that at the second level, bddng the sgnal remans as the only comprehensve ratonalzable bd functon. 4.4 Relatonshp to other Iteratve Soluton Concepts Bernhem (1984) and Pearce (1984) ntroduced ratonalzablty. Let P 0 = A 0 and B 0 = f 2 (A )g and de ne for k 1, B k+1 = f 2 (P k )g and P k+1 = fa 2 A : a s a best reply to some 2 B k+1 g. The set of ratonalzable actons s 10
12 P 1 = T 1 k=0 P k. 2 Proposton 4 For every 2 I and k 0, R k ratonalzable acton s ratonalzable. P k. Moreover, every comprehensve Proof. By de nton, R 0 = P 0 = A for all 2 I. Moreover, for all 2 I, R 1 P 1 snce any lexcographc best reply to = ( 1 ; ::; n ) on A, for some nte n 1, s a best reply to 1. Suppose that R k P k for all 2 I and some k 2. We show that R k+1 P k+1. Suppose to the contrary that there exsts an acton a 2 R k+1 wth a =2 P k+1. There s no conjecture 2 (P k ) for whch a s a best reply. Yet, a s a lexcographc best reply to some lexcographc conjecture ( 1 ; :::; n ) 2 (R k ) n. Thus, t s n partcular a best reply to 1 on R k. Snce by the nducton hypothess, R k P k, a contradcton. P k, 1 s a conjecture on Snce well-known games such as Guess-the-Average are solvable by ratonalzablty, the result mples that they are solvable by comprehensve ratonalzablty as well. Dekel and Fudenberg (1990) ntroduce one round elmnaton of weakly domnated actons followed by teratve elmnaton of strctly domnated actons as soluton concept. Let W S 1 denote the maxmal reducton of ths procedure. Proposton 5 For every player 2 I, R 1 W S 1. Proof. At the rst level, for all players the set of actons that are not weakly domnated are the set of lexcographc best reples. Ths follows from Blume, Brandenburger and Dekel (1991b, Proposton 1). Next, by Pearce (1984, Lemma 3), every acton not strctly domnated s a best reply to some conjecture over opponents actons and vce versa. Thus ratonalzablty and terated strct domnance concde. Thus the result follows from Proposton Comprehensve Ratonalzable Implementaton Let X denote the set of smply lotteres (.e. wth nte support) over an arbtrary set of alternatves. Each player 2 I has now preferences over lotteres represented by 2 Best reply refers to the pure acton best reply. 11
13 u : X R! R, where s a nte set of utlty parameters for player. Dstnct parameters n assocated dstnct preferences orderngs over X. Moreover, we assume that a player s never nd erent between all lotteres n X. The functon s lnear n ts rst argument. We assume that the preference pro le 2 = 2I s common knowledge among players n I. Yet, the socal planner does not know and want to mplement some lottery over alternatves n X. A socal choce functon f :! X assocates wth each preference pro le a lottery over alternatves. We consder a nte mechansm wth transfers hm 1 ; :::; M n ; g; t de ned by a nte acton set M for each player, an outcome functon g : M! X that assocates wth each acton pro le n M = 2I M a lottery over outcomes, and a transfer rule t = (t ) 2I : M! R n that for each player assocates wth each acton pro le a ne. (The second argument n each player s utlty functon refers to the nes. Less nes are preferred to more.) A mechansm hm 1 ; :::; M n ; g; t and a preference pro le n de ne a strategc game wth complete nformaton. A mechansm exactly mplements a socal choce functon f n comprehensve ratonalzable actons wth nes bounded by t > 0 f and only f jt(m)j t for all m 2 M, and for any 2, there exsts m ( ) 2 M such that g(m ( )) = f( ), t(m ( )) = 0, and R 1 ( ) = fm ( )g. A socal choce functon f s exactly mplementable n comprehensve ratonalzable actons wth small nes f for all t > 0, there exsts a mechansm whch exactly mplements f wth nes bounded by t. Proposton 6 Suppose that there are at least three players. functon s exactly mplementable n comprehensve ratonalzable actons. Then any socal choce Proof. Abreu and Matsushma (1994) show that any socal choce functon s exactly mplementable n terated admssble actons. In ther proof, they show that n fact t s exactly mplementable wth one round elmnaton of weakly domnated actons followed by many rounds of elmnaton of strctly domnated actons. In ther mechansm, the maxmal reductons of both procedures concde. Thus the result follows from Proposton 5 above. 12
14 5 Epstemc characterzaton of comprehensve ratonalzablty Our next goal wll be to provde an epstemc characterzaton of comprehensve ratonalzable strateges. To ths e ect we de ne lexcographc conjectures type spaces, construct the unversal space n the category of lexcographc conjectures type spaces, and de ne wthn t the graded event whose prmary part corresponds to the set of comprehensve ratonalzable strateges. 5.1 Lexcographc conjectures type spaces For a gven game wth a set of players I and a Hausdor space of strategy pro les S = Q 2I S ; T 2I s a lexcographc conjectures type space f 8 2 I; T = S n1 T n and each Tn s a Hausdor space, wth contnuous mappngs : T! S specfyng each type s strategy, de ned by the contnuous mappngs n : T n! S ; n 1 and the contnuous mappngs 3 : T! [ n1 T n specfyng each type s state of mnd regardng the other players types, as de ned by the contnuous mappngs n : Tn! T n ; n 1 havng the property that the n-tuple of belefs n (t n) js of t n 2 Tn on S de ned by n t n () js n t n j j6= 1 () 3 In what follows we wll use the notatonal conventon Y = Q j2j;j6= Y j : 13
