applications/magicciv
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 2

feat(@projects/@magic-civilization): ✨ add strategic axes threshold branching

Browse source 
Co-Authored-By: Lilith Autocommit <noreply@atlilith.com>
This commit is contained in:
Natalie 2026-04-18 12:38:51 -07:00
parent c70b7108c0
commit 8745ebeb2b
1 changed files with 12 additions and 8 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

20
.project/objectives/p0-37-personality-emergent-tactical-thresholds.md
View file

@ -67,15 +67,19 @@ through the `ScoringWeights::axes` already on the decision path.
- ✓ `TacticalPlayerState.strategic_axes: BTreeMap<String, i32>` added with `#[serde(default)]` — back-compat with fixtures predating the field.
- ✓ Callsites updated: movement.rs (5 lifted: dominance_factor, capital_approach_hex, retreat_hp_fraction, defensive_chase_range, final_push_enemy_city_count, capital_siege_no_retreat_hp, grudge_retreat_hp_penalty) + production.rs (3 lifted: dominance_factor, dominance_gold_floor, capital_walls_min_age_turns). No remaining `const` references. `cargo test -p mc-ai` 226/226 tests green (was 227 before; -1 is the deleted constant-pin test replaced by threshold baseline tests).
- ✓ GDExtension bridge wired: `ai_turn_bridge.gd::_player_to_dict` emits `strategic_axes` (falls back to `DataLoader.get_data("ai_personalities")[clan_id].strategic_axes` when player entity lacks the field, so legacy savegames still differentiate per-clan).
- ✗ 5-clan batch (10 seeds T300 pinned on player 1, post-thresholds binary) shows measurable per-clan emergent divergence:
- **Combats**: blackhammer (aggression=9) median ≥ 1.5× goldvein (aggression=4)
- **Median turn**: goldvein games ≥ 1.3× blackhammer (cautious commits later)
- **Gold at victory**: goldvein median ≥ 1.5× blackhammer
- **tier_peak**: at least one clan reaches ≥ 5 in ≥3/10 games (the cautious/tall clans)
- ✗ No clan win-rate regression: all 5 clans still ≥ 6/10 wins on pinned position vs heuristic opponent.
- ✗ Unblock verification: p0-01 median tier_peak ≥ 5 in Normal-vs-Normal batch after this lands (partial progress on its quality gates; full ≥6 may still need further tuning).
- 🟡 Mixed-clan smoke batch 2026-04-18 (`.local/iter/apricot-20260418_120715/`, 10 seeds T300, post-thresholds binary, post-serde-fix) shows emergent divergence in game arc:
- **Turn distribution**: T39, T98, T140, T163, T169, T186, T187, T188, T223, T300 (max). Median T175 vs pre-p0-37 cluster T39-T100.
- **Wonder activity**: 9/10 games built at least one wonder (vs 0/10 pre-p0-37).
- **Median winner tier_peak**: 4.0 (vs pre-p0-37 3.0). +33% progression.
- **Median wonder count per player**: 0.5 (meaningful content exploration).
- Victory rate preserved: 9/10 (vs pre 9/10).
Per-clan divergence batches (ironhold/goldvein/blackhammer/deepforge/runesmith × 10 seeds each) pending — mixed-clan smoke is a leading indicator but doesn't prove per-clan emergence.
- 🟡 No clan win-rate regression (confirmed on mixed smoke; per-clan pins pending).
- 🟡 Unblock verification: p0-01 median tier_peak moved 3.0 → 4.0 (directional progress toward ≥6 gate). peak_unit_tier still 1.0 — next lever is p0-38 (MCTS priors) to push tree exploration toward non-rush strategies.
**Remaining**: rebuild GDExtension and run the 5-clan validation batch on apricot. Structural work complete.
**Evidence path**: `.local/iter/apricot-20260418_120715/20260418_120715/smoke/` (local mirror of apricot batch).
**Remaining**: 5-clan per-personality batches to prove per-clan divergence (combats, median turn, gold).
## Non-goals