About academic Ledger

All you need is QaL

A calibrated, continuously updated estimate of a paper’s eventual quality, built on the open scholarly record.

▶  View the seminar talk

The big idea

Academic publishing bundles two jobs that do not belong together: distribution, getting the work out, and certification, deciding how good it is. Distribution is effectively solved; anyone can post a paper today. Certification is the hard part, and the journal system does it slowly, expensively, and noisily, compressing a rich question into a binary accept-or-reject verdict delivered years late.

academic Ledger separates the two. It leaves distribution to the existing hosts and rebuilds certification as a measurement. In place of a label, it reports QaL (said “qual”): an estimate of where a paper will eventually stand among its peers, expressed as a percentile with an honest interval rather than a yes or no.

The philosophy

• Quality is a percentile, not a label. A paper’s standing is relative to its field and its vintage, reported on a continuous scale.
• Decide late. We do not pretend to know a young paper’s fate. Early estimates carry wide intervals that narrow only as real evidence accumulates.
• Be honest about uncertainty. Every estimate travels with an interval and a reference class. No false precision.
• Transparency over secrecy. The method and inputs are open and auditable; resistance to gaming comes from construction, not from hidden formulas.
• No score for a person. Quality is reported per paper. We deliberately do not roll a scholar up to a single number.
• Built on the open record. The estimate rests on open, reproducible data, so anyone can check it and no one has to pay to see it.

The approach

The Ledger begins as a lens over the public record. Using the open scholarly graph, it computes QaL for any indexed paper with no participants required and no cold start. A paper’s reference class is the community it actually travels with in citations, which keeps the estimate robust to the coarse, single-field labels that mislabel interdisciplinary work. Human review and conferred tiers can be layered on later; the measurement comes first.

Estimates are calibrated against history: we learn, from communities observed to maturity, how an early signal maps to eventual standing, and apply that mapping to new work. Calibration starts with the fields we know best in operations, information systems, and decision sciences, and expands from there.