How it's measured
Positions come from engine-checked Chess960 play. Each is reduced to a few measured facets — material sacrificed on the forcing line, mate distance, whether the key move is quiet, whether it forks. A ranking picks one facet as its named metric and orders its positions by it. The metric sits at the top of every list, so you always know what "best" means.
- Each list states its own metric in plain words.
- Positions keep their measured facets as tags — nothing hidden behind a single score.
- Diagrams render from the position itself, so non-standard Chess960 start squares show correctly.
Frequently asked
What is Chess960 (Fischer Random)?
Chess960, also called Fischer Random, is a chess variant introduced by Bobby Fischer in 1996. The pawns and the rules of movement are unchanged, but the back-rank pieces are shuffled into one of 960 starting positions — the same for both sides, with the bishops on opposite colours and the king between its rooks. Opening theory does not apply, so play is a test of understanding from the first move.
How are the positions measured?
Every position comes from engine-checked Chess960 play and is reduced to a few measured facets — material sacrificed on the forcing line, distance to mate, whether the key move is quiet, whether it forks, and how long the forced line runs. Those measurements, not opinion, decide the order of each list.
How is "best" decided in each ranking?
Each ranking names one facet as its metric and orders its positions by it — most material sacrificed, quickest forced mate, longest forced line, and so on. The metric sits at the top of every list, so "best" always means one stated, checkable thing.
Are the puzzles from real games or composed?
They are positions from Chess960 play, verified by a chess engine — found and selected, not hand-composed. The solution lines are engine-accurate.
What does "mate in N" mean?
"Mate in N" means a forced checkmate exists within N moves against any defence — the defender can delay it but cannot escape it.
Why do some rankings show a number and others don't?
Some lists rank by a number shown on the card — material sacrificed, mate distance, forced-line length. Others rank by a quality with no single number, like the cleanest discovered attacks or the most devastating forks, where the order is the curation. Either way, the measured facets behind every position stay open.
Can I use the data?
Yes. The whole corpus is published as an open dataset, in JSON and CSV, under CC BY 4.0 — free to use with attribution.
Where can I learn the tactics themselves?
The glossary defines every tactic the rankings are built on — discovered attacks, forks, pins, quiet moves and more — with how each one works and how Chess960 changes it.