Awhile ago chess.com reworked and simplified it's Brilliancy detector to basically look for two criterion:
-
The move gave the player a material disadvantage by either
- Moving a piece that will be captured next turn and not capturing a stronger piece with that move OR
- Not defending a piece that is under attack and will be captured next turn or
- Revealing a piece to capture next turn that was otherwise safe.
-
The played move (that 'hung' a piece) was the 'best' analysis move by a wide enough margin.
The feels pretty good to me, but the currently lichessTools brilliancy analysis doesn't seem to use the same criteria.
for instance:
I don't know why this was marked as a brilliancy
I kinda feel like this one shoulda been marked a brilliancy; but it wasn't even marked good?
Anyways, just some food for thought.
-Xyvir
https://lichess.org/XhoMkNgX#63