{"paper_id":"2504.16093","title":"Efficient Portfolio Selection through Preference Aggregation with Quicksort and the Bradley--Terry Model","authors":["Yurun Ge","Lucas Böttcher","Tom Chou","Maria R. D'Orsogna"],"paper_date":"2025-04-06","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2504.16093v1","page_url":"https://thequant.space/flowcharts/efficient-portfolio-selection-through-preference-aggregation/","scores":{"empirical_rigor":3,"math_complexity":7.5,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes pairwise‑comparison aggregation rules based on Quicksort and the Bradley–Terry model to rank projects using agents’ win probabilities.","quote":"We propose comparison rules based on Quicksort and the Bradley--Terry model, which connects rankings to pairwise \"win\" probabilities. In our model, each agent determines win probabilities of a pair of projects based on his or her specific evaluation of the projects' long-term benefit.","summary":"How to allocate limited resources to projects that will yield the greatest long-term benefits is a problem that often arises in decision-making under uncertainty. For example, organizations may need to evaluate and select innovation projects with risky returns. Similarly, when allocating resources…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The noise in perceived project values is independently and normally distributed.","type":"statistical","stated":"explicit","quote":"(7) Under the assumption that the noise in the perceived value is independently and normally distributed, the quantity ηiℓ−η jℓfollows a normal distribution with a mean of 0 and a standard deviation of q σ2 iℓ+ σ2 jℓ.","support":"full"},{"assumption":"Agents evaluate win probabilities wℓij as the probability that their perceived value for project i exceeds that for project j.","type":"model","stated":"explicit","quote":"Suppose agent ℓis evaluating projects i and j, with perceived values viℓand v jℓ, respectively, and corresponding noise terms ηiℓand η jℓ. As a starting point, we define wℓ ij B Pr \u0010 vi > v j \u0011 = Pr \u0010 (viℓ−ηiℓ) > (vjℓ−ηjℓ) \u0011 = Pr \u0010 (ηiℓ−ηjℓ) < (viℓ−vjℓ) \u0011 .","support":"full"},{"assumption":"Project intrinsic values are fixed ground‑truth numbers and agents have no direct knowledge of them.","type":"model","stated":"implied","quote":"To model organizational decision-making, agents can be considered as having specific expertise, while projects are characterized by an intrinsic value (representing their long-term benefit and assumed to be ground truth) and a defining type [1].","support":"full"},{"assumption":"Monte‑Carlo simulations with additive Gaussian noise are used to generate synthetic project evaluations.","type":"other","stated":"explicit","quote":"The uncer-tainty in agent ℓ’s project evaluations is quantified by additive Gaussian noise with zero mean and standard deviation σiℓ= |ti −eℓ|, where the expertise level eℓof agent ℓis given by Eq. (6). All our results are based on Monte Carlo simulations.","support":"full"}]},"datasets":{"status":"available","items":[{"name":"Monte Carlo simulation of 30 projects with Gaussian noise","provider":null,"kind":"synthetic","access":"unknown","period":null,"universe":"30 projects (i = 1,…,30)","quote":"We define the value of project i as vi = i (i ∈{1, . . . , 30}). The uncer-tainty in agent ℓ’s project evaluations is quantified by additive Gaussian noise with zero mean and standard deviation σiℓ= |ti −eℓ|, where the expertise level eℓof agent ℓis given by Eq. (6).","support":"full"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"synthetic"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"Several of the proposed aggregation methods perform better than the two most effective existing aggregation methods.","kind":"performance","quote":"Several of the methods we propose perform better than the two most effective aggregation methods currently available. In Section 6, we examine the performance of these aggregation methods to show that several of the methods we propose perform better than existing aggregation methods.","support":"full"},{"finding":"When win probabilities are restricted to a finite set, the relative performance ranking of the methods remains unchanged.","kind":"performance","quote":"The relative performance ranking of the methods remains unchanged.","support":"full"},{"finding":"The Two‑Phase Bradley–Terry method achieves favorable performance with a relatively small number of pairwise comparisons, making it the most practical approach for real‑world use.","kind":"performance","quote":"In contrast, the Two-Phase Bradley–Terry method achieves favorable performance with a relatively small number of pairwise comparisons, making it the most practical approach for real-world use. 7.","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":18,"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":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"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":"2026-11-14T19:50:56Z"},"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/"}