15 consttutes a lexcographc conjecture over S (.e., these belefs are mutually sngular and the unon of ther supports s S ): Some (but not all) of the T n may be empty. Notce that n ths de nton, only the belefs of the types t n on the other players strateges S are requred to form a lexcographc conjecture; n contrast, the belefs of the types t n on the other players types T are not requred to have supports whose unon cover T. The motvaton for ths dstncton s that pror to playng the game, player may potentally get surprsng ver able evdence that her prmary belef on the other players strategy pro le S was wrong (.e. that the other players strategy pro les not ruled out by that evdence was assgned probablty zero by the prmary belef), n whch case she resorts to her secondary belef, and so forth. In contrast, no drect ver able evdence s feasble regardng the other players belefs. Hence, pror to playng the game there cannot arse a necessty for player to replace her prmary belef about the other players belefs, and therefore player need not necessarly entertan an exhaustve arsenal of mutually sngular alternatve belefs on the other players types. Ths does not preclude, of course, that a swtch of player to a secondary (or ternary, etc.) belef about the other players strateges may be correlated wth a correspondng swtch n belef about the other players types. Even though the lexcographc belef system (t ) need not have full support, the notons of n ntely more lkely and assumpton are de ned for (t ) verbatm. De nton 1 t assumes E f (t ) assumes E and for every s 2 (E ), (t ) deems ft 2 E j (t ) = s g as n ntely more lkely than ft =2 E j (t ) = s g. Lemma 2 In any lexcographc conjecture type space, for any measurable sets Y and Z wth Y \ Z = ; we have ft that deem Y n ntely more lkely as Zg s measurable. Proof. Let Y \ Y = ; and Y; Z be measurable. For any n, T n s measurable. For ` n the set B ;>0 ` (Y ) := ft 2 Tnj (t )`(Y ) > 0g as well as the set ft 2 Tnj (t )`(Z) = 0g are measurable. 14
16 Hence B I;`(Y; Z)n = ft nj (t n)`(y ) > 0; (t n) k (Y ) = 0; for k > `; (t n) j (Z) = 0; for j `g s measurable, snce t s an ntersecton of measurable sets. The set BI (Y; Z)n := ft n 2 Tnj (t n) deems Y as n ntely more lkely than Zg s S `=1;:::;n B I;`(Y; Z), and therefore measurable. Hence the set S k1 B I (Y; Z)n s measurable. Lemma 3 In any lexcographc conjecture type space: For any measurable event E T, the event ft jt assumes E g s measurable n T. Proof. For all s 2 (E ) the set ft j (t ) deems (t ) 1 (fs g)\e as n ntely more lkely than (t ) 1 (fs g) \ (T n E )g s measurable by Lemma 2. The set ft jt assumes E g s an ntersecton of the measurable set ft j (t ) assumes E g and the measurable sets (t ) 1 (fs g) \ E, where s 2 (E ). 5.2 The unversal lexcographc conjectures type space For a gven game wth a set of players I and strategy pro les S = Q 2I S ; de ne, nductvely, for each player 2 I: 0 = S 1;n = n s ; 1;` `n 1 = S n1 1;n o n : 1;` `n s a lexcographc conjecture on S Suppose, nductvely, that 8n 1 we have already de ned k;n and k. De ne k+1;n = ns ; 1;`; : : : ; k;`; k+1;` k+1 = S n1 k+1;n `n 2 k;n o n k : 8` n margj k 1 k+1;` = k;` In the lmt, de ne 1;n = n s ; 1;`; : : : `n o : s ; 1;`; : : : ; k;` 2 k;n 8k 0 `n 1 = S n1 1;n 15
17 By the Kolmogorov extenson theorem (the verson of Metver or Bourbak 1969), for each 4! 1;n s ; 1;`; : : : `n 2 1;n there exsts a unque tuple of n regular Borel probablty measures n! 1;n 2 ( (S 1 )) n such that 8k 0 and 8` n marg j k n! 1;n = ` k+1;` In partcular, from ths condton for k = 0, we get that s a lexcographc conjecture over S : marg js Moreover, the mappngs n are contnuous, and so are also the mappngs n : 1;n! S n! 1;n ` `n de ned by n! 1;n = s Hence h 1 2I s a lexcographc conjectures type space. The unversalty and completeness of h 1 2I follow the lnes of by-now standard arguments. ( Unversalty here refers to termnalty n the category of jij-player lexcographc conjectures type spaces based on the strategy pro les set S: ) Type möpse: Let ht 2I and h ^T 2I be lexcographc type spaces on S = Q 2I S. Then h : T! ^T contnuous s.t. T h! ^T & S. ^ commutes,.e. ^ (h (t )) = (t ) for all 2 I, for all t 2 T, such that h T n (Tn) ^T n and s contnuous for all n js j and such that (t ) (h ) 1 ( ^E ) = ^ (h (t ))( ^E ) for all measurable ^E ^T. Unversalty: Let h ^T 2I be any lexcographc conjecture type space on S. Let ^t 2 ^T n. 4 Each 1;n s non-empty t contans e.g. the n-tuple of heraches expressng player s belef that all the players lexcographc conjectures over K are commonly known. 16
18 De ne 0;n(^t ) := h 0;n(^t ) := (^t ), h 0;n : ^T Also de ne 1;n(^t ) :=! T 0;n. Ths s obvously contnuous. ^ (^t ); (( ) 1 ()). Ths s a lexcographc conjecture on `n S snce h ^T s a lexcographc conjecture type space. Suppose j m;n for j 2 I and for 0 m k have already been de ned s.t. h k;n := 0;n; 1;n; :::; k;n : ^T! T k; n s contnuous. Then de ne k+1;n va k+1;n(^t )` := ^ (^t )` (h k ) 1 (). k+1;n s contnuous f h k s contnuous, whch holds by the nducton hypothess. Coherency of Let k 1 and let proj jt 2 I that marg T have marg T k 1 k 1 0;n(^t ); 1;n(^t ); :::; k;n(^t ); ::: k 1 : T k! T k 1 k+1(^t )` = k(^t )`. Let E k 1 k+1(^t )`(E k ) = k+1(^t )` (proj jt ) 1 (E k 1 k 1 ) = ^ (^t )` (h k ) 1 ((proj jt k 1 = ^ (^t )` (proj jt h k ) 1 (E k 1 ) The last equaton follows from the fact that we have proj T (h j k 1 ) j6=. And therefore marg T ^T! T for all 2 I. k 1 k 1 be the projecton. We have to show for be some measurable event n Tk 1. We ) 1 (E k+1 )) k 1 h k = ^ (^t )`(h k 1 (E k 1 )): = (proj T j h j k ) j6= = k 1 k+1(^t ) = k(^t ). Hence we have shown that h := ( 0; 1; :::) : Snce the nte cylnder sets form an algebra that generates the -algebra on T for all 2 I and snce (h (^t )) and ^ (^t ) ((h ) 1 ()) are two -addtve probablty measures that - by constructon - concde on these cylnder sets, by the Caratheodory extenson theorem, they must be equal. Hence, snce by constructon (h (^t )) = ^ (^t ), (h ) 2I s a type mops from ^T := h ^T 2I to the unversal type space. Unqueness: Let := h 2I denote from now on the lexcographc conjecture type space constructon n ths secton (5.2) and for any lexcographc conjecture type space T := ht 2I let (w T ) 2I be the type mops from T to constructed before. A close examnaton of the de nton shows Remark 2 (w ) 2I = (d ) 2I. 17
