{"paper_id":"2307.11846","title":"Social and individual learning in the Minority Game","authors":["Bryce Morsky","Fuwei Zhuang","Zuojun Zhou"],"paper_date":"2023-07-21","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2307.11846v2","page_url":"https://thequant.space/flowcharts/social-and-individual-learning-in-the-minority-game/","scores":{"empirical_rigor":2,"math_complexity":7.5,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes to analyze how social and individual learning affect outcomes in the Minority Game model of a financial market, showing that social learning can reduce market efficiency while individual learning can mitigate these inefficiencies.","quote":"We study the roles of social and individual learning on outcomes of the Minority Game model of a financial market. We show how individual learning can rescue a population engaged in social learning from such inefficiencies.","summary":"We study the roles of social and individual learning on outcomes of the Minority Game model of a financial market. Social learning occurs via agents adopting the strategies of their neighbours within a social network, while individual learning results in agents changing their strategies without…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"After each turn agents learn socially with probability ℓs or individually with probability ℓi.","type":"model","stated":"explicit","quote":"After every turn, agents will learn socially or individually with probabilities ℓs and ℓi, respectively.","support":"full"},{"assumption":"Social learning occurs via pairwise imitation across a fixed social network.","type":"model","stated":"explicit","quote":"We consider the case where this imitation occurs across a social network. Once the graph is constructed, we assume that it remains fixed throughout the game.","support":"full"},{"assumption":"The social network is represented by either a complete graph or an Erdős‑Rényi random graph constructed with edge probability p = d/N.","type":"model","stated":"explicit","quote":"Specifically, we construct complete graphs and Erd˝os-R´enyi random graphs where agents can only socially learn from agents with which they share an edge. To construct the Erd˝os-R´enyi random graphs, a pair of agents is connected with probability p = d/N, where d is the mean node degree.","support":"full"},{"assumption":"Imitation probability follows a logistic (Fermi) selection function based on payoff differences.","type":"model","stated":"explicit","quote":"Once a focal agent has been paired with one of its neighbours on the social network to learn from, we simulate social learning using a logistic function (as used in previous research on social learning and evolutionary game theory literature where it is sometimes called Fermi selection [48, 50, 37, 38]): Pi→j(t) = 1 1 + exp(κ(min(πik(t)) −max(πjk(t)))),","support":"full"},{"assumption":"Agents know the true historic success of others’ strategy tables.","type":"model","stated":"explicit","quote":"Additionally, we have assumed that agents know the true historic success of others’ strategy tables.","support":"full"},{"assumption":"No costs are associated with either social or individual learning.","type":"other","stated":"explicit","quote":"For one, we have not assumed any costs associated with either social or individual learning.","support":"full"}]},"datasets":{"status":"available","items":[{"name":"Synthetic Minority Game simulations","provider":null,"kind":"synthetic","access":"unknown","period":null,"universe":null,"quote":"After every turn, agents will learn socially or individually with probabilities ℓs and ℓi, respectively. 3 Results We begin our results by depicting several informative time series of the attendance A(t) in Figure 1.","support":"full"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"synthetic"},"transaction_costs":{"status":"available","modeled":"unclear","headline_net_of_costs":null,"detail":null,"quote":null,"source":"conflict","rule":"R4b","candidates":{"front_matter":"n/a","evidence_card":"no","evidence_card_quoted":true},"conflict_kind":"applicability","card_detail":null,"card_quote":"For one, we have not assumed any costs associated with either social or individual learning.","meaning":"the platform classifier and the paper-text extraction disagree, so no value is asserted"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"Social learning alone leads to large swings in attendance and high volatility, reducing market efficiency.","kind":"theoretical","quote":"Social learning, however, on its own results in large swings in the population from buying to selling and thus a high volatility relative to the other cases (Figure 1c).","support":"full"},{"finding":"Social learning causes a crash in strategy‑table diversity, potentially locking the population into a maximally inefficient state.","kind":"theoretical","quote":"This phe-nomenon occurs because there is a crash in strategy table diversity as agents lose strategy tables 5 (a) No social or individual learning, N = 1001 (b) Only individual learning, N = 1001 (c) Only social learning, N = 1001 (d) Individual and social learning, N = 1001 (e) Only social learning, N = 101 Figure 1: Time series of the attendance A for no individual or social learning (a), only individual learn-ing","support":"full"},{"finding":"Individual learning alone yields marginally lower attendance variability compared to no learning.","kind":"theoretical","quote":"We find that when there is only individual learning, the variability in attendance is marginally lower than in the case with no social or individual learning (Figure 1b).","support":"full"},{"finding":"Adding individual learning to social learning reduces volatility to levels comparable to the no‑learning case.","kind":"theoretical","quote":"Figure 1d depicts a representative example showing that the addition of indi-vidual learning reduces the volatility to levels comparable to the case where there is no learning (Figure 1a).","support":"full"},{"finding":"Individual learning can rescue a population from the inefficiencies caused by social learning.","kind":"theoretical","quote":"We show how individual learning can rescue a population engaged in social learning from such inefficiencies.","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"available","supports":[],"contradicts":[{"paper":{"paper_id":"2511.13614","title":"Market-Dependent Communication in Multi-Agent Alpha Generation","paper_date":"2025-11-17","authors":["Jerick Shi","Burton Hollifield"],"empirical_rigor":6,"page_url":"https://thequant.space/flowcharts/market-dependent-communication-in-multi-agent-alpha-generati/"},"question":"Does information sharing affect strategy diversity in multi‑agent market games?","this_paper_finding":"Social learning causes a crash in strategy‑table diversity, potentially locking the population into a maximally inefficient state.","other_paper_finding":"All organizational structures converge to similar strategy correlations regardless of information sharing, eliminating diversity loss.","other_paper_quote":"All organizational structures converge to similar strategy correlations regardless of information sharing, eliminating \"diversity loss\" as communication’s primary failure mode.","note":"A reports a crash in diversity due to social learning, while B reports no diversity loss as all structures converge, contradicting A.","confidence":0.499}],"qualifies":[],"compared_with":14,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":true,"code_url":"https://github.com/bmorsky/minority-games","note":null},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":2,"extracted_at":"2026-11-14T20:02:01Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}