wc-ownership-meta — effective ownership, the field model, and the leverage math

Implements footballfantasy/context/frameworks/game-theory-meta.md. This game is scored against the whole field at once, not a weekly opponent, so the entire opponent collapses into one number per player: ownership. This skill is the engine that turns ownership into the rank-relative leverage the system runs on — the ownership_leverage term in wc-fitness-eval, the differential/cover read on the decision board, and the per-pick leverage multiplier the strategists weight by.

The one sentence that governs everything below: rank moves on the gap between your points and the field's points, so what matters is never "did my player score?" but "did my player score relative to what the field owns?"

wc-ownership-analyst calls this to build its read; the strategists call it to weight picks; wc-fitness-eval calls it for the rank term. Never re-derive ownership downstream — read this skill's ownership signal.

Workflow

- [ ] 1. Gather field ownership% per relevant player (web search — public ownership trends) and captaincy share for the round
- [ ] 2. Compute EO = field_ownership% x (1 + captaincy_share) for every player in play; classify each into the 3 regimes vs our squad
- [ ] 3. Build the TEMPLATE SET (high-EO core you cannot afford to miss) and the DIFFERENTIAL LIST (live-ceiling low-ownership picks)
- [ ] 4. Run the FIELD MODEL — project "what will most managers do?" this round; flag over-concentration (fade) and under-reaction (get-there-first)
- [ ] 5. Compute ownership_leverage(c | θ) for wc-fitness-eval: cover-penalty when θ=protect, differentiate-reward when θ=gain
- [ ] 6. Overlay mini-league rival-relative leverage (block a rival you lead; match a rival's punt you trail)
- [ ] 7. Emit the `ownership` signal with field_ownership, effective_ownership, template_set, differential_list, and the flags