---
_id: '17061'
abstract:
- lang: eng
  text: Across many domains of interaction, both natural and artificial, individuals
    use past experience to shape future behaviors. The results of such learning processes
    depend on what individuals wish to maximize. A natural objective is one’s own
    success. However, when two such “selfish” learners interact with each other, the
    outcome can be detrimental to both, especially when there are conflicts of interest.
    Here, we explore how a learner can align incentives with a selfish opponent. Moreover,
    we consider the dynamics that arise when learning rules themselves are subject
    to evolutionary pressure. By combining extensive simulations and analytical techniques,
    we demonstrate that selfish learning is unstable in most classical two-player
    repeated games. If evolution operates on the level of long-run payoffs, selection
    instead favors learning rules that incorporate social (other-regarding) preferences.
    To further corroborate these results, we analyze data from a repeated prisoner’s
    dilemma experiment. We find that selfish learning is insufficient to explain human
    behavior when there is a trade-off between payoff maximization and fairness.
acknowledgement: The authors are grateful to Jörg Oechssler for many helpful comments.
  A.M. was supported by a Simons Postdoctoral Fellowship (Math+X) at the University
  of Pennsylvania; K.C. was supported by the European Research Council Consolidator
  Grant 863818 (ForM-SMArt); and C.H. was supported by the European Research Council
  Starting Grant 850529 (E-DIRECT).
article_number: pgac141
article_processing_charge: Yes
article_type: original
arxiv: 1
author:
- first_name: Alex
  full_name: McAvoy, Alex
  last_name: McAvoy
- first_name: Julian
  full_name: Kates-Harbeck, Julian
  last_name: Kates-Harbeck
- first_name: Krishnendu
  full_name: Chatterjee, Krishnendu
  id: 2E5DCA20-F248-11E8-B48F-1D18A9856A87
  last_name: Chatterjee
  orcid: 0000-0002-4561-241X
- first_name: Christian
  full_name: Hilbe, Christian
  id: 2FDF8F3C-F248-11E8-B48F-1D18A9856A87
  last_name: Hilbe
  orcid: 0000-0001-5116-955X
citation:
  ama: McAvoy A, Kates-Harbeck J, Chatterjee K, Hilbe C. Evolutionary instability
    of selfish learning in repeated games. <i>PNAS Nexus</i>. 2022;1(4). doi:<a href="https://doi.org/10.1093/pnasnexus/pgac141">10.1093/pnasnexus/pgac141</a>
  apa: McAvoy, A., Kates-Harbeck, J., Chatterjee, K., &#38; Hilbe, C. (2022). Evolutionary
    instability of selfish learning in repeated games. <i>PNAS Nexus</i>. Oxford University
    Press. <a href="https://doi.org/10.1093/pnasnexus/pgac141">https://doi.org/10.1093/pnasnexus/pgac141</a>
  chicago: McAvoy, Alex, Julian Kates-Harbeck, Krishnendu Chatterjee, and Christian
    Hilbe. “Evolutionary Instability of Selfish Learning in Repeated Games.” <i>PNAS
    Nexus</i>. Oxford University Press, 2022. <a href="https://doi.org/10.1093/pnasnexus/pgac141">https://doi.org/10.1093/pnasnexus/pgac141</a>.
  ieee: A. McAvoy, J. Kates-Harbeck, K. Chatterjee, and C. Hilbe, “Evolutionary instability
    of selfish learning in repeated games,” <i>PNAS Nexus</i>, vol. 1, no. 4. Oxford
    University Press, 2022.
  ista: McAvoy A, Kates-Harbeck J, Chatterjee K, Hilbe C. 2022. Evolutionary instability
    of selfish learning in repeated games. PNAS Nexus. 1(4), pgac141.
  mla: McAvoy, Alex, et al. “Evolutionary Instability of Selfish Learning in Repeated
    Games.” <i>PNAS Nexus</i>, vol. 1, no. 4, pgac141, Oxford University Press, 2022,
    doi:<a href="https://doi.org/10.1093/pnasnexus/pgac141">10.1093/pnasnexus/pgac141</a>.
  short: A. McAvoy, J. Kates-Harbeck, K. Chatterjee, C. Hilbe, PNAS Nexus 1 (2022).
date_created: 2024-05-28T14:23:12Z
date_published: 2022-09-01T00:00:00Z
date_updated: 2025-06-11T13:54:20Z
day: '01'
ddc:
- '000'
department:
- _id: KrCh
doi: 10.1093/pnasnexus/pgac141
ec_funded: 1
external_id:
  arxiv:
  - '2105.06199'
  pmid:
  - '36714856'
file:
- access_level: open_access
  checksum: 79a8e3e4be7e8a2b407b4efddd65f3f3
  content_type: application/pdf
  creator: dernst
  date_created: 2024-08-06T07:33:30Z
  date_updated: 2024-08-06T07:33:30Z
  file_id: '17400'
  file_name: 2022_PNASNexus_McAvoy.pdf
  file_size: 2410962
  relation: main_file
  success: 1
file_date_updated: 2024-08-06T07:33:30Z
has_accepted_license: '1'
intvolume: '         1'
issue: '4'
language:
- iso: eng
month: '09'
oa: 1
oa_version: Published Version
pmid: 1
project:
- _id: 0599E47C-7A3F-11EA-A408-12923DDC885E
  call_identifier: H2020
  grant_number: '863818'
  name: 'Formal Methods for Stochastic Models: Algorithms and Applications'
publication: PNAS Nexus
publication_identifier:
  issn:
  - 2752-6542
publication_status: published
publisher: Oxford University Press
quality_controlled: '1'
related_material:
  link:
  - relation: software
    url: https://github.com/alexmcavoy/fmtl/
scopus_import: '1'
status: public
title: Evolutionary instability of selfish learning in repeated games
tmp:
  image: /images/cc_by.png
  legal_code_url: https://creativecommons.org/licenses/by/4.0/legalcode
  name: Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)
  short: CC BY (4.0)
type: journal_article
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 1
year: '2022'
...