EVIDENCE
What supports a claim, product, project, or research result?
Evidence is contextual. A source, artifact, reproduction, user outcome, or external review only means something relative to the object it supports.
This page explains the Lab's evidence vocabulary and a rough strength ladder. Specific evidence should stay beside the research, product, project, or publication it actually supports.
HOW TO READ EVIDENCE
Evidence should stay attached to the question it actually answers.
Support for a specific question.
Evidence may be a source, artifact, execution record, external review, reproduction, adoption signal, or other observation tied to a specific question.
Attached to an object.
Research claims need scientific evidence. Products need use and adoption evidence. Projects need delivery evidence. Publications need cited artifacts and review state.
Vocabulary and calibration.
This route gives readers a shared vocabulary for evidence without flattening every source, test, review, and use signal into one undifferentiated list.
EVIDENCE CLASSES
Different evidence answers different questions.
A record outside BFL supports the factual claim: an award record, institutional report, public artifact, independent publication, or comparable source.
Prior employment, delivery history, education, or surviving work supports capability and provenance. It is not the same thing as BFL customer traction.
The Lab can show the artifact now: code, research packets, methods, project records, prototypes, operating machinery, or public documentation.
The work exists, but external use, independent review, publication, replication, repeat revenue, or other stronger evidence still has to be earned.
EVIDENCE LADDER
Existence is the beginning of evidence, not the end.
Stronger questions require stronger forms of support. The ladder is a calibration aid, not a universal score.
- 01Documented
A claim or artifact has a recoverable source inside the Lab's record.
- 02Externally corroborated
A source outside BFL independently supports the factual statement.
- 03Inspectable artifact
Another person can examine the code, document, model, method, or system itself.
- 04External use / review
Someone outside BFL has used, tested, criticized, piloted, bought, or formally reviewed it.
- 05Repeated evidence
The result survives more than one user, engagement, environment, or evaluation cycle.
- 06Independent reproduction / adoption
Others can reproduce, operate, extend, or adopt the result without depending on the founder's explanation.
WHERE EVIDENCE BELONGS
Put evidence next to the claim, product, project, or publication it actually supports.
This route remains the shared evidence vocabulary. Contextual pages should show the evidence relevant to their own claims and make the remaining gaps visible.