19 Proposton 7 Let (h ) 2I be a type mops from the lexcographc conjecture type space T = ht 2I to ^T = h ^T 2I. Then we have w ^T h = w T for all 2 I. Proof. Analogous to Hefetz and Samet (1998, add reference to correspondng proposton/lemma) wth the obvous changes. Corollary 1 Let T be a lexcographc conjecture type space. Then (w T ) 2I s the unque type mops from T to. Proof. Let (h ) 2I be a type mops from T to. Then we have w T = w h = h. Proposton Remark. Remark 3 (After the constructon of the unversal type space.) The projectons proj k+1 k Proof. : k+1! k are onto. Analogous to the proof of Theorem 6 n Hefetz (1993) or the onto part of the p n n (3) of Theorem 1.1 Chapter III n Mertens, Sorn and Zamr (2003, p. 108). De nton 2 Let ^( ) be the set of all 2 (( ) n for n js j s.t. ((( ) 1 ())`)`n consttutes a lexcographc conjecture on S, endowed wth the relatve topology nherted from the dsjont unon topology of the S njs j (( )) n. Proposton 8 For all 2 I, :! ^( ) s a homeomorphsm. Proof. Smlar to Meer (forthcomng) usng unversalty or by drect examnaton of the maps n the unversal type space. 5.3 The epstemc characterzaton Let ht 2I be a type space for a gven game wth strategy pro les S = Q 2I S : De ne the followng sequence of events of player s Ratonalty and Mutual Assumpton of degree k 0 of Ratonalty: RMA 0 R = nt 2 T : t s a best reply to t o js 18
20 and nductvely RMA k+1 R = t 2 RMA k R : t assumes RMA k R Furthermore, de ne the event of s ratonalty and common assumpton of ratonalty to be 1\ RCAR = RMA k R k=0 Lemma 4 For k 0: If t 2 RMA k R, then (t ) js 1 assumes (RMA k 1 R ). Proof. For k = 0, there s nothng to prove. Suppose the clam s shown for all j 2 I and all m k. Let t 2 RMA k+1 R. By the nducton hypothess (t ) js assumes (RMA m R ) for all m < k. We have to show that (t ) js assumes (RMA k R ). Snce (t ) assumes RMA k R, there s an ndex ` 1 such that (t )`0 (RMA`0R ) = 1 for all `0 ` and (t )`0 (RMA`0R ) = 0 for all `0 > `. Ths mples that f (t )`0jS (fs g) > 0, for some `0 `, then s 2 (RMA k R ). Conversely, let s 2 (RMA k R ). By the second condton n the de nton of assumpton and the de nton of n ntely more lkely, (t )`0 (ft 2 RMA k R ; (t ) = s g) > 0 for some `0 and snce (t )`00(RMA k R ) = 0 for all `00 > `, we have `0 `. But ths mples that (t )`0jS (fs g) > 0. Together wth the rst part of the nducton step, we have that (RMA k R ) s assumed by (t )js. Theorem 1 (Epstemc characterzaton of comprehensve ratonalzablty) In the unversal lexcographc conjectures type space, for each k 0 the strateges (RMA k R ) played by the types n RMA k R are the strateges n R k R k = RMA k R and also n the lmt, the comprehensve ratonalzable strateges of player are the strateges played by s types n the event of s ratonalty and common assumpton of ratonalty: R1 = RCAR 19
21 Proof. Clearly (RMA 0 R ) R 0; snce by de nton 8t 2 RMA 0 R ; (t ) s a best reply to some 2 C 0; namely to = (t ) js : Conversely, let s 2 R 0: By de nton, s s a best reply to some lexcographc conjecture = ( 1; : : : n) 2 (S ) : Our am s hence to nd a type t 1 n the unversal lexcographc type space T 1 whose strategy (t 1) s s and whose margnal belef (t 1) js on S s : To ths e ect, let _ j 1 2 (S j ) be an arbtrary full-support belef of player j on her rvals strateges, j 2 I, and de ne nductvely _ j k+1 = _ j 0 1 ;::: _ j0 k j 0 6=j j s belef that there s k-mutual certanty that the players belefs on ther rvals strateges are _ j 1 ( s the Drac probablty measure). j2i De ne now t 1 = s ; ; _ 2; : : : _ k; : : : where _ 2 = _ 2;1; : : : ; _ 2;n s de ned by _ 2;` = ì _ j ; ` n 1 j6= and nductvely _ k = _ k;1; : : : ; _ k;n s de ned by _ k;` = _ k 1;` _ j ; ` n: k 1 j6= t 1 = s ; ; _ 2; : : : _ k; : : : s thus a type n RMA 0 R T1 whose strategy (t 1) = s s a best reply to ts margnal belef (t 1) js = on the other players strateges as requred. Suppose, by nducton, that 8 2 I we have already proved that 8m k ts s the case that R m = (RMA m R ), and hence that R m = Q j6= j (RMA m R j ). We now have to prove that R k+1 = (RMA k+1 R ). If t 2 RMA k R, then (t ) s a best reply to (t ) js. By the nducton hypothess and Lemma 4, f t 2 RMA k+1 R, then (t ) s a best reply to (t ) js, whch assumes (RMA m R ) = R m for all m k (and hence 20
22 (t ) js 1 s n C k+1 ). Thus, (t ) 2 R k+1. Conversely, let s 2 R k+1, we want to nd a t 2 RMA k+1 R such that (t ) = s. There s a 2 C k+1 such that s s a best reply to. By the nducton hypothess, (RMA k R ) = R k. has the form = ( 1 ; :::; n ) for some n 1. Recall that s full support and mutually sngular. Moreover, recall that S s nonempty and nte. Consder s = (s j ) j6=. There s exactly one ` n such that `(s ) > 0. For each j 6= choose a t j (s j ) such that j (t j (s j )) = s j and t j (s j ) 2 RMA m R j for the maxmal m k among the ^t j wth j (^t j ) = s j. Now let t be the unque ^t 2 T such that (^t ) s a xed best reply to and such that (^t )` := P s 2S `(s ) t (s )() for all ` = 1; :::; n. Ths measure assumes RMA m R for all m k, t sats es the second condton of the de nton of RMA k+1 R for all m k and ts margnal on S s. And (t ) s a best reply to. That R 1 = (RMA 1 ) s proved n Lemma 6. Constructon: Let k be such that C k+1 = C k and R k+1 = R k for all 2 I. For each s 2 R k choose 1;n(s ) (s ) 2 C k for some n(s ) wth s a lexcographc best reply to 1;n(s ) (s ) = ( 1;1; :::; 1;n). Smplfy noton by wrtng 1(s ) for 1;n(s ). Lkewse, for any s =2 R k, chose a (s ) 2 C m wth m maxmal such that s s a lexcographc best reply to (s ), f such a (s ) exsts. Otherwse let (s ) be any lexcographc conjecture. De ne 2;n(s ) such that 2(s )() := 2;n(s )`() := P s 2S 1;`(s )(fs g) (s j ; j 1 (sj )) j6= (). By de nton marg js 2;n(s )`() = 1;`(s )(). Fx k 1. Assume for each 2 I and s 2 S, we have already de ned by nducton t ;m (s ) n(s ) = (s ; ( 1;`(s ); :::; m;`(s ))`n(s ) 2 T m;n(s ). Then de ne m+1;`(s )() := P s 2S 1;`(s )(fs g) (s j ;t j;m+1 (s j )) j6= (). By the nducton hypothess m;`(s ) = P s 2S 1;`(s )(fs g) (s j ;t ;m 1 (s j )) j6= and we have that marg jt ;m (s j ;t j;m (s j )) j6= () = (s j ;t j;m 1 (s j )) j6= () snce t j;m (s j ) extends t j;m 1 (s j ) by constructon. If we let t (s ) = (s ; ( 1;`(s ); :::)`n(s )) then we get by constructon of the unversal type space that (t (s ))` = P s 2S 1;`(s )(fs g) (s j ;t j (s j )) j6= (). By de nton: f s 2 R m, then 1;n(s ) assumes C p for all p < m. Therefore, by construc- 21
