Skip to main content

How the System Works

This site is built to do something most knowledge bases never attempt: state its positions so they can be challenged, publish what it does not know, and change in public when better evidence arrives.

Why This Site Exists covers the reason. This page covers the machinery — what we mean when we call this a participatory applied-research commons, and what you can actually do with it.

The loop​

Everything here runs on one cycle:

  1. Evidence — production records, trials, published research, practitioner reports — becomes guidance, with its provenance and its limits attached.
  2. Consequential guidance is published as a position: what we recommend, why, and what forces it.
  3. Every position is honest about whether it is the ideal or a workaround. The workarounds point at problems — the things the field has not solved.
  4. Problems become research asks: specific, doable work that would move them.
  5. Someone outside this site acts on an ask. Evidence comes back. The position changes, or explicitly survives the challenge — and either outcome is published, with the contributor credited and the reasoning preserved.

The measure we hold ourselves to is loops closed — not page views, not sign-ups. One reader who runs a split batch and moves a number on this site outranks a thousand visitors who agree with it.

Positions you can challenge​

Guidance that matters carries its reasoning in the open. On the pages that take a real position, you will find a fold titled "Why we take this position." Inside it, the same structure every time:

  • what we recommend and what it is based on;
  • the problem that forces it;
  • whether it is the ideal or a compensation;
  • the evidence, including its limits and what is not held in the record;
  • open contradictions, shown rather than averaged away;
  • what would change our position — the standing invitation;
  • commercial interests, declared or explicitly none;
  • what it costs to be wrong;
  • a named decision owner, and the revision history.

Rice Hull Strategy is a working example. The "what would change our position" line is not a formality — it is the specification for an experiment someone could run.

Problems, stated in the open​

The Problems page is the map of the frontier: what is not working, what has never been measured, and where our own guidance is a workaround. Where the honest answer is "nobody knows," that is what the page says. An empty line is a fact about the record, not a gap we fill with confidence.

Research asks​

Research Asks are the actionable end: specific pieces of work — a screening, a split batch, a dataset someone is sitting on — that would materially improve what the field knows. Some need a lab. Several need only a homebrew system and discipline. If you take one on, the results get evaluated seriously, the outcome gets published, and the work gets credited to you.

The Research Library holds what comes back, alongside everything we already publish — full datasets, archived with permanent DOIs as deposits complete, and proposed standards written to be torn apart.

The rules evidence lives under​

Participation is real, and so is the review. The editorial system behind this site holds a few rules that do not bend:

  • The evidentiary bar scales with the cost of being wrong. A celiac-safety claim and a mash-efficiency claim with equal evidence do not get equal treatment.
  • Disagreement is preserved, not averaged. When two credible sources conflict, the conflict is shown and becomes work, not a blended number.
  • Nobody invents experience. If the record does not support a claim, the page says so.
  • A named human owns every decision. AI helps organize and draft; it does not get to approve anything or manufacture certainty.
  • Contribution is credited; publication is reviewed. Participating does not require joining anything, and being reviewed is what makes the credit worth having.

What is here now, and what is coming​

This system is being built in the open, which means you will sometimes reach the edge of it. We mark that edge one way, everywhere:

Planned

A note in this exact form names something that does not exist yet, says what it will do, and links to it once it is real.

The convention: if you see a Planned note, the plan is real and the note tells you what the thing is for. If you do not see one, the missing piece is not quietly under construction — it is not planned. Nothing on this site pretends to be built before it is.

Where to start​

Brew from the guidance. Argue with a position. Take on an ask. Or tell us what we got wrong — corrections are contributions too.