Negatively correlated bandits
We analyse a two-player game of strategic experimentation with two-armed bandits. Either player has to decide in continuous time whether to use a safe arm with a known pay-off or a risky arm whose expected pay-off per unit of time is initially unknown. This pay-off can be high or low and is negative...
Gespeichert in:
| Veröffentlicht in: | The Review of economic studies 2011-04, Vol.78 (2), p.693-732 |
|---|---|
| Hauptverfasser: | , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 732 |
|---|---|
| container_issue | 2 |
| container_start_page | 693 |
| container_title | The Review of economic studies |
| container_volume | 78 |
| creator | Klein, Nicolas Rady, Sven |
| description | We analyse a two-player game of strategic experimentation with two-armed bandits. Either player has to decide in continuous time whether to use a safe arm with a known pay-off or a risky arm whose expected pay-off per unit of time is initially unknown. This pay-off can be high or low and is negatively correlated across players. We characterize the set of all Markov perfect equilibria in the benchmark case where the risky arms are known to be of opposite type and construct equilibria in cut-offf strategies for arbitrary negative correlation. All strategies and pay-offs are in closed form. In marked contrast to the case where both risky arms are of the same type, there always exists an equilibrium in cut-off strategies, and there always exists an equilibrium exhibiting efficient long-run patterns of learning. These results extend to a three-player game with common knowledge that exactly one risky arm is of the high pay-off type. |
| doi_str_mv | 10.1093/restud/rdq025 |
| format | Article |
| fullrecord | <record><control><sourceid>jstor_proqu</sourceid><recordid>TN_cdi_proquest_miscellaneous_865526074</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><jstor_id>23015871</jstor_id><oup_id>10.1093/restud/rdq025</oup_id><sourcerecordid>23015871</sourcerecordid><originalsourceid>FETCH-LOGICAL-c473t-20f852f2eb6e2e0e4cd670a57be1b7e9dc4c181d02735aedaaea588ef021d3353</originalsourceid><addsrcrecordid>eNqF0EtLxDAQB_AgCq6PoxdhYfGil7qTpGnSoyy-YNGLgreQJlPp0t12k1bYb28kouDF0xzmxzz-hJxRuKZQ8rnHMIxu7t0WmNgjE5oXMiu5fNsnEwCeZ4VgxSE5CmEFAFQpOSHTJ3w3Q_OB7W5mO--xNQO6WWU2rhnCCTmoTRvw9Lsek9e725fFQ7Z8vn9c3Cwzm0s-ZAxqJVjNsCqQIWBuXSHBCFkhrSSWzuaWKuqASS4MOmPQCKWwBkYd54Ifk8s0t_fddoxv6HUTLLat2WA3Bq0KEU8HmUd58UeuutFv4nG6BM4og1JGlCVkfReCx1r3vlkbv9MU9FdUOkWlU1TRXyXfjf2_dJroKgyd_8GMAxVK0tg_T33X9L9ry1wJzvkngDt9pQ</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>903212097</pqid></control><display><type>article</type><title>Negatively correlated bandits</title><source>Jstor Complete Legacy</source><source>Oxford University Press Journals All Titles (1996-Current)</source><source>Business Source Complete</source><creator>Klein, Nicolas ; Rady, Sven</creator><creatorcontrib>Klein, Nicolas ; Rady, Sven</creatorcontrib><description>We analyse a two-player game of strategic experimentation with two-armed bandits. Either player has to decide in continuous time whether to use a safe arm with a known pay-off or a risky arm whose expected pay-off per unit of time is initially unknown. This pay-off can be high or low and is negatively correlated across players. We characterize the set of all Markov perfect equilibria in the benchmark case where the risky arms are known to be of opposite type and construct equilibria in cut-offf strategies for arbitrary negative correlation. All strategies and pay-offs are in closed form. In marked contrast to the case where both risky arms are of the same type, there always exists an equilibrium in cut-off strategies, and there always exists an equilibrium exhibiting efficient long-run patterns of learning. These results extend to a three-player game with common knowledge that exactly one risky arm is of the high pay-off type.</description><identifier>ISSN: 0034-6526</identifier><identifier>ISSN: 1467-937X</identifier><identifier>ISSN: 0034-6527</identifier><identifier>EISSN: 1467-937X</identifier><identifier>DOI: 10.1093/restud/rdq025</identifier><language>eng</language><publisher>Oxford: Review of Economic Studies Ltd., Blackwell Publishing</publisher><subject>Approximation ; Correlation ; Correlation analysis ; Correlations ; Economic transitions ; Efficient strategies ; Entscheidung ; Equilibrium ; Erwartungshaltung ; Experimentation ; Game theory ; Gewinn ; Laws of Motion ; Learning ; Lernen ; Markov analysis ; Markovian processes ; Markowscher Prozess ; Odes ; Opportunity costs ; Pay-off ; Risiko ; Risikomanagement ; Risk ; Spieltheorie ; Strategic behaviour ; Studies ; Time ; Uniqueness</subject><ispartof>The Review of economic studies, 2011-04, Vol.78 (2), p.693-732</ispartof><rights>The Review of Economic Studies Ltd 2011</rights><rights>The Author 2011. Published by Oxford University Press on behalf of The Review of Economic Studies Limited. 