Phase 4b (rosters): per-zone outlier re-classification

Phase 4-A's per-monster trace named one or two miscategorised entries
on each of the four outlier zones — Phase 2c's single-boss-elite
dilution pattern hadn't been applied uniformly, and several over-tier
creatures were sitting on standard slots they couldn't carry.

Roster changes (no monster stat-block touches — bestiary is shared):

  crypt_valdris (T1, 8.5% → 46.0%):
    Dropped flameskull (CR3, 68 kills — still in underforge T3 where
    its CR fits). Promoted specter to IsElite at SW=3 as the soft
    dilutor, mirroring Phase 2c's worg/owlbear shape so the elite
    bracket no longer auto-picks a one-shot.

  forest_shadows (T2, 0.0% → 7.5%):
    Standard pool was carrying two real killers — Displacer Beast
    (38% win) and Bandit Captain (57% win, 14hp loss/win). Both
    re-flagged IsElite; Owlbear SW cut 4→2 to keep the now-4-elite
    pool diluted instead of front-loading Owlbear (9% win) at 50% of
    elite rolls. Standards collapse to Dire Wolf + Dryad, matching
    sunken_temple's "all standards ≥90% win" floor.

  manor_blackspire (T3, 0.5% → 14.5%):
    Wraith was the dominant standard killer (85 attributed kills,
    45hp loss per win) — too punishing for the standard slot.
    Promoted to IsElite; standard floor becomes
    Shadow+Poltergeist+Banshee (all ≥92% win in trace).

  abyss_portal (T5, 0.0% → 25.0%):
    Nalfeshnee was a CR-13 demon on a T5 standard slot (2.8% win, 86
    kills); promoted to IsElite. Second pass: Vrock at 86% win as
    standard still bled 65 kills via 43hp/win attrition — promoted to
    elite too. Standard collapses to Quasit alone (99.9% win, 1.3hp
    loss), matching dragons_lair's near-zero standard-pool kill rate.

Tier-level effect on the Phase 3-B (b=50, f=tier) sweep:

  T1: 26.8% → 44.0%  (+17 pp)
  T2:  6.2% → 10.0%  (+4)
  T3:  2.5% →  8.5%  (+6)
  T4:  7.8% →  7.8%  (no T4 outlier identified)
  T5: 27.5% → 41.0%  (+14)

Outliers are no longer outliers (forest_shadows 7.5% vs sibling 13%
gap; abyss_portal 25% vs sibling 57.5%); the remaining T2/T3/T5
shortfall vs the target band (62-82%/54-74%/36-56%) is tier-wide —
both zones at each problem tier sit below band, so a per-zone tool
won't close it. Phase 4's exit condition ("all per-zone cells within
band") is therefore conditioned on a follow-up tier-wide pass, but
the per-zone bug-fix it was scoped to address is done.

-short suite: same two pre-existing failures
(TestAdv2Scenario_ZoneRunGoblinWarrens, TestMageSpellbookLineInRender).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

