{"id":1767,"date":"2023-07-07T23:03:02","date_gmt":"2023-07-07T14:03:02","guid":{"rendered":"https:\/\/saraheee.com\/?p=1767"},"modified":"2023-07-08T02:21:05","modified_gmt":"2023-07-07T17:21:05","slug":"game-theory-10-chap20-computing-pbe-in-signaling-games","status":"publish","type":"post","link":"https:\/\/saraheee.com\/ko\/2023\/07\/game-theory-10-chap20-computing-pbe-in-signaling-games\/","title":{"rendered":"Game Theory 10 \u2013 chap20. Computing PBE in signaling games"},"content":{"rendered":"<p>We study how to compute a PBE by going through the next classic example<\/p>\n\n\n\n<p><mark style=\"background-color:var(--global-color-10)\" class=\"has-inline-color\">EXAMPLE 5.7 (THE BEER-QUICHE GAME, CHO AND KREPS (1987))<\/mark>.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Player 1 can be weak (w) or strong (s), and let Pr(\\(t_1\\) = w) = 0.1. The strong type likes beer for breakfast, while the weak type likes quiche.<\/li>\n\n\n\n<li>Player 1 is ordering his breakfast, and player 2 is watching and contemplating whether to pick a fight with player 1. Player 2 wants to fight with the weak type but walk away from the strong type.<\/li>\n\n\n\n<li>Player 1 likes to avoid a fight: he gets a payoff of one from the preferred breakfast, and a payoff of two from avoiding the fight.<\/li>\n<\/ul>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-large is-resized\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-35-1024x531.png\" alt=\"\" class=\"wp-image-1773\" width=\"512\" height=\"266\" srcset=\"https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-35-1024x531.png 1024w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-35-300x156.png 300w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-35-768x398.png 768w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-35-1536x797.png 1536w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-35-2048x1062.png 2048w\" sizes=\"(max-width: 512px) 100vw, 512px\" \/><\/figure><\/div>\n\n\n<p>(1) Represent player 1&#8217;s <strong><mark style=\"background-color:rgba(0, 0, 0, 0)\" class=\"has-inline-color has-global-color-8-color\">strategy<\/mark><\/strong> \\(\\sigma_1\\) with two parameters:<\/p>\n\n\n\n<p>\\(\\sigma_1 = (\\sigma_1^w(B), \\sigma_1^s(B)) = (x, y)\\)<br>\\(\\sigma_1^w(B)\\): for wimpy type, \\(\\sigma_1^s(B)\\): for surly type<\/p>\n\n\n\n<p>(2) Compute player 2&#8217;s <strong><mark style=\"background-color:rgba(0, 0, 0, 0)\" class=\"has-inline-color has-global-color-8-color\">Bayesian beliefs<\/mark><\/strong> given \\(\\sigma_1\\):<\/p>\n\n\n\n<p>p \u2261 \\(\\mu_2^B(w) = \\frac{P_\\sigma(a)}{P_\\sigma(I_B)} = \\frac{0.1x}{0.1x+0.9y} = \\frac{x}{x+9y}\\)<\/p>\n\n\n\n<p>q \u2261 \\(\\mu_2^Q(w) = \\frac{0.1(1-x)}{0.1(1-x) + 0.9(1-y)} = \\frac{1-x}{10-x-9y}\\)<\/p>\n\n\n\n<p>(3) Find player 2&#8217;s <strong><mark style=\"background-color:rgba(0, 0, 0, 0)\" class=\"has-inline-color has-global-color-8-color\">best response<\/mark><\/strong> to \\(\\sigma_1\\) given beliefs:<\/p>\n\n\n\n<p>player 2&#8217;s BR @ \\(I_B = \\begin {cases}\u00a0F &amp; \\text{if } p \\geq \\frac{1}{2} \\\\\u00a0W &amp; \\text{if } p &lt; \\frac{1}{2} \\end {cases}\\)<\/p>\n\n\n\n<p>player 2&#8217;s BR @ \\(I_Q = \\begin {cases}\u00a0F &amp; \\text{if } q \\geq \\frac{1}{2} \\\\\u00a0W &amp; \\text{if } q &lt; \\frac{1}{2} \\end {cases}\\)<\/p>\n\n\n\n<p><strong>(4)<\/strong> Divide analysis into cases according to player 1&#8217;s choice of x and y:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>(x, y) = (1, 1), a <em>pooling strategy<\/em> in which both types send message B<\/li>\n<\/ul>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"alignleft size-large