2011</rights><rights>Copyright Blackwell Publishing Ltd. Apr 2011</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c473t-20f852f2eb6e2e0e4cd670a57be1b7e9dc4c181d02735aedaaea588ef021d3353</citedby><cites>FETCH-LOGICAL-c473t-20f852f2eb6e2e0e4cd670a57be1b7e9dc4c181d02735aedaaea588ef021d3353</cites></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktopdf>$$Uhttps://www.jstor.org/stable/pdf/23015871$$EPDF$$P50$$Gjstor$$H</linktopdf><linktohtml>$$Uhttps://www.jstor.org/stable/23015871$$EHTML$$P50$$Gjstor$$H</linktohtml><link.rule.ids>314,776,780,799,1578,27901,27902,57992,58225</link.rule.ids><backlink>$$Uhttp://www.fachportal-paedagogik.de/fis_bildung/suche/fis_set.html?FId=948533$$DAccess content in the German Education Portal$$Hfree_for_read</backlink></links><search><creatorcontrib>Klein, Nicolas</creatorcontrib><creatorcontrib>Rady, Sven</creatorcontrib><title>Negatively correlated bandits</title><title>The Review of economic studies</title><description>We analyse a two-player game of strategic experimentation with two-armed bandits. Either player has to decide in continuous time whether to use a safe arm with a known pay-off or a risky arm whose expected pay-off per unit of time is initially unknown. This pay-off can be high or low and is negatively correlated across players. We characterize the set of all Markov perfect equilibria in the benchmark case where the risky arms are known to be of opposite type and construct equilibria in cut-offf strategies for arbitrary negative correlation. All strategies and pay-offs are in closed form. In marked contrast to the case where both risky arms are of the same type, there always exists an equilibrium in cut-off strategies, and there always exists an equilibrium exhibiting efficient long-run patterns of learning. These results extend to a three-player game with common knowledge that exactly one risky arm is of the high pay-off type.</description><subject>Approximation</subject><subject>Correlation</subject><subject>Correlation analysis</subject><subject>Correlations</subject><subject>Economic transitions</subject><subject>Efficient strategies</subject><subject>Entscheidung</subject><subject>Equilibrium</subject><subject>Erwartungshaltung</subject><subject>Experimentation</subject><subject>Game theory</subject><subject>Gewinn</subject><subject>Laws of Motion</subject><subject>Learning</subject><subject>Lernen</subject><subject>Markov analysis</subject><subject>Markovian processes</subject><subject>Markowscher Prozess</subject><subject>Odes</subject><subject>Opportunity costs</subject><subject>Pay-off</subject><subject>Risiko</subject><subject>Risikomanagement</subject><subject>Risk</subject><subject>Spieltheorie</subject><subject>Strategic behaviour</subject><subject>Studies</subject><subject>Time</subject><subject>Uniqueness</subject><issn>0034-6526</issn><issn>1467-937X</issn><issn>0034-6527</issn><issn>1467-937X</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2011</creationdate><recordtype>article</recordtype><recordid>eNqF0EtLxDAQB_AgCq6PoxdhYfGil7qTpGnSoyy-YNGLgreQJlPp0t12k1bYb28kouDF0xzmxzz-hJxRuKZQ8rnHMIxu7t0WmNgjE5oXMiu5fNsnEwCeZ4VgxSE5CmEFAFQpOSHTJ3w3Q_OB7W5mO--xNQO6WWU2rhnCCTmoTRvw9Lsek9e725fFQ7Z8vn9c3Cwzm0s-ZAxqJVjNsCqQIWBuXSHBCFkhrSSWzuaWKuqASS4MOmPQCKWwBkYd54Ifk8s0t_fddoxv6HUTLLat2WA3Bq0KEU8HmUd58UeuutFv4nG6BM4og1JGlCVkfReCx1r3vlkbv9MU9FdUOkWlU1TRXyXfjf2_dJroKgyd_8GMAxVK0tg_T33X9L9ry1wJzvkngDt9pQ</recordid><startdate>20110401</startdate><enddate>20110401</enddate><creator>Klein, Nicolas</creator><creator>Rady, Sven</creator><general>Review of Economic Studies Ltd., Blackwell Publishing</general><general>Oxford University Press</general><scope>9S6</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>8BJ</scope><scope>FQK</scope><scope>JBE</scope></search><sort><creationdate>20110401</creationdate><title>Negatively correlated bandits</title><author>Klein, Nicolas ; Rady, Sven</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c473t-20f852f2eb6e2e0e4cd670a57be1b7e9dc4c181d02735aedaaea588ef021d3353</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2011</creationdate><topic>Approximation</topic><topic>Correlation</topic><topic>Correlation analysis</topic><topic>Correlations</topic><topic>Economic transitions</topic><topic>Efficient strategies</topic><topic>Entscheidung</topic><topic>Equilibrium</topic><topic>Erwartungshaltung</topic><topic>Experimentation</topic><topic>Game theory</topic><topic>Gewinn</topic><topic>Laws of Motion</topic><topic>Learning</topic><topic>Lernen</topic><topic>Markov analysis</topic><topic>Markovian processes</topic><topic>Markowscher Prozess</topic><topic>Odes</topic><topic>Opportunity costs</topic><topic>Pay-off</topic><topic>Risiko</topic><topic>Risikomanagement</topic><topic>Risk</topic><topic>Spieltheorie</topic><topic>Strategic behaviour</topic><topic>Studies</topic><topic>Time</topic><topic>Uniqueness</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Klein, Nicolas</creatorcontrib><creatorcontrib>Rady, Sven</creatorcontrib><collection>FIS Bildung Literaturdatenbank</collection><collection>CrossRef</collection><collection>International Bibliography of the Social Sciences (IBSS)</collection><collection>International Bibliography of the Social Sciences</collection><collection>International Bibliography of the Social Sciences</collection><jtitle>The Review of economic studies</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Klein, Nicolas</au><au>Rady, Sven</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Negatively correlated bandits</atitle><jtitle>The Review of economic studies</jtitle><date>2011-04-01</date><risdate>2011</risdate><volume>78</volume><issue>2</issue><spage>693</spage><epage>732</epage><pages>693-732</pages><issn>0034-6526</issn><issn>1467-937X</issn><issn>0034-6527</issn><eissn>1467-937X</eissn><abstract>We analyse a two-player game of strategic experimentation with two-armed bandits. Either player has to decide in continuous time whether to use a safe arm with a known pay-off or a risky arm whose expected pay-off per unit of time is initially unknown. This pay-off can be high or low and is negatively correlated across players. We characterize the set of all Markov perfect equilibria in the benchmark case where the risky arms are known to be of opposite type and construct equilibria in cut-offf strategies for arbitrary negative correlation. All strategies and pay-offs are in closed form. In marked contrast to the case where both risky arms are of the same type, there always exists an equilibrium in cut-off strategies, and there always exists an equilibrium exhibiting efficient long-run patterns of learning. These results extend to a three-player game with common knowledge that exactly one risky arm is of the high pay-off type.</abstract><cop>Oxford</cop><pub>Review of Economic Studies Ltd., Blackwell Publishing</pub><doi>10.1093/restud/rdq025</doi><tpages>40</tpages><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext |
| identifier | ISSN: 0034-6526 |
| ispartof | The Review of economic studies, 2011-04, Vol.78 (2), p.693-732 |
| issn | 0034-6526 1467-937X 0034-6527 1467-937X |
| language | eng |
| recordid | cdi_proquest_miscellaneous_865526074 |
| source | Jstor Complete Legacy; Oxford University Press Journals All Titles (1996-Current); Business Source Complete |
| subjects | Approximation Correlation Correlation analysis Correlations Economic transitions Efficient strategies Entscheidung Equilibrium Erwartungshaltung Experimentation Game theory Gewinn Laws of Motion Learning Lernen Markov analysis Markovian processes Markowscher Prozess Odes Opportunity costs Pay-off Risiko Risikomanagement Risk Spieltheorie Strategic behaviour Studies Time Uniqueness |
| title | Negatively correlated bandits |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-01T19%3A06%3A53IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-jstor_proqu&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Negatively%20correlated%20bandits&rft.jtitle=The%20Review%20of%20economic%20studies&rft.au=Klein,%20Nicolas&rft.date=2011-04-01&rft.volume=78&rft.issue=2&rft.spage=693&rft.epage=732&rft.pages=693-732&rft.issn=0034-6526&rft.eissn=1467-937X&rft_id=info:doi/10.1093/restud/rdq025&rft_dat=%3Cjstor_proqu%3E23015871%3C/jstor_proqu%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=903212097&rft_id=info:pmid/&rft_jstor_id=23015871&rft_oup_id=10.1093/restud/rdq025&rfr_iscdi=true |