23 ton of the t (s ), (t (s )) assumes ft (s )js 2 Rp g =: t (Rp ). By also, by constructon, f s 2 (t (Rp )) = Rp, then t (s ) deems ft ( s ^ )j (^s ) = s and ^s 2 Rp g = ft (s )g as n ntely more lkely than ft =2 t (Rp )j (t ) = s g. Note, ths latter set gets by constructon probablty 0 at every level of (t (s )), whle (t (s ))`(ft (s )g) > 0 f and only f 1;`(s )(fs g) > 0. But snce 1;n(s ) s a lexcographc conjecture on S ths happens for some ` n(s ). We have just shown the followng: Lemma 5 For all 2 I, for all m 0, for all p < m, and for all s 2 R m, t (s ) assumes ft (s )js 2 R p g. Lemma 6 For all 2 I, R 1 = (RMA 1 R ). Proof. We now prove by nducton the followng clam: For all m 0: s 2 R m f and only f t (s ) 2 RMA m R. The ( drecton follows already from the rst part of the proof of the Theorem. We now prove the ) drecton. If m = 0, by constructon s s a best reply to t (s ) s rst order belef. Observe that by constructon (t (s ))` (ft (s )js 2 S g) = 1 for all ` n(s ). Now x any s 2 R m+1. By the nducton hypothess, for all j: ft j (s j )js j 2 R j mg = ft j (s j )jt j (s j ) 2 RMA m R j g. Now, by the above observaton, for all ` n(s ) we have (t (s ))` (ft (s )js 2 R m g) = (t (s ))`(RMA m R ). But by Lemma t (s ) assumes (ft (s )js 2 Rm g), and hence assumes RMA m R. Snce t also assumes RMA m 1 R by the nducton hypothess (for m 1), we have shown that t (s ) 2 RMA m+1 R. To sum up, snce Rk = R m for m k for all 2 I, and snce each Rk 6= ; and snce the clam holds we have s 2 R1 f and only f Rm for m k f and only f t (s ) 2 RMA m R for all m k f and only f t (s ) 2 RMA 1 R and snce t (s ) plays s, we have that R 1 (RMA 1 R ). On the other hand, f t 2 RMA 1 R t follows that t 2 RMA m R for all m 0 and therefore (t ) 2 R m for all m 0 by the rst part of the proof of the theorem. But (t ) 2 R m for all m 0 mples (t ) 2 R 1. Hence we have show that (RMA 1 R ) = R 1 for all 2 I. 22
24 References [1] Abreu, D. and H. Matsushma (1994). Exact mplementaton, Journal of Economc Theory 64, [2] Bernhem, B.D. (1984). Ratonalzable strategc behavor, Econometrca 52, [3] Blume, L., Brandenburger, A., and E. Dekel (1991a). Lexcographc probabltes and choce under uncertanty, Econometrca 59, [4] Blume, L., Brandenburger, A., and E. Dekel (1991b). Lexcographc probabltes and equlbrum re nements, Econometrca 59, [5] Brandenburger, A., Fredenberg, A. and J. Kesler (2008). Admssblty n games, Econometrca 76, [6] Brams, S.J., Klgour, D.M., and M.D. Davs (1993). Unravelng n games of sharng and exchange, n: Bnmore, K.G., Krman, A., and P. Tan (eds.), Fronters of Game Theory, MIT Press, [7] Meer, M. (2001). unpublshed paper. [8] Mertens, J.F., Sorn, S., and S. Zamr (2003). Repeated games, mmeo. [9] Mouln, H. (1986). Game theory for the socal scences, 2nd edton, New York Unversty Press. [10] Harstad, R. and D. Levn (1985). A class of domnance-solvable common value auctons, Revew of Economc Studes 52, [11] Hefetz, A. (1993). The Bayesan formulaton of ncomplete nformaton - The noncompact case, Internatonal Journal of Game Theory 21, [12] Hefetz, A. and D. Samet (1998). Topology-free typology of belefs, Journal of Economc Theory 82, [13] Osborne, M. (2004). An ntroducton to game theory, Oxford Unversty Press. [14] Pearce, D. (1984). Ratonalzable strategc behavor and the problem of perfecton, Econometrca 52,
25 [15] Spohn, W. (1982). How to make sense of game theory, n: Stegmüller, W., Balzer, W., and W. Spohn (eds.), Phlosophy of economcs, Sprnger-Verlag, p [16] Stahl, D.O. (1995). Lexcographc ratonalzablty and terated admssblty, Economcs Letters 47,
Genericity of Critical Types
Genercty of Crtcal Types Y-Chun Chen Alfredo D Tllo Eduardo Fangold Syang Xong September 2008 Abstract Ely and Pesk 2008 offers an nsghtful characterzaton of crtcal types: a type s crtcal f and only f
More informationInterim Correlated Rationalizability 1
Interm Correlated Ratonalzablty Edde Dekel Northwestern Unversty and Tel Avv Unversty Drew Fudenberg Harvard Unversty Stephen Morrs Prnceton Unversty Frst Draft: May 2003. Ths Draft: November 2006 Ths
More informationRATIONALIZABLE IMPLEMENTATION. Dirk Bergemann and Stephen Morris. May 2009 COWLES FOUNDATION DISCUSSION PAPER NO. 1697
RATIONALIZABLE IMPLEMENTATION By Drk Bergemann and Stephen Morrs May 2009 COWLES FOUNDATION DISCUSSION PAPER NO. 1697 COWLES FOUNDATION FOR RESEARCH IN ECONOMICS YALE UNIVERSITY Box 208281 New Haven, Connectcut
More informationSubjective Uncertainty Over Behavior Strategies: A Correction
Subjectve Uncertanty Over Behavor Strateges: A Correcton The Harvard communty has made ths artcle openly avalable. Please share how ths access benefts you. Your story matters. Ctaton Publshed Verson Accessed
More informationRobust Implementation: The Role of Large Type Spaces
Robust Implementaton: The Role of Large Type Spaces Drk Bergemann y Stephen Morrs z Frst Verson: March 23 Ths Verson: June 25 Abstract A socal choce functon s robustly mplemented f every equlbrum on every
More informationRobust Implementation: The Role of Large Type Spaces
Robust Implementaton: The Role of Large Type Spaces Drk Bergemann y Stephen Morrs z Frst Verson: March 2003 Ths Verson: Aprl 2004 Abstract We analyze the problem of fully mplementng a socal choce functon
More informationMicroeconomics: Auctions
Mcroeconomcs: Auctons Frédérc Robert-coud ovember 8, Abstract We rst characterze the PBE n a smple rst prce and second prce sealed bd aucton wth prvate values. The key result s that the expected revenue
More informationTopologies on Types: Connections
Topologes on Types: Connectons Y-Chun Chen y Syang Xong z May 28, 2008 Abstract For d erent purposes, economsts may use d erent topologes on types. We characterze the relatonshp among these varous topologes.