is-resized\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-37-1024x797.png\" alt=\"\" class=\"wp-image-1780\" width=\"512\" height=\"399\" srcset=\"https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-37-1024x797.png 1024w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-37-300x233.png 300w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-37-768x597.png 768w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-37.png 1054w\" sizes=\"(max-width: 512px) 100vw, 512px\" \/><\/figure><\/div>\n\n\n<p>p = 0.1 but q is unrestricted<br>If B was sent, W is BR given p = 0.1<\/p>\n\n\n\n<p>Suppose Q was sent and q &lt; 0.5 \u2192 W is BR<br>Not an eqbm \u2235 the w-type would then deviate to Q<\/p>\n\n\n\n<p>Suppose Q was sent but q \u2265 0.5 \u2192 F is BR<br>Both types wouldn&#8217;t deviate, and thus it is an eqbm<\/p>\n\n\n\n<p>There exists on perfect Bayesian eqbm in pooling strategies:<\/p>\n\n\n\n<p><strong>\u27e8(BB, WF), p = 0.1, q \u2265 0.5\u27e9<\/strong><\/p>\n\n\n\n<p><strong>(4-2)<\/strong> (x, y) = (1, 0), a separating strategy in which the weak type sends Q but the strong type sends B.<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"alignleft size-large is-resized\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-38-1024x778.png\" alt=\"\" class=\"wp-image-1781\" width=\"537\" height=\"408\" srcset=\"https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-38-1024x778.png 1024w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-38-300x228.png 300w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-38-768x584.png 768w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-38.png 1074w\" sizes=\"(max-width: 537px) 100vw, 537px\" \/><\/figure><\/div>\n\n\n<p>Bayesian beliefs under \\(\\sigma_1\\) = (1, 0) are<br>p = 1 and q = 0<\/p>\n\n\n\n<p>That is, player 2 exactly infers player 1&#8217;s type from the message sent<br>F is a BR to B and W is a BR to Q<\/p>\n\n\n\n<p>Not an eqbm b\/c the wimpy type would deviate to Q<\/p>\n\n\n\n<p>No PBE exists in this case<\/p>\n\n\n\n<p>There is <em>no<\/em> PBE in which player 1 uses the separating strategy BQ.<\/p>\n\n\n\n<p><strong>(4-3)<\/strong> (x, y) = (0, 1), another separating strategy<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"alignleft size-large is-resized\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-41-1024x815.png\" alt=\"\" class=\"wp-image-1784\" width=\"512\" height=\"408\" srcset=\"https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-41-1024x815.png 1024w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-41-300x239.png 300w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-41-768x611.png 768w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-41.png 1164w\" sizes=\"(max-width: 512px) 100vw, 512px\" \/><\/figure><\/div>\n\n\n<p>Bayesian beliefs are<br>p = 0 and q = 1<\/p>\n\n\n\n<p><mark style=\"background-color:rgba(0, 0, 0, 0)\" class=\"has-inline-color has-global-color-15-color\">That is, player 2 exactly infers player 1&#8217;s type from the message sent<br>F is a BR to Q and W is a BR to B<\/mark><\/p>\n\n\n\n<p>Not an eqbm b\/c the wimpy type would deviate to<mark style=\"background-color:rgba(0, 0, 0, 0)\" class=\"has-inline-color has-global-color-15-color\"> B<\/mark><\/p>\n\n\n\n<p>No PBE exists in this case<\/p>\n\n\n\n<p>Hence <em>no<\/em> PBE exists in which player employs a separating strategy <mark style=\"background-color:rgba(0, 0, 0, 0)\" class=\"has-inline-color has-global-color-15-color\">QB.