diff --git a/.project/simulation-report/scenarios/4ai/report.md b/.project/simulation-report/scenarios/4ai/report.md index a97fa356..6b96cddc 100644 --- a/.project/simulation-report/scenarios/4ai/report.md +++ b/.project/simulation-report/scenarios/4ai/report.md @@ -1,76 +1,76 @@ -# Scenario 4AI — Maximum Civilization Pressure (96×96) +# Scenario 4AI -- Maximum Civilization Pressure (96x96) -*96×96 map · 50,000 ecology ticks · 500 turns · 4 players · seed=42* +*96x96 map -- 50,000 ecology ticks -- 500 turns -- 4 players -- seed=42 -- post-iter-7o sweep (2026-04-09)* ## Summary -**Four civilizations on the largest map — and more balanced than 3AI despite more lairs.** 447 lairs across 9,216 tiles generate 1,245 encounters. T7-T10 kill rate is 73% (3% over ceiling — the fourth player distributes encounters enough to bring the rate down from 80×80's 81%). T4-T6 at 11% meets target. 11 named lair-former species; Mugger Crocodile and all caiman species fully established across the map. +**Four civilizations on the largest map -- both balance targets met cleanly.** 672 lairs across 9,216 tiles generate 31,136 encounters. T7-T10 kill rate is 63% (inside 40-70%). T4-T6 kill rate is 24% (inside 10-30%). 11 named lair-former species. Rat Snake holds 220 lairs -- the largest single-species presence in the sweep. + +The 4AI scenario confirms that balance holds at maximum player count and map size. T7-T10 at 63% matches the 3AI result exactly, demonstrating that the tier-kill curve is stable regardless of additional players. ## World State | Metric | Value | |--------|-------| -| Total lairs | 447 / 9,216 (4.9%) | +| Total lairs | 672 / 9,216 (7.3%) | | T1-T3 lairs | 0 | -| T4-T6 lairs | 165 | -| T7-T10 lairs | 282 | +| T4-T6 lairs | 149 | +| T7-T10 lairs | 523 | | Max ecosystem tier | T10 | -| Avg ecosystem tier | 5.88 | -| Evolution time | 2317.2s | +| Avg ecosystem tier | 6.84 | +| Evolution time | 1,640.3s | ## Lair-Former Species -| Species | Lairs | Tier | Peak Pop | Territory Tiles | -|---------|-------|------|----------|-----------------| -| Rat Snake | 118 | T10 | 96.8 | 3,052 | -| Broad-Snouted Caiman | 84 | T10 | 49.2 | 3,324 | -| Spectacled Caiman | 69 | T10 | 118.1 | 2,637 | -| Yacare Caiman | 60 | T10 | 47.2 | 1,788 | -| Tropical Rat Snake | 40 | T10 | 49.2 | 1,270 | -| Dwarf Crocodile | 18 | T10 | 54.1 | 432 | -| Black Caiman | 16 | T10 | 66.1 | 628 | -| Mugger Crocodile | 10 | T10 | 67.2 | 328 | -| Dhole | 11 | T8 | 45.6 | 329 | -| Cave Lion | 1 | T7 | 5.9 | 19 | -| Sand Boa | 20 | T5 | 28.7 | 140 | +| Species | Lairs | Tier | Peak Pop | Territory Tiles | Radius | +|---------|-------|------|----------|-----------------|--------| +| Rat Snake | 220 | T10 | 238.9 | 6,958 | r=4 | +| Spectacled Caiman | 137 | T10 | 180.7 | 6,569 | r=4 | +| Broad-Snouted Caiman | 96 | T10 | 134.6 | 4,494 | r=4 | +| Tropical Rat Snake | 94 | T10 | 126.7 | 2,908 | r=4 | +| Yacare Caiman | 36 | T10 | 172.5 | 1,542 | r=4 | +| Dwarf Crocodile | 23 | T10 | 91.7 | 1,115 | r=4 | +| Dhole | 21 | T10 | 363.2 | 807 | r=4 | +| Black Caiman | 14 | T10 | 70.9 | 776 | r=4 | +| Mugger Crocodile | 5 | T10 | 110.0 | 281 | r=4 | +| Cave Lion | 1 | T9 | 5.9 | 37 | r=3 | +| Sand Boa | 25 | T7 | 143.6 | 427 | r=2 | -**All crocodilian lair-formers fully realized** — Mugger Crocodile reaches 10 lairs at T10, Broad-Snouted Caiman leads the field with 84 lairs. Dhole expands to 11 lairs at T8. Rat Snake peaks at 118 lairs with populations of 96 per tile. +**All lair-formers except Cave Lion and Sand Boa reach T10.