Playbook.
/advocacy is built around a two-surface schema and a publication discipline. Every intervention has a public five-field card and a deep page with twenty-five structured fields. Every claim has a citation, a coalition map, an adversary list, and a falsifier. Every template is bounded by length and named by failure mode. Every adoption is tracked publicly.
The discipline exists because the alternative is what most policy sites already are: position papers without standing, claims without falsifiers, coalitions without tiering, and templates that ship without acknowledging the failure modes they meet on first contact with a real legislative calendar. The schema is the structural answer to those failure modes. This page lays it out.
Every intervention targets a specific substrate. The substrate landscape — the six deepest non-biological substrates of modern society, the fifteen-layer compact stack, the four-phase expansion roadmap — lives at /advocacy/substrates. The Playbook here is the methodology applied within that landscape; every Phase 2-4 domain reuses these primitives.
Every intervention has two surfaces: a public card for the Tuesday-morning advocate, and a deep page for the school board member, the journalist, the congressional staffer who needs to defend the case in print or in committee.
- Fight — one sentence: what this intervention does.
- Who Wins — the substrate that returns to the public.
- Who Loses — named adversaries and money flow.
- What You Do Monday — the Tuesday-morning bundle: 200-word talking point, link to a 3-page resolution, link to a 1-page objection responder.
- Status — vehicle status (Live / Dormant / Pre-political / Post-political), jurisdictions adopting count, scoreboard entry.
The audience is school board members, journalists, congressional staff, and the parents in their districts who would like to intervene before they have to read a fifty-page memo. The register is plain English on the public card; framework primitives are reserved for the deep page where the reader is asking for them.
Klein-tier accessibility. Galloway-tier provocative framing where it earns its place. Schneier-tier honest about adversaries. No marketing language. No emoji. No declarative mood-affiliation that would make the intervention easier to dismiss as tribal.
Most policy work isn't substrate-engineering. That's not the problem. The problem is pretending it is. The filter — who pays the maintenance, who's defaulting — separates the work that repairs substrate from the work that does something else, often something useful, but categorically different.
Below: the twenty-two categories of intervention that fail the filter, with named structural reasons and real examples. Substrate-engineering becomes legible by contrast. Click any row for the deep diagnosis and the constructive pivot.
Knowing the difference doesn't make the other categories useless — most have legitimate roles. It does mean advocacy budgets get spent on what they can actually move. The filter below tests proposed interventions against the substrate-engineering test.
The filter as a testable instrument. Three diagnostic questions, deterministically classified — no language model, no inference, just rules. Paste an intervention; the strip develops; the verdict is named, with the matching category from the twenty-two above where applicable.
The classifier is rule-based and published; when it gets a case wrong, send the reasoning. The rule table grows with new cases.
The canonical synthesis of this section — the four-step pedagogy, the three required components, and the framework's full posture on measurement — lives at /advocacy/kpis. Read that page first if you arrived here looking for "what counts as a KPI?"
Every substrate-engineering intervention needs its own measurement reform. Otherwise the inverse pedagogy is incomplete — naming what fails and naming what would work, without measuring whether the work actually delivers, returns the substrate move to the same theater the filter was meant to expose.
Each substrate move ships with three required components:
- A rigorous KPI for the substrate condition itself — measure the substrate's health (interpretive capacity, attention residence, civic-decision quality), not the symptom (test scores, screen-time minutes, voter turnout).
- A gameable KPI to avoid — Goodhart's Law applied. Name the measure opponents will pivot to once the rigorous one becomes the target.
- The mitigation that makes the rigorous KPI usable — typically triangulation across sources, behavioral verification, or stratified sampling that resists coaching.
- Filter exposes the proposed intervention.
- Category names what it fails as, when it fails.
- Adjacent intervention names what would actually work.
- Measurement discipline names how we'd know it did.
Each step is necessary. Skipping step four turns substrate-engineering into the same theater the filter was meant to expose.
The K-12 Measurement Reform agenda is the canonical implementation of these four cross-cutting principles. Phase 2-4 domain expansions (healthcare, civic infrastructure, information environment) reuse them, with domain-specific construct families and instruments.
Every domain's measurement should distinguish diagnostic, formative, public reporting, accountability, and research tiers — each with its own audience, stakes, and validity-evidence requirements. The current collapse of all five into a single scoreboard is the core failure mode across substrate domains, not a K-12 idiosyncrasy.
Every public metric ships with structured documentation: construct, instrument, population, reliability, validity evidence, use license, forbidden uses, known gaming risks, equity risks, update cadence, uncertainty display, audit trail. A metric without its passport is not a measurement; it is a rumor with a citation.
Canonical implementation — K-12 Measurement Reform →For any substrate KPI surfaced on /advocacy, the gaming model is published alongside the metric, not in a separate audit document. Each metric, the predictable gaming behavior it produces under consequential pressure, and the structural countermeasure (typically triangulation, sampled audit, or uncertainty display). Required, not optional.
Some data must be firewalled from punitive use to enable honest improvement. The principle is cross-cutting: healthcare improvement data firewalled from malpractice discovery, civic-infrastructure incident reports firewalled from political reprisal, classroom formative data firewalled from teacher-evaluation pipelines. Without protected diagnostic space, the system stops being able to see itself.
The Measurement Reform page is the canonical implementation. The Playbook is where the principles live for cross-domain reuse.
- Not a manifesto. Manifestos commit to positions; this site commits to falsifiers.
- Not a closed system. Phase 1 is K-12. Phases 2-4 will extend and may revise.
- Not a comments section. Astroturf vector closed by design. Submissions through /advocacy/contribute.
- Not finished. Versioned, dated, change-logged. Corrections welcome on a 14-day SLA.