<\/mark><\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong>(4-4)<\/strong> Lastly, we examine (x, y) = (0, 0) in which both types send Q<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"alignleft size-large is-resized\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-42-1024x828.png\" alt=\"\" class=\"wp-image-1785\" width=\"512\" height=\"414\" srcset=\"https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-42-1024x828.png 1024w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-42-300x243.png 300w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-42-768x621.png 768w, https:\/\/saraheee.com\/wp-content\/uploads\/2023\/07\/image-42.png 1182w\" sizes=\"(max-width: 512px) 100vw, 512px\" \/><\/figure><\/div>\n\n\n<ul class=\"wp-block-list\">\n<li>q = 0.1 but p is unrestricted<\/li>\n\n\n\n<li>If Q was sent <strong><mark style=\"background-color:rgba(0, 0, 0, 0)\" class=\"has-inline-color has-global-color-8-color\">W<\/mark><\/strong> is a BR<\/li>\n<\/ul>\n\n\n\n<p><mark style=\"background-color:rgba(0, 0, 0, 0)\" class=\"has-inline-color has-global-color-15-color\"><strong>\u27e8(QQ, FW), p \u2265 0.5, q = 0.1\u27e9<\/strong><\/mark><\/p>\n\n\n\n<p><mark style=\"background-color:rgba(0, 0, 0, 0)\" class=\"has-inline-color has-global-color-15-color\">If p &lt; 1\/2, player 2 will walk away whether player 1 orders a Beer or a Quiche<\/mark><\/p>\n\n\n\n<p><mark style=\"background-color:rgba(0, 0, 0, 0)\" class=\"has-inline-color has-global-color-15-color\">So player 1 has an incentive to debate.<\/mark><br><mark style=\"background-color:rgba(0, 0, 0, 0)\" class=\"has-inline-color has-global-color-15-color\">In the weak type, player 2 has no incentive to debate because player 1 gets 3 from Q,<\/mark><br><mark style=\"background-color:rgba(0, 0, 0, 0)\" class=\"has-inline-color has-global-color-15-color\">In the strong type, player 1 has an incentive to debate because he gets a payoff of 3 if he orders Beer.<\/mark><\/p>\n\n\n\n<p><mark style=\"background-color:rgba(0, 0, 0, 0)\" class=\"has-inline-color has-global-color-15-color\">So there is no PBE when p is less than 1\/2.<\/mark><\/p>\n\n\n\n<p><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Reference: Chang-Koo Chi, (39\/50) Game Theory and Applications 10 \u2013 Computing PBE in signaling games, Jul 14, 2020,&nbsp;<a href=\"https:\/\/youtu.be\/KrpsV-91YTI\" rel=\"noopener\">https:\/\/youtu.be\/KrpsV-91YTI<\/a><\/li>\n<\/ul>\n\n\n\n<p><\/p>","protected":false},"excerpt":{"rendered":"<p>Describe how to find a PBE in a signaling game.<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[5],"tags":[104,103,54,106,4,98,101,97,102,99,33],"class_list":["post-1767","post","type-post","status-publish","format-standard","hentry","category-game-theory-and-applications","tag-bayesian-beliefs","tag-beer-quiche-game","tag-best-response","tag-cho-kreps","tag-game-theory","tag-jul-7-2023","tag-pbe","tag-perfect-bayesian-equilibrium","tag-pooling-strategy","tag-signaling-games","tag-strategy"],"_links":{"self":[{"href":"https:\/\/saraheee.com\/ko\/wp-json\/wp\/v2\/posts\/1767"}],"collection":[{"href":"https:\/\/saraheee.com\/ko\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/saraheee.com\/ko\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/saraheee.com\/ko\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/saraheee.com\/ko\/wp-json\/wp\/v2\/comments?post=1767"}],"version-history":[{"count":10,"href":"https:\/\/saraheee.com\/ko\/wp-json\/wp\/v2\/posts\/1767\/revisions"}],"predecessor-version":[{"id":1815,"href":"https:\/\/saraheee.com\/ko\/wp-json\/wp\/v2\/posts\/1767\/revisions\/1815"}],"wp:attachment":[{"href":"https:\/\/saraheee.com\/ko\/wp-json\/wp\/v2\/media?parent=1767"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/saraheee.com\/ko\/wp-json\/wp\/v2\/categories?post=1767"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/saraheee.com\/ko\/wp-json\/wp\/v2\/tags?post=1767"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}