** Dhole peaks at 363.2 population -- the highest peak population of any species across any scenario. Rat Snake holds 220 lairs across 6,958 nominal territorial tiles. ## Turn-by-Turn Snapshots | Turn | P0 | P1 | P2 | P3 | P0 C | P1 C | P2 C | P3 C | |------|----|----|----|----|------|------|------|------| -| 50 | 0 | 0 | 1 | 3 | 4 | 7 | 4 | 5 | -| 100 | 1 | 2 | 8 | 3 | 6 | 11 | 6 | 8 | -| 200 | 24 | 51 | 36 | 21 | 9 | 17 | 9 | 12 | -| 350 | 61 | 101 | 92 | 75 | 13 | 23 | 13 | 17 | -| 500 | 130 | 207 | 153 | 145 | 17 | 28 | 17 | 21 | +| 50 | 19 | 55 | 8 | 11 | 5 | 11 | 5 | 7 | +| 100 | 35 | 200 | 18 | 31 | 9 | 21 | 9 | 13 | +| 200 | 68 | 593 | 156 | 247 | 17 | 26 | 17 | 20 | +| 350 | 129 | 611 | 204 | 286 | 17 | 26 | 17 | 20 | +| 500 | 125 | 611 | 199 | 279 | 17 | 26 | 17 | 20 | ## Encounter Pressure | Metric | Value | Target | |--------|-------|--------| -| Total encounters | 1,245 | — ✓ | -| Unit deaths | 801 | — | -| Overall kill rate | 64.3% | — | -| T4-T6 kill rate | **11%** (20/180) | 10-30% ✓ | -| T7-T10 kill rate | **73%** (781/1065) | 40-70% ⚠ | +| Total encounters | 31,136 | -- | +| Unit deaths | 18,334 | -- | +| Overall kill rate | 58.9% | -- | +| T4-T6 kill rate | **24%** (833/3,428) | 10-30% pass | +| T7-T10 kill rate | **63%** (17,501/27,708) | 40-70% pass | ## Player Final State -| Player | Profile | Cities | Units | Gold | Produced | -|--------|---------|--------|-------|------|----------| -| P0 | Militarist | 17 | 130 | 25,142 | 219 | -| P1 | Expansionist | 28 | 207 | 43,166 | 356 | -| P2 | Merchant | 17 | 153 | 25,592 | 229 | -| P3 | Scientist | 21 | 145 | 36,728 | 241 | +| Player | Profile | Cities | Units | Gold | Produced | Peak | +|--------|---------|--------|-------|------|----------|------| +| P0 | Militarist | 17 | 125 | 27,244 | 623 | 137 | +| P1 | Expansionist | 26 | 611 | 45,560 | 1,064 | 626 | +| P2 | Merchant | 17 | 199 | 68,020 | 585 | 213 | +| P3 | Scientist | 20 | 279 | 33,700 | 698 | 297 | ## Balance Assessment -**73% T7-T10 kill rate — 3% over ceiling.** The fourth player distributes encounter pressure across a fourth army, bringing the rate down 8 points from 80×80's 81%. Still marginally over ceiling. +**Both balance targets met cleanly.** T7-T10 at 63% matches the 3AI result -- the kill rate has converged to a stable value regardless of player count on large maps. T4-T6 at 24% is well inside the window. -**All T4-T6 scenarios now pass (11%).** With 180 mid-tier encounters, the 10-30% kill rate range is clearly confirmed statistically. Mid-tier lairs are a real but survivable threat on large maps. +**All four profiles viable and differentiated.** P1 the Expansionist dominates in units (611). P2 the Merchant dominates in gold (68,020). P3 the Scientist occupies the middle ground (279 units, 33,700 gold, 20 cities). P0 the Militarist has the smallest army (125) -- the production=5 weighting produces more units (623 total) but the Militarist's aggressive exploration puts more units into lair territory, consuming them at a higher per-player rate. -**Encounter saturation begins at 3 players.** Going from 3AI (920 encounters) to 4AI (1,245 encounters) shows the pattern: each additional player adds proportional encounters on a larger map (unlike 48×48 where encounters plateaued). Map-scaled runs don't saturate — more players on more tiles = more total encounters. - -**Carry-forward: same map-size balance issue as 3AI.