Transparent calculations
Data & Methodology
Every result is tied to a named source snapshot, generated checksum, and disclosed calculation boundary.
Last updated:
Current data snapshot
- Pinned commit
c59712e24b839a0bedef16b06a1a0117e8741fe3- Database label
- v27
- Coverage
- 299 Pals and 44,851 combination rows
Import integrity boundary
The SHA-256 checksums confirm that the generated files match the pinned PalCalc import. They confirm import integrity, not manual in-game verification of every row. Product overrides are disclosed below and are not silently mixed into the source fields.
Breeding and path calculations
Parent order is normalized. Named special combinations take precedence over ordinary mappings. Path results use breadth-first search over the same combination graph and optimize for the fewest breeding generations, not for Cake cost, gender-roll time, or passive-skill probability.
Passive-skill estimates
The predictor enumerates the unique inherited parent pool, applies the published inherited-count weights, then models eligible random additions without replacement under the four-slot cap. It is a planning estimate, not a guarantee for any individual egg.
Mutation estimates
Mutation Eggs are kept outside the ordinary combination graph. The 1%, estimated 2%, and 3% Cake rates shown in the planner are labeled fan-database estimates reviewed from paldb.cc's v1.0.2 Mutation notes. Palspawn can calculate repeated-attempt probability from those inputs, but does not claim they are official guarantees. Mutation species prediction remains unavailable because the audited PalCalc snapshot does not include an independently versioned species formula, exclusion list, and golden fixtures.
Overrides and corrections
Product-required overrides are stored in the generated manifest so they can be audited separately from the imported source. Dataset updates must rerun validation and checksum tests before being published.