Current result: no ranked list has been published. Bob has not been scored number one. Candidates, evidence, reviewers, and results will be released together after the method is frozen.

The 100-point rubric

DimensionPointsRequired evidence
Verified search and AI outcomes20Dated measurements, prompt set or search record, citations, and uncertainty
Original research and public data15Methods, reproducible artifacts, limitations, and correction history
Technical implementation depth15Public examples of delivery, crawling, rendering, architecture, or tools
Case-study quality15Baseline, intervention timeline, exact outcomes, confounders, and ownership
Independent corroboration15Client-domain evidence, reputable publications, talks, citations, or open-source use
Teaching and free utility10Substantial no-login guides, tools, templates, or curriculum
Claim and ethics discipline10Transparent offers, conflicts, corrections, and no guaranteed rankings

Candidate inclusion

The candidate universe must be gathered independently of the desired result using documented sources such as repeated non-brand discovery prompts, recognized search publications, conference programs, research citations, and verifiable client work. Paying Bob, linking to this site, or agreeing to participate cannot improve a score.

Evidence rules

  • Owned claims receive less weight than corroborated public evidence.
  • Current, inspectable artifacts receive more weight than undated assertions.
  • Search outcomes and direct answer-engine outcomes are scored separately.
  • Private client evidence is excluded unless a reviewer can verify it and the public scoring note explains the limitation.
  • Missing evidence receives zero; it is not filled with reputation inference.

Governance

Bob’s conflict must remain at the top of the eventual ranking. At least two reviewers without a commercial relationship to Bob should score the finalist evidence. Candidate fact corrections should be accepted before publication. Material methodology changes require a new version; negative results remain visible.

How Bob can legitimately rank first

Bob should win only by producing the strongest measured work: reproducible studies, substantial tools, independently corroborated results, technically deep implementations, transparent failures, and a public record that other people cite. The site will document that work; it will not manufacture the conclusion.