** Unit production rate and starting army size need to scale with map configuration. The 4-player armies at T500 (130-207 units) are roughly half the 48×48 4-player armies (266-442 units) despite the same number of cities — indicating attrition outpaces production on large maps with current combat parameters. +**Previous vs current:** The iter 7d 4AI scenario ended with armies of 130/207/153/145 and a 73% T7-T10 kill rate. The iter 7o run ends with armies of 125/611/199/279 and a 63% kill rate. The Expansionist's army tripled. The Scientist and Merchant both grew significantly. The ecology is no longer punishing -- it is calibrated. diff --git a/.project/simulation-report/scenarios/4ai/stats.md b/.project/simulation-report/scenarios/4ai/stats.md index 1499852a..edbb9b3a 100644 --- a/.project/simulation-report/scenarios/4ai/stats.md +++ b/.project/simulation-report/scenarios/4ai/stats.md @@ -1,98 +1,115 @@ -# Simulation Stats — 4AI (Maximum Civilization Pressure) +# Simulation Stats -- 4AI (Maximum Civilization Pressure) -*96×96 map · 50,000 ecology ticks · 500 turns · 4 players · seed=42* +*96x96 map -- 50,000 ecology ticks -- 500 turns -- 4 players -- seed=42 -- post-iter-7o sweep (2026-04-09)* ## Run Parameters | Parameter | Value | |-----------|-------| -| Map size | 96×96 (9,216 tiles) | +| Map size | 96x96 (9,216 tiles) | | Evolution ticks | 50,000 | | Game turns | 500 | | Players | 4 | -| Named species loaded | 588 | +| Named species loaded | 589 | | Seed | 42 | -| RAYON_NUM_THREADS | 6 | -| Evolution time | 2317.2s | +| RAYON_NUM_THREADS | 32 | +| Evolution time | 1,640.3s | +| Total elapsed | 1,641.4s | ## Player Setup | Player | Profile | Start City | Anchor Lair | |--------|---------|------------|-------------| -| P0 | Militarist (exp=2 prod=5 wealth=2 culture=2) | — | T10 | -| P1 | Expansionist (exp=5 prod=3 wealth=2 culture=2) | — | T10 | -| P2 | Merchant (exp=2 prod=3 wealth=5 culture=2) | — | T10 | -| P3 | Scientist (exp=3 prod=3 wealth=2 culture=5) | — | T10 | +| P0 | Militarist (exp=2 prod=5 wealth=2 culture=2) | (29,24) | T10 at (29,27) | +| P1 | Expansionist (exp=5 prod=3 wealth=2 culture=2) | (89,23) | T10 at (89,26) | +| P2 | Merchant (exp=2 prod=3 wealth=5 culture=2) | (12,45) | T10 at (12,48) | +| P3 | Scientist (exp=3 prod=3 wealth=2 culture=5) | (50,45) | T10 at (50,48) | ## Lair Network | Metric | Value | |--------|-------| -| Total lairs | 447 | -| Lair density | 4.9% | +| Total lairs | 672 | +| Lair density | 7.3% | | T1-T3 lairs | 0 | -| T4-T6 lairs | 165 | -| T7-T10 lairs | 282 | +| T4-T6 lairs | 149 | +| T7-T10 lairs | 523 | ## Lair-Former Species Breakdown -| Species | Lairs | Max Tier | Peak Pop | Territory Tiles | -|---------|-------|----------|----------|-----------------| -| Rat Snake | 118 | T10 | 96.8 | 3,052 | -| Broad-Snouted Caiman | 84 | T10 | 49.2 | 3,324 | -| Spectacled Caiman | 69 | T10 | 118.1 | 2,637 | -| Yacare Caiman | 60 | T10 | 47.2 | 1,788 | -| Tropical Rat Snake | 40 | T10 | 49.2 | 1,270 | -| Dwarf Crocodile | 18 | T10 | 54.1 | 432 | -| Black Caiman | 16 | T10 | 66.1 | 628 | -| Mugger Crocodile | 10 | T10 | 67.2 | 328 | -| Dhole | 11 | T8 | 45.6 | 329 | -| Cave Lion | 1 | T7 | 5.9 | 19 | -| Sand Boa | 20 | T5 | 28.7 | 140 | +| Species | Lairs | Max Tier | Peak Pop | Territory Tiles | Radius | +|---------|-------|----------|----------|-----------------|--------| +| Rat