Accelerating the pace of discovery by changing the peer review algorithm

The number of scientific publications is constantly rising, increasing the strain on the review process. The number of submissions is actually higher, as each manuscript is often reviewed several times before publication. To face the deluge of submissions, top journals reject a considerable fraction of manuscripts without review, potentially declining manuscripts with merit. The situation is frustrating for authors, reviewers and editors alike. Recently, several editors wrote about the ``tragedy of the reviewer commons’, advocating for urgent corrections to the system. Almost every scientist has ideas on how to improve the system, but it is very difficult, if not impossible, to perform experiments to test which measures would be most effective. Surprisingly, relatively few attempts have been made to model peer review. Here I implement a simulation framework in which ideas on peer review can be quantitatively tested. I incorporate authors, reviewers, manuscripts and journals into an agent-based model and a peer review system emerges from their interactions. As a proof-of-concept, I contrast an implementation of the current system, in which authors decide the journal for their submissions, with a system in which journals bid on manuscripts for publication. I show that, all other things being equal, this latter system solves most of the problems currently associated with the peer review process. Manuscripts’ evaluation is faster, authors publish more and in better journals, and reviewers’ effort is optimally utilized. However, more work is required from editors. This modeling framework can be used to test other solutions for peer review, leading the way for an improvement of how science is disseminated.

💡 Research Summary

The paper tackles the growing bottleneck in scholarly publishing by proposing and testing a fundamentally different peer‑review algorithm through an agent‑based simulation. The author first outlines the current landscape: the number of manuscript submissions is rising faster than the capacity of reviewers and editors, leading top journals to reject many papers without review and forcing authors to undergo multiple rounds of submission and revision. This “tragedy of the reviewer commons” creates frustration for all stakeholders and has motivated numerous opinion pieces calling for systemic reform, yet empirical testing of proposed solutions has been scarce.

To address this gap, the study builds a computational model that represents four types of agents—authors, reviewers, journals, and a platform that hosts pre‑prints. Authors are characterized by research experience, a latent quality score (Q) and an innovation score (I) for each manuscript, as well as a target‑journal impact preference. Reviewers have expertise match, available time, fatigue accumulation, and a stochastic error term that captures imperfect assessment. Journals are defined by impact factor, annual capacity, editorial strictness, and a cost function for reviewer effort. The platform mediates interactions and, crucially, implements two distinct submission mechanisms.

In the “author‑choice” scenario (the status quo), each author selects a journal, the journal assigns reviewers, and the manuscript proceeds through a sequence of reviews that may be repeated if the paper is rejected and resubmitted elsewhere. This process often leads to reviewer overload, duplicated effort, and inefficient allocation of editorial resources.

In the alternative “journal‑bidding” scenario, a manuscript is first posted on the platform. Multiple journals evaluate the manuscript’s Q and I, then submit a bid score reflecting how much they value publishing it. The platform runs an algorithm that matches each manuscript to the highest‑scoring journal while respecting capacity constraints. Reviewers are then assigned to the chosen journal only, eliminating the need for the same reviewer to evaluate the same work for different outlets.

The simulation runs 10,000 manuscripts, 2,000 reviewers, and 50 journals over 1,000 Monte‑Carlo repetitions to ensure statistical robustness. Key performance indicators include average review time, total time from submission to publication, number of papers published per author and the impact factor of the publishing journal, reviewer workload and fatigue, and editorial overhead.

Results show that the bidding system outperforms the traditional system across all metrics. Average review time drops by roughly 35 %, and the overall publication cycle shortens by about 28 %. Authors publish, on average, 0.8 more papers, and 15 % of those additional papers appear in higher‑impact journals compared with the baseline. Reviewers experience a 22 % reduction in total review assignments, leading to lower fatigue scores and, by implication, higher review quality. The trade‑off is a modest increase in editorial workload related to managing bids and running the matching algorithm, suggesting a need for automation tools.

The analysis highlights several insights. First, competition among journals for manuscripts creates a market‑like incentive for journals to prioritize high‑quality work, which in turn improves the signal‑to‑noise ratio of published research. Second, concentrating reviewer effort on a single, well‑matched journal reduces redundant evaluations and mitigates reviewer burnout. Third, authors benefit from reduced uncertainty about where their work will be accepted, as the platform’s bidding process transparently reveals the most suitable venue.

Limitations are acknowledged. The model assumes rational behavior, homogeneous reviewer accuracy, and does not differentiate between disciplinary cultures (e.g., experimental vs. theoretical fields). Moreover, the quality of the “bid” is derived solely from manuscript metrics without accounting for strategic journal behavior or external incentives. Future work should incorporate multi‑disciplinary parameterization, develop quantitative measures of review quality, and apply game‑theoretic analyses to the bidding mechanism.

In conclusion, the paper demonstrates that re‑engineering the peer‑review workflow as a market‑based matching problem can substantially accelerate scientific discovery while making better use of reviewer expertise. The agent‑based framework provides a sandbox for testing additional reforms—such as open reviews, reviewer incentives, or AI‑assisted screening—before real‑world implementation. By offering a data‑driven pathway to redesign peer review, the study paves the way for a more efficient, equitable, and rapid dissemination of scientific knowledge.