Measuring How Many Qualified Candidates Keyword Filters Wrongly Reject
Quantifying Algorithmic Friction in Automated Resume Screening Systems
Ibrahim Denis Fofanah · arXiv:2602.04087 (PDF) · submitted February 3, 2026
Summary by Andrew Johnson, OneTwo Resume editorial team · updated · We are not affiliated with the authors. arXiv papers are preprints and may not be peer-reviewed.
The follow-up to the Algorithmic Barrier framework, focused on measurement. Screening is modeled as a classification task, and "algorithmic friction" is defined precisely: excess false-negative rejections — qualified candidates wrongly filtered — caused by semantic misinterpretation.
In controlled simulations holding qualifications identical, deterministic keyword-based screening is compared against vector-space semantic matching. Keyword screening exhibits high friction; semantic matching substantially less. In plain terms: the older and cruder the matching technology, the more qualified people it throws away on wording alone.
What the paper reports
- Defines a measurable friction metric: excess false-negative rejection under identical qualification conditions.
- Keyword-based screening shows high algorithmic friction in controlled simulation.
- Vector-space semantic matching reduces (without eliminating) that friction.
What this means for your resume
Our editorial interpretation — the paper does not give job-seeker advice.
- You rarely know which technology an employer runs, and the punishing case is the crude one — so write for the keyword matcher: exact terms from the posting, spelled the way the posting spells them, in machine-readable text. If the employer runs something smarter, that costs you nothing.
Read it with these caveats
arXiv preprint (Feb 2026, v1); results are from controlled simulations, not audits of named commercial systems, and single-author work awaiting peer review.
Primary source: Quantifying Algorithmic Friction in Automated Resume Screening Systems — always read the paper before citing it. Spotted an error in our summary? Tell us and we'll fix it with a visible correction.
More in how ai screeners work — and fail
Forget Bias for a Moment — Do LLM Screeners Even Rank Correctly?
Bias audits ask if screeners are fair. This Princeton-line study asks the prior question: are their rankings valid at all?
General-Purpose LLMs vs a Purpose-Built Hiring Model, on 10,000 Real Pairs
OpenAI, Anthropic, Google, Meta, and DeepSeek models benchmarked on ~10,000 real candidate-job pairs — against a domain-specific model that beat them all.
A Name for the "200 Applications, Zero Replies" Problem
High vacancies AND prolonged unemployment, together — this framework blames deterministic screening rejecting qualified candidates via semantic misinterpretation.