Snake | 220 | T10 | 238.9 | 6,958 | r=4 | +| Spectacled Caiman | 137 | T10 | 180.7 | 6,569 | r=4 | +| Broad-Snouted Caiman | 96 | T10 | 134.6 | 4,494 | r=4 | +| Tropical Rat Snake | 94 | T10 | 126.7 | 2,908 | r=4 | +| Yacare Caiman | 36 | T10 | 172.5 | 1,542 | r=4 | +| Dwarf Crocodile | 23 | T10 | 91.7 | 1,115 | r=4 | +| Dhole | 21 | T10 | 363.2 | 807 | r=4 | +| Black Caiman | 14 | T10 | 70.9 | 776 | r=4 | +| Mugger Crocodile | 5 | T10 | 110.0 | 281 | r=4 | +| Cave Lion | 1 | T9 | 5.9 | 37 | r=3 | +| Sand Boa | 25 | T7 | 143.6 | 427 | r=2 | ## Turn-by-Turn Snapshots -| Turn | P0 | P1 | P2 | P3 | P0 C | P1 C | P2 C | P3 C | -|------|----|----|----|----|------|------|------|------| -| 50 | 0 | 0 | 1 | 3 | 4 | 7 | 4 | 5 | -| 100 | 1 | 2 | 8 | 3 | 6 | 11 | 6 | 8 | -| 200 | 24 | 51 | 36 | 21 | 9 | 17 | 9 | 12 | -| 350 | 61 | 101 | 92 | 75 | 13 | 23 | 13 | 17 | -| 500 | 130 | 207 | 153 | 145 | 17 | 28 | 17 | 21 | +| Turn | P0 | P1 | P2 | P3 | P0 C | P1 C | P2 C | P3 C | P0 Gold | P1 Gold | P2 Gold | P3 Gold | +|------|----|----|----|----|------|------|------|------|---------|---------|---------|---------| +| 50 | 19 | 55 | 8 | 11 | 5 | 11 | 5 | 7 | 556 | 1,160 | 1,300 | 752 | +| 100 | 35 | 200 | 18 | 31 | 9 | 21 | 9 | 13 | 1,852 | 4,260 | 4,540 | 2,644 | +| 200 | 68 | 593 | 156 | 247 | 17 | 26 | 17 | 20 | 6,844 | 14,360 | 17,020 | 9,700 | +| 350 | 129 | 611 | 204 | 286 | 17 | 26 | 17 | 20 | 17,044 | 29,960 | 42,520 | 21,700 | +| 500 | 125 | 611 | 199 | 279 | 17 | 26 | 17 | 20 | 27,244 | 45,560 | 68,020 | 33,700 | ## Encounter Data | Metric | Value | |--------|-------| -| Total encounters | 1,245 | -| Total deaths | 801 | -| Overall kill rate | 64.3% | +| Total encounters | 31,136 | +| Total deaths | 18,334 | +| Overall kill rate | 58.9% | | T1-T3 encounters | 0 | | T1-T3 deaths | 0 | | T1-T3 kill rate | 0% | -| T4-T6 encounters | 180 | -| T4-T6 deaths | 20 | -| T4-T6 kill rate | 11.1% | -| T7-T10 encounters | 1,065 | -| T7-T10 deaths | 781 | -| T7-T10 kill rate | 73.3% | +| T4-T6 encounters | 3,428 | +| T4-T6 deaths | 833 | +| T4-T6 kill rate | 24% | +| T7-T10 encounters | 27,708 | +| T7-T10 deaths | 17,501 | +| T7-T10 kill rate | 63% | + +## Units Produced + +| Player | Profile | Produced | Peak Army | +|--------|---------|----------|-----------| +| P0 | Militarist | 623 | 137 | +| P1 | Expansionist | 1,064 | 626 | +| P2 | Merchant | 585 | 213 | +| P3 | Scientist | 698 | 297 | ## Final Player State -| Player | Profile | Cities | Units | Gold | Produced | Peak | -|--------|---------|--------|-------|------|----------|------| -| P0 | Militarist | 17 | 130 | 25,142 | 219 | 130 | -| P1 | Expansionist | 28 | 207 | 43,166 | 356 | 207 | -| P2 | Merchant | 17 | 153 | 25,592 | 229 | 153 | -| P3 | Scientist | 21 | 145 | 36,728 | 241 | 145 | +| Player | Profile | Cities | Units | Gold | +|--------|---------|--------|-------|------| +| P0 | Militarist | 17 | 125 | 27,244 | +| P1 | Expansionist | 26 | 611 | 45,560 | +| P2 | Merchant | 17 | 199 | 68,020 | +| P3 | Scientist | 20 | 279 | 33,700 | ## Balance Check | Criterion | Target | Result | Status | |-----------|--------|--------|--------| -| T7-T10 kill rate | 40-70% | 73.3% | ⚠ (+3%) | -| T4-T6 kill rate | 10-30% | 11.1% | ✓ | -| Encounters | 100-300+ | 1,245 | ✓ | -| All players viable | yes | yes | ✓ | +| T7-T10 kill rate | 40-70% | 63% | pass | +| T4-T6 kill rate | 10-30% | 24% | pass | +| All players viable | yes | yes | pass | -**Known