More informationEconomics 101. Lecture 4 - Equilibrium and Efficiency
Economcs 0 Lecture 4 - Equlbrum and Effcency Intro As dscussed n the prevous lecture, we wll now move from an envronment where we looed at consumers mang decsons n solaton to analyzng economes full of
More informationInterim Rationalizability. Eddie Dekel, Drew Fudenberg and Stephen Morris. Working Paper No September, 2005
Interm Ratonalzablty By Edde Dekel, Drew Fudenberg and Stephen Morrs Workng Paper No. 6-2005 September, 2005 The Foerder Insttute for Economc Research and The Sackler Insttute of Economc Studes Interm
More informationImplementation in Mixed Nash Equilibrium
Department of Economcs Workng Paper Seres Implementaton n Mxed Nash Equlbrum Claudo Mezzett & Ludovc Renou May 2012 Research Paper Number 1146 ISSN: 0819-2642 ISBN: 978 0 7340 4496 9 Department of Economcs
More informationFACTORIZATION IN KRULL MONOIDS WITH INFINITE CLASS GROUP
C O L L O Q U I U M M A T H E M A T I C U M VOL. 80 1999 NO. 1 FACTORIZATION IN KRULL MONOIDS WITH INFINITE CLASS GROUP BY FLORIAN K A I N R A T H (GRAZ) Abstract. Let H be a Krull monod wth nfnte class
More informationContinuous Implementation
Contnuous Implementaton Maron Oury HEC Pars Olver Terceux Pars School of Economcs and CNRS Abstract It s well-known that mechansm desgn lterature makes many smplfyng nformatonal assumptons n partcular
More informationBayesian epistemology II: Arguments for Probabilism
Bayesan epstemology II: Arguments for Probablsm Rchard Pettgrew May 9, 2012 1 The model Represent an agent s credal state at a gven tme t by a credence functon c t : F [0, 1]. where F s the algebra of
More informationImplementation and Detection
1 December 18 2014 Implementaton and Detecton Htosh Matsushma Department of Economcs Unversty of Tokyo 2 Ths paper consders mplementaton of scf: Mechansm Desgn wth Unqueness CP attempts to mplement scf
More informationj) = 1 (note sigma notation) ii. Continuous random variable (e.g. Normal distribution) 1. density function: f ( x) 0 and f ( x) dx = 1
Random varables Measure of central tendences and varablty (means and varances) Jont densty functons and ndependence Measures of assocaton (covarance and correlaton) Interestng result Condtonal dstrbutons
More informationMath 205A Homework #2 Edward Burkard. Assume each composition with a projection is continuous. Let U Y Y be an open set.
Math 205A Homework #2 Edward Burkard Problem - Determne whether the topology T = fx;?; fcg ; fa; bg ; fa; b; cg ; fa; b; c; dgg s Hausdor. Choose the two ponts a; b 2 X. Snce there s no two dsjont open
More informationCowles Foundation for Research in Economics at Yale University
Cowles Foundaton for Research n Economcs at Yale Unversty Cowles Foundaton Dscusson Paper No. 666 ROBUST IMPLEMENTATION IN GENERAL MECHANISMS Drk Bergemann and Stephen Morrs Month 28 An author ndex to
More informationCMS-EMS Center for Mathematical Studies in Economics And Management Science. Discussion Paper #1586
CMS-EMS Center for Mathematcal Studes n Economcs And Management Scence Dscusson Paper #1586 "When Do Types Induce the Same Belef Herarchy?" Andres Perea and Wllemen Kets Northwestern Unversty December
More informationAPPENDIX A Some Linear Algebra
APPENDIX A Some Lnear Algebra The collecton of m, n matrces A.1 Matrces a 1,1,..., a 1,n A = a m,1,..., a m,n wth real elements a,j s denoted by R m,n. If n = 1 then A s called a column vector. Smlarly,
More informationHierarchies of Beliefs and the Belief-invariant. Bayesian Solution
Herarches of Belefs and the Belef-nvarant Bayesan Soluton Qanfeng Tang March 26, 2015 Abstract The belef-nvarant Bayesan soluton s a noton of correlated equlbrum n games wth ncomplete nformaton proposed
More informationPerfect Competition and the Nash Bargaining Solution
Perfect Competton and the Nash Barganng Soluton Renhard John Department of Economcs Unversty of Bonn Adenauerallee 24-42 53113 Bonn, Germany emal: rohn@un-bonn.de May 2005 Abstract For a lnear exchange
More informationHIERARCHIES OF BELIEF AND INTERIM RATIONALIZABILITY
HIERARCHIES OF BELIEF AND INTERIM RATIONALIZABILITY JEFFREY C. ELY AND MARCIN PESKI Abstract. In games wth ncomplete nformaton, conventonal herarches of belef are ncomplete as descrptons of the players
More informationA note on the one-deviation property in extensive form games
Games and Economc Behavor 40 (2002) 322 338 www.academcpress.com Note A note on the one-devaton property n extensve form games Andrés Perea Departamento de Economía, Unversdad Carlos III de Madrd, Calle
More informationLinear, affine, and convex sets and hulls In the sequel, unless otherwise specified, X will denote a real vector space.
Lnear, affne, and convex sets and hulls In the sequel, unless otherwse specfed, X wll denote a real vector space. Lnes and segments. Gven two ponts x, y X, we defne xy = {x + t(y x) : t R} = {(1 t)x +
More informationHigher Order Beliefs in Dynamic Envinronments.
Hgher Order Belefs n Dynamc Envnronments. Antono Penta y Dept. of Economcs, Unversty of Pennsylvana May 2008 (prelmnary and ncomplete verson) Abstract Ths paper explores the role of hgher order belefs
More informationPROCEEDINGS OF THE AMERICAN MATHEMATICAL SOCIETY Volume 125, Number 7, July 1997, Pages 2119{2125 S (97) THE STRONG OPEN SET CONDITION
PROCDINGS OF TH AMRICAN MATHMATICAL SOCITY Volume 125, Number 7, July 1997, Pages 2119{2125 S 0002-9939(97)03816-1 TH STRONG OPN ST CONDITION IN TH RANDOM CAS NORBRT PATZSCHK (Communcated by Palle. T.
More informationTHE CHINESE REMAINDER THEOREM. We should thank the Chinese for their wonderful remainder theorem. Glenn Stevens
THE CHINESE REMAINDER THEOREM KEITH CONRAD We should thank the Chnese for ther wonderful remander theorem. Glenn Stevens 1. Introducton The Chnese remander theorem says we can unquely solve any par of
More informationEpistemic Game Theory: Beliefs and Types
Epstemc Game Theory: Belefs and Types Marcano Snscalch March 28, 2007 1 Introducton John Harsany [19] ntroduced the formalsm of type spaces to provde a smple and parsmonous representaton of belef herarches.
More informationAlgorithms for cautious reasoning in games
Algorthms for cautous reasonng n games Ger B. Ashem a Andrés Perea b 12 October 2018 Abstract We provde comparable algorthms for the Dekel-Fudenberg procedure, terated admssblty, proper ratonalzablty and
More informationInfinitely Split Nash Equilibrium Problems in Repeated Games
Infntely Splt ash Equlbrum Problems n Repeated Games Jnlu L Department of Mathematcs Shawnee State Unversty Portsmouth, Oho 4566 USA Abstract In ths paper, we ntroduce the concept of nfntely splt ash equlbrum
More informationProblem Set 9 Solutions
Desgn and Analyss of Algorthms May 4, 2015 Massachusetts Insttute of Technology 6.046J/18.410J Profs. Erk Demane, Srn Devadas, and Nancy Lynch Problem Set 9 Solutons Problem Set 9 Solutons Ths problem
More informationExercise Solutions to Real Analysis
xercse Solutons to Real Analyss Note: References refer to H. L. Royden, Real Analyss xersze 1. Gven any set A any ɛ > 0, there s an open set O such that A O m O m A + ɛ. Soluton 1. If m A =, then there
More informationA combinatorial problem associated with nonograms
A combnatoral problem assocated wth nonograms Jessca Benton Ron Snow Nolan Wallach March 21, 2005 1 Introducton. Ths work was motvated by a queston posed by the second named author to the frst named author
More informationStrategic Distinguishability and Robust Virtual Implementation
Strategc Dstngushablty and Robust Vrtual Implementaton Drk Bergemann y Stephen Morrs z Frst Verson: March 2006 Ths Verson: Aprl 2008 Abstract In a general nterdependent preference envronment, we characterze
More information20. Mon, Oct. 13 What we have done so far corresponds roughly to Chapters 2 & 3 of Lee. Now we turn to Chapter 4. The first idea is connectedness.