internal/plugin/dnd_zone.go

@@ -261,12 +261,21 @@ func zoneCryptValdris() ZoneDefinition {
 		MinRooms:   6,
 		MaxRooms:   7,
 		Enemies: []ZoneEnemy{
+			// Phase 4-B (outlier fix): the live roster was the only
+			// dual-killer-elite zone (wight + flameskull, both CR3+ on
+			// the T1 standards floor), so the InterruptElite bracket
+			// was effectively "pick your one-shot." Phase 2c skipped
+			// this zone because it was already dual-elite; Phase 4-A's
+			// per-monster trace caught the consequence (167/183
+			// deaths attributed to those two). Flameskull dropped here
+			// (still in underforge T3 where its CR fits); specter
+			// promoted from standard to the soft-elite dilutor slot
+			// at SW=3, matching Phase 2c's worg/owlbear shape.
 			{BestiaryID: "skeleton", SpawnWeight: 7},
 			{BestiaryID: "zombie", SpawnWeight: 6},
 			{BestiaryID: "shadow", SpawnWeight: 4},
-			{BestiaryID: "specter", SpawnWeight: 3},
+			{BestiaryID: "specter", SpawnWeight: 3, IsElite: true},
 			{BestiaryID: "wight", SpawnWeight: 1, IsElite: true},
-			{BestiaryID: "flameskull", SpawnWeight: 1, IsElite: true},
 		},
 		Boss: ZoneBoss{
 			BestiaryID:  "boss_valdris_unburied",
@@ -308,11 +317,29 @@ func zoneForestShadows() ZoneDefinition {
 		MinRooms:   6,
 		MaxRooms:   8,
 		Enemies: []ZoneEnemy{
+			// Phase 4-B (outlier fix): standard pool was carrying two
+			// real killers — Displacer Beast (38% win as a standard
+			// pick) and Bandit Captain (57% win). Re-flag Displacer
+			// Beast as elite so it leaves the standard slot, joining
+			// Owlbear/Hag as a third elite for pool dilution; trim
+			// Bandit Captain SW so the soft Dire Wolf carries more of
+			// the standard roll mass. Boss + standards now match the
+			// sibling sunken_temple's lethality shape.
+			//
+			// Second pass: a first attempt (Displacer→elite alone)
+			// only moved comp 0% → 0.5%. Phase 4-A re-trace showed
+			// Bandit Captain still bleeding the standard pool (321
+			// appearances, 54.8% win, 53 kills) and Owlbear monopolising
+			// the elite roll at SW=4. Promoted Captain to elite as
+			// well; cut Owlbear SW 4→2 so the 4-elite pool dilutes
+			// instead of front-loading Owlbear (9% win). Standard
+			// becomes Dire Wolf + Dryad, matching sunken_temple's
+			// "all standards ≥90% win" floor.
 			{BestiaryID: "dire_wolf", SpawnWeight: 6},
-			{BestiaryID: "bandit_captain", SpawnWeight: 4},
-			{BestiaryID: "owlbear", SpawnWeight: 4, IsElite: true},
 			{BestiaryID: "dryad_corrupted", SpawnWeight: 3},
-			{BestiaryID: "displacer_beast", SpawnWeight: 3},
+			{BestiaryID: "owlbear", SpawnWeight: 2, IsElite: true},
+			{BestiaryID: "bandit_captain", SpawnWeight: 3, IsElite: true},
+			{BestiaryID: "displacer_beast", SpawnWeight: 3, IsElite: true},
 			{BestiaryID: "green_hag", SpawnWeight: 1, IsElite: true},
 		},
 		Boss: ZoneBoss{
@@ -399,10 +426,17 @@ func zoneManorBlackspire() ZoneDefinition {
 		MinRooms:   7,
 		MaxRooms:   9,
 		Enemies: []ZoneEnemy{
+			// Phase 4-B (outlier fix): Wraith was the dominant
+			// standard-pool killer (45 hp loss/win, 85 attributed
+			// kills in Phase 4-A) — too punishing for the standard
+			// slot at T3. Promoted to elite; Spawn+Wraith+Revenant
+			// becomes a 3-elite pool with balanced dilution, leaving
+			// Shadow/Poltergeist/Banshee as a clean standard floor
+			// (all ≥92% win in the trace).
 			{BestiaryID: "shadow", SpawnWeight: 6},
 			{BestiaryID: "poltergeist", SpawnWeight: 5},
 			{BestiaryID: "banshee", SpawnWeight: 3},
-			{BestiaryID: "wraith", SpawnWeight: 3},
+			{BestiaryID: "wraith", SpawnWeight: 3, IsElite: true},
 			{BestiaryID: "vampire_spawn", SpawnWeight: 3, IsElite: true},
 			{BestiaryID: "revenant", SpawnWeight: 1, IsElite: true},
 		},
@@ -635,10 +669,30 @@ func zoneAbyssPortal() ZoneDefinition {
 		MinRooms:   9,
 		MaxRooms:   10,
 		Enemies: []ZoneEnemy{
+			// Phase 4-B (outlier fix): Nalfeshnee was mis-classified
+			// as a standard at T5 — Phase 4-A measured 2.8% win rate
+			// across 180 standard-pool picks, attributing 86 kills
+			// (43% of all deaths in the trace). Re-flagged as elite
+			// where its CR-13 stat block belongs; the new elite pool
+			// Hezrou+Nalfeshnee+Marilith mirrors the sibling
+			// dragons_lair's three-elite shape. Vrock SW shaved 5→4
+			// so the soft Quasit (≥99% win) takes a larger share of
+			// the standard mass, since Vrock was the secondary
+			// standard killer (53 kills, 44hp loss/win).
+			//
+			// Second pass: a first attempt (Nalf→elite, Vrock SW 5→4)
+			// only moved comp 0% → 9%. Re-trace showed Vrock still
+			// dominant in the standard pool (1013 appearances, 86.3%
+			// win, but 43.6 hp loss per win → 65 kills via attrition).
+			// The sibling dragons_lair posts zero standard-pool kills;
+			// promoting Vrock to elite brings abyss in line. Standard
+			// floor collapses to Quasit alone, which is fine — Quasit
+			// posts ≥99.9% win in trace and the elite bracket
+			// triggers ~7% of rolls anyway.
 			{BestiaryID: "quasit", SpawnWeight: 6},
-			{BestiaryID: "vrock", SpawnWeight: 5},
+			{BestiaryID: "vrock", SpawnWeight: 4, IsElite: true},
 			{BestiaryID: "hezrou", SpawnWeight: 4, IsElite: true},
-			{BestiaryID: "nalfeshnee", SpawnWeight: 2},
+			{BestiaryID: "nalfeshnee", SpawnWeight: 3, IsElite: true},
 			{BestiaryID: "marilith", SpawnWeight: 1, IsElite: true},
 		},
 		Boss: ZoneBoss{