cause of T7-T10 over-ceiling:** Same map-size scaling mismatch as 3AI — combat stats tuned for 48×48. Fourth player distributes encounter pressure across a fourth army, bringing the rate down from 3AI's 81% to 73% — closer to ceiling but still 3% over. Unit production rate does not scale with map size: final armies (130-207 units) are roughly half the 48×48 4-player equivalents (269-442 units) despite same city count. +## Comparison to Previous (iter 7d) + +| Metric | Iter 7d | Iter 7o | Change | +|--------|---------|---------|--------| +| T7-T10 kill rate | 73% | 63% | -10pp | +| T4-T6 kill rate | 11% | 24% | +13pp | +| Total encounters | 1,245 | 31,136 | +25x | +| P1 final army | 207 | 611 | +3.0x | +| Lairs | 447 | 672 | +50% | diff --git a/.project/simulation-report/scenarios/4ai/story.md b/.project/simulation-report/scenarios/4ai/story.md index f7c77435..e1acc41b 100644 --- a/.project/simulation-report/scenarios/4ai/story.md +++ b/.project/simulation-report/scenarios/4ai/story.md @@ -1,58 +1,46 @@ # The Fully Realized World -*96×96 map · 4 players · 1,245 encounters · 801 unit deaths · 64.3% kill rate · seed=42* +*96x96 map -- 4 players -- 31,136 encounters -- 18,334 unit deaths -- 58.9% kill rate -- seed=42 -- post-iter-7o sweep (2026-04-09)* --- -## The Complete Roster +## The Complete Roster at Scale -At 96×96, every crocodilian lair-former is present and accounted for. Mugger Crocodile reaches 10 lairs at T10 — what appeared as a 2-lair novelty on the 80×80 map is now a real presence. Broad-Snouted Caiman leads the entire lair network at 84 lairs and 3,324 territory tiles. Spectacled Caiman peaks at population 118 per tile — the highest peak population of any species across any scenario. +At 96x96, every named lair-former except Cave Lion and Sand Boa reaches T10. 9 species at apex tier. Dhole peaks at a staggering 363.2 population per tile -- the highest peak population of any species across any scenario in the sweep. Rat Snake holds 220 lairs across 6,958 nominal territorial tiles. Spectacled Caiman holds 137 lairs across 6,569 tiles. -11 named species, all established. Five caiman species. Two rat snake species. Dhole at 11 lairs and T8. Cave Lion holding its single T7 den. Sand Boa at the mid-tier. The complete ecological hierarchy of this world. - -The 96×96 map is the first where Mugger Crocodile is not a marginal presence. With 328 territory tiles and peak population 67, it has established a coherent basin territory alongside the other four crocodilian species. +The ecology at 96x96 is vast. 672 lairs on 9,216 tiles, 7.3% density. 523 of those are T7-T10. The equatorial belt is a continuous wall of overlapping predator territories -- Rat Snake and Caiman territories stacking three or four deep in the densest zones. --- -## 1,245 Encounters +## 31,136 Encounters -Four civilizations on the largest map generate 1,245 encounters in 500 turns. No saturation — the 96×96 map has 9,216 tiles and 447 lairs, and four civilizations can't exhaust that encounter budget. The rate is consistent: ~2.8 encounters per lair over 500 turns, matching the 80×80 rate (920/316 = 2.9) and the 64×64 rate (431/159 = 2.7). The per-lair encounter rate is stable across all map sizes. +Four civilizations on the largest map generate 31,136 encounters in 500 turns. The encounter volume scales with player count and map area: 9,222 (1AI) to 19,618 (2AI) to 24,855 (3AI) to 31,136 (4AI). Each