20. Mon, Oct. 13 What we have done so far corresponds roughly to Chapters 2 & 3 of Lee. Now we turn to Chapter 4. The frst dea s connectedness. Essentally, we want to say that a space cannot be decomposed
More informationCorrelated Equilibrium in Games with Incomplete Information
Correlated Equlbrum n Games wth Incomplete Informaton Drk Bergemann y Stephen Morrs z Frst Verson: October Current Verson: May 4, Abstract We de ne a noton of correlated equlbrum for games wth ncomplete
More informationThe basic point with mechanism design is that it allows a distinction between the underlying
14 Mechansm Desgn The basc pont wth mechansm desgn s that t allows a dstncton between the underlyng economc envronment and the rules of the game. We wll take as gven some set of possble outcomes (alternatves,
More informationModule 17: Mechanism Design & Optimal Auctions
Module 7: Mechansm Desgn & Optmal Auctons Informaton Economcs (Ec 55) George Georgads Examples: Auctons Blateral trade Producton and dstrbuton n socety General Setup N agents Each agent has prvate nformaton
More informationThe Second Anti-Mathima on Game Theory
The Second Ant-Mathma on Game Theory Ath. Kehagas December 1 2006 1 Introducton In ths note we wll examne the noton of game equlbrum for three types of games 1. 2-player 2-acton zero-sum games 2. 2-player
More informationCritical Types. September 24, Abstract
Crtcal Types Jeffrey C. Ely Marcn Pesk September 24, 2010 Abstract How can we know n advance whether smplfyng assumptons about belefs wll make a dfference n the conclusons of game-theoretc models? We defne
More informationDIFFERENTIAL SCHEMES
DIFFERENTIAL SCHEMES RAYMOND T. HOOBLER Dedcated to the memory o Jerry Kovacc 1. schemes All rngs contan Q and are commutatve. We x a d erental rng A throughout ths secton. 1.1. The topologcal space. Let
More informationHierarchies of belief and interim rationalizability
Theoretcal Economcs 1 (2006), 19 65 1555-7561/20060019 Herarches of belef and nterm ratonalzablty JEFFREY C. ELY Department of Economcs, Northwestern Unversty MARCIN PESKI Department of Economcs, Unversty
More informationEndogenous timing in a mixed oligopoly consisting of a single public firm and foreign competitors. Abstract
Endogenous tmng n a mxed olgopoly consstng o a sngle publc rm and oregn compettors Yuanzhu Lu Chna Economcs and Management Academy, Central Unversty o Fnance and Economcs Abstract We nvestgate endogenous
More informationThe Folk Theorem for Games with Private Almost-Perfect Monitoring
The Folk Theorem for Games wth Prvate Almost-Perfect Montorng Johannes Hörner y Wojcech Olszewsk z October 2005 Abstract We prove the folk theorem for dscounted repeated games under prvate, almost-perfect
More informationMaximizing the number of nonnegative subsets
Maxmzng the number of nonnegatve subsets Noga Alon Hao Huang December 1, 213 Abstract Gven a set of n real numbers, f the sum of elements of every subset of sze larger than k s negatve, what s the maxmum
More informationBackward Induction Reasoning in Games with Incomplete Information
Backward Inducton Reasonng n Games wth Incomplete Informaton Antono Penta y Unversty of Wsconsn - Madson, Dept. of Economcs Ths verson: September 26, 2011 Abstract Backward Inducton s one of the central
More informationGames of Threats. Elon Kohlberg Abraham Neyman. Working Paper
Games of Threats Elon Kohlberg Abraham Neyman Workng Paper 18-023 Games of Threats Elon Kohlberg Harvard Busness School Abraham Neyman The Hebrew Unversty of Jerusalem Workng Paper 18-023 Copyrght 2017
More informationON THE EQUIVALENCE OF ORDINAL BAYESIAN INCENTIVE COMPATIBILITY AND DOMINANT STRATEGY INCENTIVE COMPATIBILITY FOR RANDOM RULES
ON THE EQUIVALENCE OF ORDINAL BAYESIAN INCENTIVE COMPATIBILITY AND DOMINANT STRATEGY INCENTIVE COMPATIBILITY FOR RANDOM RULES Madhuparna Karmokar 1 and Souvk Roy 1 1 Economc Research Unt, Indan Statstcal
More information2.3 Nilpotent endomorphisms
s a block dagonal matrx, wth A Mat dm U (C) In fact, we can assume that B = B 1 B k, wth B an ordered bass of U, and that A = [f U ] B, where f U : U U s the restrcton of f to U 40 23 Nlpotent endomorphsms
More informationPricing Network Services by Jun Shu, Pravin Varaiya
Prcng Network Servces by Jun Shu, Pravn Varaya Presented by Hayden So September 25, 2003 Introducton: Two Network Problems Engneerng: A game theoretcal sound congeston control mechansm that s ncentve compatble
More informationAnti-van der Waerden numbers of 3-term arithmetic progressions.
