Tyler’s Methodology

Tyler’s methodology (the Fracture Clock) runs on a few stacked layers:

1. Time primitives
Three distinct clocks, not one: calendrical time (the calendar), event time (what actually happened), and Produced Delay — a newer addition distinguishing ordinary observed delay from delay a platform profits from maintaining (i.e., the gap itself becomes a monetized asset for whoever controls it).

2. The core causal chain
Every fracture is scored through: Action X → Consequence Y → Actor Z must respond → New equilibrium W (where the pre-X state is no longer reachable). This is the operational definition of “fracture” — not just a notable event, but one that structurally forecloses a prior baseline.

3. Irreversibility Classification (the scoring taxonomy)
Four categories, plus a permanent Escalation sub-type:

• Military Thresholds (first force use, mobilization, boundary crossings)
• Diplomatic Commitments (public statements as limiting records, formal pacts, UNSC/ICC referrals)
• Economic/Logistical (hard-to-lift sanctions, supply-chain rerouting, infrastructure damage with multi-year repair)
• Escalation Dynamics (retaliatory cycles, precedent-setting actions, domestic political pressure)

4. Discrete vs. Compounding sub-typing
Within Escalation Dynamics: Discrete fractures have one datable Action X and log directly to a state transition. Compounding fractures have no single trigger — closure emerges across repeated cycles, so they’re logged as provisional until a measurable baseline shift is confirmed across multiple data points, and only upgraded to a full state transition once reversal would require overcoming sunk behavioral/identity cost.

5. Built-in falsification
A Compounding classification is invalidated by: rapid non-lagged recovery after a policy reversal, evidence the trend was already reverting before logging, or a data correction showing the original reading was an outlier. This is the piece that separates it from pure narrative pattern-matching — there are explicit conditions under which a call is wrong.

6. The Prediction Ledger discipline
Intake-before-scoring (the call is logged before the outcome is known), full-population transparency (misses are logged too, not cherry-picked), and anti-gaming safeguards — this is the evidentiary backbone that lets track-record entries like the Franc Call or Negative Rate Call be checked against what was actually claimed at the time, not retrofitted after the fact.

7. Proposed extensions
Two newer ledger fields — reach trajectory and echo-to-recovery ratio — aimed at tracking how a call’s visibility and correction dynamics evolve after logging, on top of whether the call itself was right.

In short: it’s a causal-chain classification system for identifying which events cross a structural point of no return, timestamped and scored against a falsifiable rubric, with the ledger designed to make the track record auditable rather than asserted.