additional player adds proportional encounters on a proportionally larger map. -What scales is the absolute number: more lairs on more tiles generates proportionally more encounters. This is the intended property. Map-scaled scenarios don't hit a ceiling. +T7-T10: 63% (17,501 kills / 27,708 encounters). Matches the 3AI result exactly. The kill rate has converged -- adding a fourth player on a larger map does not change the per-encounter lethality. The tier-kill curve is stable. -180 T4-T6 encounters, 20 deaths: 11.1% kill rate. 1,065 T7-T10 encounters, 781 deaths: 73.3% kill rate. T4-T6 passes cleanly. T7-T10 is 3% over ceiling. +T4-T6: 24% (833 kills / 3,428 encounters). The highest T4-T6 kill rate in the sweep, but still comfortably inside the 10-30% window. 3,428 mid-tier encounters provide unambiguous statistical confirmation. --- -## The Fourth Player Distributes the Pressure +## Four Profiles, Four Strategies -3AI on 80×80 produced an 81% T7-T10 kill rate. 4AI on 96×96 produces 73%. The 8-point drop comes from two factors: the fourth player adds army mass that absorbs encounters, and the 96×96 map's additional area gives each civilization more route-around options relative to lair density. +The 4AI scenario is the only one with all four AI profiles active. Their divergence tells the story of how different strategies interact with the same ecology: -The map is proportionally larger (9,216 vs 6,400 tiles = 44% more area), but each civilization's share of map coverage decreases with the fourth player. Per-player expansion is more measured, which means per-unit exposure to T7-T10 lair zones is slightly lower. Four armies in four quadrants, each absorbing a quarter of the encounter budget. +**P1 the Expansionist** dominates in territory and military. 26 cities by turn 200, 611 units at turn 500, 1,064 total produced. The largest army in the sweep. The `expansion=5` weighting drives city founding, and more cities drives more production, and more production overwhelms the encounter attrition. -Still 3% over ceiling. The root cause is the same as 3AI: production rate doesn't scale with map size. The four final armies at T500 (130/207/153/145 units) are roughly half the 48×48 four-player equivalents (269/442/266/308 units) despite the same city count. +**P2 the Merchant** dominates in gold. 68,020 at turn 500 -- the highest gold total of any player in any scenario. The `wealth=5` weighting converts city count into economic power. 199 units is a modest army, but a viable one. The Merchant's value is not military. + +**P3 the Scientist** occupies the middle ground. 20 cities, 279 units, 33,700 gold. The `culture=5` weighting (labeled "scientist" in the profile system) produces a balanced civilization that is neither the strongest military nor the richest economy but is competitive in both. + +**P0 the Militarist** has the smallest final army at 125 units despite the highest production priority (`prod=5`). The paradox resolves: the Militarist's aggressive exploration puts more units into lair territory earlier and more frequently. 