Ant-van der Waerden numbers of 3-term arthmetc progressons. Zhanar Berkkyzy, Alex Schulte, and Mchael Young Aprl 24, 2016 Abstract The ant-van der Waerden number, denoted by aw([n], k), s the smallest
More informationVickrey Auction VCG Combinatorial Auctions. Mechanism Design. Algorithms and Data Structures. Winter 2016
Mechansm Desgn Algorthms and Data Structures Wnter 2016 1 / 39 Vckrey Aucton Vckrey-Clarke-Groves Mechansms Sngle-Mnded Combnatoral Auctons 2 / 39 Mechansm Desgn (wth Money) Set A of outcomes to choose
More informationDIEGO AVERNA. A point x 2 X is said to be weakly Pareto-optimal for the function f provided
WEAKLY PARETO-OPTIMAL ALTERNATIVES FOR A VECTOR MAXIMIZATION PROBLEM: EXISTENCE AND CONNECTEDNESS DIEGO AVERNA Let X be a non-empty set and f =f 1 ::: f k ):X! R k a functon. A pont x 2 X s sad to be weakly
More informationGame Theory. Lecture Notes By Y. Narahari. Department of Computer Science and Automation Indian Institute of Science Bangalore, India February 2008
Game Theory Lecture Notes By Y. Narahar Department of Computer Scence and Automaton Indan Insttute of Scence Bangalore, Inda February 2008 Chapter 10: Two Person Zero Sum Games Note: Ths s a only a draft
More informationRobust virtual implementation
Theoretcal Economcs 4 (2009), 45 88 1555-7561/20090045 Robust vrtual mplementaton DIRK BERGEMANN Department of Economcs, Yale Unversty STEPHEN MORRIS Department of Economcs, Prnceton Unversty In a general
More informationCS286r Assign One. Answer Key
CS286r Assgn One Answer Key 1 Game theory 1.1 1.1.1 Let off-equlbrum strateges also be that people contnue to play n Nash equlbrum. Devatng from any Nash equlbrum s a weakly domnated strategy. That s,
More informationFirst day August 1, Problems and Solutions
FOURTH INTERNATIONAL COMPETITION FOR UNIVERSITY STUDENTS IN MATHEMATICS July 30 August 4, 997, Plovdv, BULGARIA Frst day August, 997 Problems and Solutons Problem. Let {ε n } n= be a sequence of postve
More informationIncomplete Information and Robustness in Strategic Environments
Unversty of Pennsylvana ScholarlyCommons Publcly Accessble Penn Dssertatons Sprng 5-17-2010 Incomplete Informaton and Robustness n Strategc Envronments Antono Penta Unversty of Pennsylvana, apenta@ssc.wsc.edu
More information2E Pattern Recognition Solutions to Introduction to Pattern Recognition, Chapter 2: Bayesian pattern classification
E395 - Pattern Recognton Solutons to Introducton to Pattern Recognton, Chapter : Bayesan pattern classfcaton Preface Ths document s a soluton manual for selected exercses from Introducton to Pattern Recognton
More informationGames in Preference Form and Preference Rationalizability
Games n Preference Form and Preference Ratonalzablty Stephen Morrs Satoru Takahash September 8, 2012 Abstract We ntroduce a game n preference form, whch conssts of a game form and a preference structure,
More informationCOWLES FOUNDATION FOR RESEARCH IN ECONOMICS YALE UNIVERSITY
Best Response Equvalence by Stephen Morrs and Takash U July 2002 COWLES FOUNDATION DISCUSSION PAPER NO. 1377 COWLES FOUNDATION FOR RESEARCH IN ECONOMICS YALE UNIVERSITY Box 208281 New Haven, Connectcut
More informationConvexity preserving interpolation by splines of arbitrary degree
Computer Scence Journal of Moldova, vol.18, no.1(52), 2010 Convexty preservng nterpolaton by splnes of arbtrary degree Igor Verlan Abstract In the present paper an algorthm of C 2 nterpolaton of dscrete
More informationPayoff Information and. Self-Confirming Equilibrium 1
Payoff Informaton and Self-Confrmng Equlbrum 1 Frst verson: Aprl 25, 1995 Ths revson: July 12, 1999 Edde Deel Drew Fudenberg Davd K. Levne 2 1 We than Faru Gul, an assocate edtor, two referees, and semnar
More informationRyan (2009)- regulating a concentrated industry (cement) Firms play Cournot in the stage. Make lumpy investment decisions
1 Motvaton Next we consder dynamc games where the choce varables are contnuous and/or dscrete. Example 1: Ryan (2009)- regulatng a concentrated ndustry (cement) Frms play Cournot n the stage Make lumpy
More informationCOWLES FOUNDATION FOR RESEARCH IN ECONOMICS YALE UNIVERSITY
ROBUST MECHANISM DESIGN By Drk Bergemann and Stephen Morrs Aprl 2004 COWLES FOUNDATION DISCUSSION PAPER NO. 1421R COWLES FOUNDATION FOR RESEARCH IN ECONOMICS YALE UNIVERSITY Box 208281 New Haven, Connectcut
More informationPerron Vectors of an Irreducible Nonnegative Interval Matrix
Perron Vectors of an Irreducble Nonnegatve Interval Matrx Jr Rohn August 4 2005 Abstract As s well known an rreducble nonnegatve matrx possesses a unquely determned Perron vector. As the man result of
More informationOnline Appendix. t=1 (p t w)q t. Then the first order condition shows that
Artcle forthcomng to ; manuscrpt no (Please, provde the manuscrpt number!) 1 Onlne Appendx Appendx E: Proofs Proof of Proposton 1 Frst we derve the equlbrum when the manufacturer does not vertcally ntegrate
More informationMore metrics on cartesian products
More metrcs on cartesan products If (X, d ) are metrc spaces for 1 n, then n Secton II4 of the lecture notes we defned three metrcs on X whose underlyng topologes are the product topology The purpose of
More informationCommon Belief Foundations of Global Games
Common Belef Foundatons of Global Games Stephen Morrs Prnceton Unversty smorrs@prnceton.edu Hyun Song Shn Prnceton Unversty hsshn@prnceton.edu October 2007 Abstract We provde a characterzaton of when an
More information(1 ) (1 ) 0 (1 ) (1 ) 0
Appendx A Appendx A contans proofs for resubmsson "Contractng Informaton Securty n the Presence of Double oral Hazard" Proof of Lemma 1: Assume that, to the contrary, BS efforts are achevable under a blateral
More informationThe Order Relation and Trace Inequalities for. Hermitian Operators
Internatonal Mathematcal Forum, Vol 3, 08, no, 507-57 HIKARI Ltd, wwwm-hkarcom https://doorg/0988/mf088055 The Order Relaton and Trace Inequaltes for Hermtan Operators Y Huang School of Informaton Scence
More informationYong Joon Ryang. 1. Introduction Consider the multicommodity transportation problem with convex quadratic cost function. 1 2 (x x0 ) T Q(x x 0 )
Kangweon-Kyungk Math. Jour. 4 1996), No. 1, pp. 7 16 AN ITERATIVE ROW-ACTION METHOD FOR MULTICOMMODITY TRANSPORTATION PROBLEMS Yong Joon Ryang Abstract. The optmzaton problems wth quadratc constrants often
More information3.1 Expectation of Functions of Several Random Variables. )' be a k-dimensional discrete or continuous random vector, with joint PMF p (, E X E X1 E X
Statstcs 1: Probablty Theory II 37 3 EPECTATION OF SEVERAL RANDOM VARIABLES As n Probablty Theory I, the nterest n most stuatons les not on the actual dstrbuton of a random vector, but rather on a number
More informationRandić Energy and Randić Estrada Index of a Graph
EUROPEAN JOURNAL OF PURE AND APPLIED MATHEMATICS Vol. 5, No., 202, 88-96 ISSN 307-5543 www.ejpam.com SPECIAL ISSUE FOR THE INTERNATIONAL CONFERENCE ON APPLIED ANALYSIS AND ALGEBRA 29 JUNE -02JULY 20, ISTANBUL
More informationNP-Completeness : Proofs
NP-Completeness : Proofs Proof Methods A method to show a decson problem Π NP-complete s as follows. (1) Show Π NP. (2) Choose an NP-complete problem Π. (3) Show Π Π. A method to show an optmzaton problem
More informationThis article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and
Ths artcle appeared n a journal publshed by Elsever. The attached copy s furnshed to the author for nternal non-commercal research and educaton use, ncludng for nstructon at the authors nsttuton and sharng
More informationk(k 1)(k 2)(p 2) 6(p d.
