The result child is the breedable pal whose CombiRank is closest to this average.
Tiebreaker (equidistant ranks)
When two breedable pals have CombiRanks equally distant from the computed average, the tie is broken by parent rarity average.
Rule: Pick the child whose rarity is closest to the average of the two parents' rarities. If still tied, pick the child with lower rarity.
Rationale: Verified against 3 in-game conflicting cases:
Parent A
Parent B
Avg Rank
Contender 1 (rank/rarity)
Contender 2 (rank/rarity)
In-game result
Tiebreaker
Braloha (r8)
Dynamoff (r6)
1215
Quivern (1210/r7)
Azurobe Cryst (1220/r8)
Quivern
Rarity closer to parent avg (7 vs 8, parent avg=7)
Shaolong (r9)
Helzephyr Lux (r8)
500
Astegon (490/r9)
Dualith (510/r8)
Dualith
Rarity closer to parent avg (tie at 0.5), lower rarity wins (8 < 9)
Braloha (r8)
Jetragon (r20)
550
Univolt Cryst (540/r6)
Silvegis (560/r8)
Silvegis
Rarity closer to parent avg (6 vs 0, parent avg=14)
Key insight: The child's rarity is matched against the parents' average rarity, not just picked arbitrarily. Legendary parents (high rarity) bias toward higher-rarity children.
Verified Combinations
Parent A
Parent B
Result
Source
Shaolong
Wumpo Botan
Eidrolon
In-game test
Shaolong
Orserk
Aegidron
In-game test
Shaolong
Helzephyr Lux
Dualith
In-game test
Shaolong
Petallia
Ghangler
In-game test
Braloha
Dynamoff
Quivern
In-game test
Braloha
Jetragon
Silvegis
In-game test
Braloha
Gobfin Ignis
Tarantriss
In-game test
Braloha
Beegarde
Beakon
Calculated (exact match)
Chikipi
Eye of Cthulhu
Chikipi
In-game test
Eye of Cthulhu
Braloha
Penking
Calculated
Data Flow
update_game_data.py reads DT_PalMonsterParameter.json from Exports
Extracts CombiRank, Tribe, IgnoreCombi, Rarity for each pal
Filters out unnamed pals and non-breedable (CombiRank ≤ 0)
Builds rank_to_best map from breedable pals (non-IgnoreCombi)
Computes formula results:
breedable × breedable: All pairs → child_to_parents_formula
IgnoreCombi × all: For IgnoreCombi parents with every partner → parent_to_children_formula and child_to_parents_ignore
Unique combos: From DT_PalCombiUnique.json → unique_combos, child_to_parents_unique
Saves to resources/game_data/breedingdata.json
Changelog
Date
Change
Reason
Initial
Weighted formula (max*7+min)//9 from betatest script
Copied from external test tool
Fix 1
Simple average (r1+r2)//2
Weighted formula gave wrong results (Shaolong+WumpoBotan→Anubis instead of Eidrolon)
Fix 2
parent_to_children_formula added for IgnoreCombi parents
Orserk, Shaolong etc. showed no children
Fix 3
parent_to_children_formula uses ALL pals as partners
IgnoreCombi×IgnoreCombi pairs missing
Fix 4
child_to_parents_ignore reverse lookup added
Parents mode didn't show IgnoreCombi parents
Fix 5
Rarity-based tiebreaker
Equidistant ranks gave wrong results; rarity matching parents' average fixed all 3 test cases
Fix 6
Removed unconditional best = p bug
Extra line outside if block overrode rarity check
File Locations
Formula logic:scripts/scrs/update_game_data.py → function update_breeding_data()
Run npx skillmds@latest add deafdudecomputers/pst-breeding in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.
Breeding Formula — Development Notes It is listed under Coding & Dev Tools on SkillMD.
This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.
This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.
Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.
deafdudecomputers (@deafdudecomputers) published this skill. Their other Agent Skills are listed on their SkillMD profile.
deafdudecomputers