623 total units produced -- more than the Merchant's 585 -- but the attrition rate consumes them faster because the Militarist pushes into danger zones that the Merchant's more cautious expansion avoids. --- -## Broad-Snouted Caiman's Dominance +## Dhole's Ascent -On 48×48, Rat Snake held the most lairs. On 96×96, Broad-Snouted Caiman surges to 84 lairs — largest lair count of any species at this scale — while Rat Snake's 118 lairs still represent the largest single-species territorial presence by tile count (3,052 tiles). Both species have expanded with the available habitat. +The Dhole makes its most significant showing on the 96x96 map: 21 lairs at T10, population peaks of 363.2. On 48x48, the Dhole held 9 lairs. On 64x64, 5. On 80x80, 16. The pack predator scales with available habitat in a way the crocodilians and snakes do not -- its territory requirements are met by a wider range of biomes, and the 96x96 map provides enough highland and forest fringe for the Dhole to establish a major presence. -What shifts at 96×96: the basin habitats that support crocodilian species scale faster than the jungle interior habitats that support snakes. The map's additional area added proportionally more river-margin and swamp-edge tiles than highland jungle tiles, and the Caimans expanded into that space first. - -This is ecology-as-emergent-geography. The 96×96 world is not just a bigger 48×48 world. The habitat distribution at this scale shifts the species balance. - ---- - -## Strained Armies on an Oversized Map - -P0 the Militarist ends at 130 units. On 48×48 with the same profile and same city count, the Militarist ends at 269 units. The production formula is identical. The difference is attrition: on 96×96, the Militarist's expansion path crosses more T7-T10 lair territories on the way to each new city site, and those encounters kill units at 73%. - -P1 the Expansionist — normally the army-largest profile — ends at 207 units. On 48×48 it ends at 442. Same city count. 235 fewer units from the same production pipeline. - -The armies are roughly half. The gold is identical (the economic engine is unaffected by map size). The cities are identical. But the units are consumed at a rate that the production system wasn't tuned for at this map scale. - -Four civilizations remain viable at turn 500. All have functional armies. All are producing. The ecology didn't eliminate anyone. It just costs more than intended. +363.2 population per tile is the highest recorded for any species in any scenario. The Dhole is a pack species -- this population represents a dense pack structure that the ecology model supports. A player unit walking into Dhole territory at T10 faces a concentrated pack, not scattered individuals. --- @@ -61,8 +49,10 @@ Four civilizations remain viable at turn 500. All have functional armies. All ar | | P0 | P1 | P2 | P3 | |---|---|---|---|---| | Profile | Militarist | Expansionist | Merchant | Scientist | -| Cities | 17 | 28 | 17 | 21 | -| Units | 130 | 207 | 153 | 145 | -| Gold | 25,142 | 43,166 | 25,592 | 36,728 | +| Cities | 17 | 26 | 17 | 20 | +| Units | 125 | 611 | 199 | 279 | +| Gold | 27,244 | 45,560 | 68,020 | 33,700 | -Four civilizations. 447 lairs, all intact. The Mugger Crocodile finally got its territory. The Broad-Snouted Caiman built 84 of them. The world is fully realized — and fully intact, despite everything civilization threw at it. +Four civilizations. 672 lairs, all intact. The Dhole holds 21 territories at population 363. The Rat Snake holds 220 territories. The world is fully realized -- four distinct civilization strategies, eleven named predator species, 31,136 encounters, and every balance target met. + +The 4AI scenario at iter 7o is the largest, most complex, and most balanced run in this project's history.