BLOCK-TRANSITIVE 3-DESIGNS WITH AFFINE AUTOMORPHISM GROUP Greg Gamble Let X = (Z p d where p s an odd prme and d N, and let B X, B = k. Then t was shown by Praeger that the set B = {B g g AGL d (p} s the
More informationSL n (F ) Equals its Own Derived Group
Internatonal Journal of Algebra, Vol. 2, 2008, no. 12, 585-594 SL n (F ) Equals ts Own Derved Group Jorge Macel BMCC-The Cty Unversty of New York, CUNY 199 Chambers street, New York, NY 10007, USA macel@cms.nyu.edu
More informationCommon Belief Foundations of Global Games
Common Belef Foundatons of Global Games Stephen Morrs Prnceton Unversty Hyun Song Shn Bank for Internatonal Settlements November 2015 Muhamet Yldz M.I.T. Abstract We study coordnaton games under general
More informationEdge Isoperimetric Inequalities
November 7, 2005 Ross M. Rchardson Edge Isopermetrc Inequaltes 1 Four Questons Recall that n the last lecture we looked at the problem of sopermetrc nequaltes n the hypercube, Q n. Our noton of boundary
More informationDeriving the X-Z Identity from Auxiliary Space Method
Dervng the X-Z Identty from Auxlary Space Method Long Chen Department of Mathematcs, Unversty of Calforna at Irvne, Irvne, CA 92697 chenlong@math.uc.edu 1 Iteratve Methods In ths paper we dscuss teratve
More informationInternational Journal of Algebra, Vol. 8, 2014, no. 5, HIKARI Ltd,
Internatonal Journal of Algebra, Vol. 8, 2014, no. 5, 229-238 HIKARI Ltd, www.m-hkar.com http://dx.do.org/10.12988/ja.2014.4212 On P-Duo odules Inaam ohammed Al Had Department of athematcs College of Educaton
More informationThe lower and upper bounds on Perron root of nonnegative irreducible matrices
Journal of Computatonal Appled Mathematcs 217 (2008) 259 267 wwwelsevercom/locate/cam The lower upper bounds on Perron root of nonnegatve rreducble matrces Guang-Xn Huang a,, Feng Yn b,keguo a a College
More informationRobust Implementation under Complete Information
Robust Implementaton under Complete Informaton Rene Saran (Prelmnary: Please do not cte) Abstract Game-theoretc soluton concepts rely on strong assumptons regardng players knowledge of each other s ratonalty
More information12 MATH 101A: ALGEBRA I, PART C: MULTILINEAR ALGEBRA. 4. Tensor product
12 MATH 101A: ALGEBRA I, PART C: MULTILINEAR ALGEBRA Here s an outlne of what I dd: (1) categorcal defnton (2) constructon (3) lst of basc propertes (4) dstrbutve property (5) rght exactness (6) localzaton
More informationWeek 2. This week, we covered operations on sets and cardinality.
Week 2 Ths week, we covered operatons on sets and cardnalty. Defnton 0.1 (Correspondence). A correspondence between two sets A and B s a set S contaned n A B = {(a, b) a A, b B}. A correspondence from
More informationLinear Algebra and its Applications
Lnear Algebra and ts Applcatons 4 (00) 5 56 Contents lsts avalable at ScenceDrect Lnear Algebra and ts Applcatons journal homepage: wwwelsevercom/locate/laa Notes on Hlbert and Cauchy matrces Mroslav Fedler
More informationOn C 0 multi-contractions having a regular dilation
SUDIA MAHEMAICA 170 (3) (2005) On C 0 mult-contractons havng a regular dlaton by Dan Popovc (mşoara) Abstract. Commutng mult-contractons of class C 0 and havng a regular sometrc dlaton are studed. We prove
More informationMarket structure and Innovation
Market structure and Innovaton Ths presentaton s based on the paper Market structure and Innovaton authored by Glenn C. Loury, publshed n The Quarterly Journal of Economcs, Vol. 93, No.3 ( Aug 1979) I.
More information2 MADALINA ROXANA BUNECI subset G (2) G G (called the set of composable pars), and two maps: h (x y)! xy : G (2)! G (product map) x! x ;1 [: G! G] (nv
An applcaton of Mackey's selecton lemma Madalna Roxana Bunec Abstract. Let G be a locally compact second countable groupod. Let F be a subset of G (0) meetng each orbt exactly once. Let us denote by df
More informationMinimal Assumptions for E ciciency in Asymmetric English Auctions.
Mnmal ssumptons for E ccency n symmetrc Englsh uctons. Juan Dubra y Unversdad de Montevdeo and Unversdad Torcuato D Tella September, 2006 bstract I ntroduce a property of player s valuatons that ensures
More informationAn Explicit Approach to Modeling Finite-Order Type Spaces and Applications
An Explct Approach to Modelng Fnte-Order Type Spaces and Applcatons Cheng-Zhong Qn and Chun-Le Yang November 18, 2009 Abstract Every abstract type of a belef-closed type space corresponds to an nfnte belef
More informationLectures - Week 4 Matrix norms, Conditioning, Vector Spaces, Linear Independence, Spanning sets and Basis, Null space and Range of a Matrix
Lectures - Week 4 Matrx norms, Condtonng, Vector Spaces, Lnear Independence, Spannng sets and Bass, Null space and Range of a Matrx Matrx Norms Now we turn to assocatng a number to each matrx. We could
More informationSTAT 309: MATHEMATICAL COMPUTATIONS I FALL 2018 LECTURE 16
STAT 39: MATHEMATICAL COMPUTATIONS I FALL 218 LECTURE 16 1 why teratve methods f we have a lnear system Ax = b where A s very, very large but s ether sparse or structured (eg, banded, Toepltz, banded plus
More informationInternet Appendix for "Nonbinding Voting for Shareholder Proposals"
Internet Appendx for "Nonbndng Votng for Shareholder Proposals" DORON LEVIT and NADYA MALENKO The Internet Appendx contans the omtted proofs from the man appendx (Secton I) and the analyss correspondng
More informationThe RepeatedPrisoners DilemmawithPrivate Monitoring: a N-player case
The RepeatedPrsoners DlemmawthPrvate Montorng: a N-player case Ichro Obara Department of Economcs Unversty of Pennsylvana obara@ssc.upenn.edu Frst verson: December, 1998 Ths verson: Aprl, 2000 Abstract
More informationExample: (13320, 22140) =? Solution #1: The divisors of are 1, 2, 3, 4, 5, 6, 9, 10, 12, 15, 18, 20, 27, 30, 36, 41,
The greatest common dvsor of two ntegers a and b (not both zero) s the largest nteger whch s a common factor of both a and b. We denote ths number by gcd(a, b), or smply (a, b) when there s no confuson
More informationn α j x j = 0 j=1 has a nontrivial solution. Here A is the n k matrix whose jth column is the vector for all t j=0
MODULE 2 Topcs: Lnear ndependence, bass and dmenson We have seen that f n a set of vectors one vector s a lnear combnaton of the remanng vectors n the set then the span of the set s unchanged f that vector
More informationLecture 7: Gluing prevarieties; products
Lecture 7: Glung prevaretes; products 1 The category of algebrac prevaretes Proposton 1. Let (f,ϕ) : (X,O X ) (Y,O Y ) be a morphsm of algebrac prevaretes. If U X and V Y are affne open subvaretes wth
More information