Do xiangqi and chess engine evaluations use the same scale, or does a plus-one score mean something different in each game?
These answers restate the rules implemented by our open-source engine, which follows the WXF laws and is regression-tested against the official 110-diagram repetition casebook. Where an answer refers to a specific position, that position lives on the linked reference page, where the build verifies it against the engine before publishing.
Direct answer
Not directly comparable, even though both look similar on the surface. A chess engine's centipawn score and a xiangqi engine's evaluation are both ultimately anchored to each game's own piece values and typical position dynamics, but those are different in each game. A xiangqi soldier is not worth the same fraction of a xiangqi chariot that a chess pawn is worth of a chess rook, so the same raw number from two different engines does not represent the same real advantage on the board.
Why the scales look similar but aren't
Both engines report a numeric advantage built from the same basic idea, material plus positional factors, converted into a single score. That structural similarity makes the numbers feel directly comparable, but each engine's scale is calibrated against its own game's piece values and typical game length, which simply are not the same between the two games.
What actually transfers between the two
The general shape of the number is meaningful within its own game: a large score reliably means a large advantage, a small one means the position is close, and that much intuition carries over fine. Treating a xiangqi +2.0 as directly equivalent to a chess +2.0 in magnitude of advantage, though, is not something the numbers actually support.
Related questions
Is there a standard xiangqi evaluation scale the way centipawns are standard in chess?
Not as tightly standardised across every engine, though point-based scales anchored to xiangqi's own piece values are common.
Does this affect how mate scores are shown?
Mate distances are generally reported similarly in both, as a move count to forced mate, which does translate directly between the two games.
.