Skip to main content

Fastest Python draughts / checkers library (import: draughts). Bitboard move generation ~200x faster than pydraughts. PDN/FEN, alpha-beta engine, HUB protocol, web UI. 8 variants: International, American, Frisian, Russian, Brazilian, Antidraughts, Breakthrough, Frysk!

Project description

py-draughts — fastest Python draughts & checkers library

GitHub Actions PyPI version Downloads Docs Python License

py-draughts (import name: draughts) is a fast, modern Python library for draughts (also known as checkers). Bitboard-backed move generation, PDN/FEN parsing, a built-in alpha-beta engine, HUB protocol bridge for external engines like Scan and Kingsrow, an interactive web UI, and tensor exports for RL / ML — all in one package.

[!IMPORTANT] The fastest pure-Python draughts library: ~200x faster legal-move generation than pydraughts, with 8 supported variants and 260+ tests.

py-draughts vs pydraughts

py-draughts pydraughts
Legal moves generation 21.4 µs 5.18 ms (243x slower)
Make move 1.2 µs 552.50 µs (460x slower)
Board init 3.3 µs 579.45 µs (176x slower)
FEN parse 27.4 µs 295.10 µs (11x slower)
Variants 8 (Standard, American, Frisian, Russian, Brazilian, Antidraughts, Breakthrough, Frysk!) 6
Built-in AI engine ✅ Alpha-beta + transposition tables ❌ External only
Engine benchmarking suite
Web UI ✅ FastAPI + interactive board
SVG rendering
ML/RL helpers (tensors, masks)
Test suite 260+ tests, real Lidraughts PDN replays Limited
HUB protocol (Scan, Kingsrow)
Implementation Bitboards (NumPy uint64) Object lists

Full comparison →

Features: 8 variants • Built-in AI engine • External engines via HUB protocol (Scan, Kingsrow) • RL/ML ready (tensors, masks) • SVG rendering • Web UI

Installation

pip install py-draughts

Core

>>> import draughts
>>> board = draughts.Board()

>>> board.legal_moves
[Move: 31->27, Move: 31->26, Move: 32->28, ...]

>>> board.push_uci("31-27")
>>> board.push_uci("18-22")
>>> board.push_uci("27x18")
>>> board.push_uci("12x23")

>>> board
 . b . b . b . b . b
 b . b . b . b . b .
 . b . b . . . b . b
 . . . . b . b . b .
 . . . b . . . . . .
 . . . . . . . . . .
 . w . w . w . w . w
 w . w . w . w . w .
 . w . w . w . w . w
 w . w . w . w . w .

>>> board.pop()  # Unmake the last move
Move: 12->23

>>> board.turn
Color.WHITE

Make and unmake moves

>>> board.push_uci("32-28")  # Make a move
>>> board.pop()              # Unmake the last move
Move: 32->28

Show ASCII board

>>> board = draughts.Board()
>>> print(board)
 . b . b . b . b . b     .  1 .  2 .  3 .  4 .  5
 b . b . b . b . b .      6 .  7 .  8 .  9 . 10 .
 . b . b . b . b . b     . 11 . 12 . 13 . 14 . 15
 b . b . b . b . b .     16 . 17 . 18 . 19 . 20 .
 . . . . . . . . . .     . 21 . 22 . 23 . 24 . 25
 . . . . . . . . . .     26 . 27 . 28 . 29 . 30 .
 . w . w . w . w . w     . 31 . 32 . 33 . 34 . 35
 w . w . w . w . w .     36 . 37 . 38 . 39 . 40 .
 . w . w . w . w . w     . 41 . 42 . 43 . 44 . 45
 w . w . w . w . w .     46 . 47 . 48 . 49 . 50 .

Detects draws and game end

>>> board.is_draw
False
>>> board.is_threefold_repetition
False
>>> board.game_over
False
>>> board.result
'-'

FEN parsing and writing

>>> board.fen
'[FEN "W:W:W31,32,33,...:B1,2,3,..."]'

>>> board = draughts.Board.from_fen("W:WK10,K20:BK35,K45")
>>> board
 . . . . . . . B . .
 . . . . . . . . . .
 . . . . . . . . . .
 . . . . . . . . . .
 . . . . . . . . . .
 . . . . . . . . . .
 . . . . W . . . . .
 . . . . . . . . . .
 . . . . W . . . . .
 B . . . . . . . . .

PDN parsing and writing

>>> board = draughts.Board()
>>> board.push_uci("32-28")
>>> board.push_uci("18-23")
>>> board.pdn
'[GameType "20"]
[Variant "Standard (international) checkers"]
[Result "-"]
1. 32-28 18-23'

>>> board = draughts.Board.from_pdn('[GameType "20"]\n1. 32-28 19-23 2. 28x19 14x23')

Variants

Variant Class Board Flying Kings Max Capture Notes
Standard StandardBoard 10×10 Yes Required International / FMJD rules
American AmericanBoard 8×8 No Not required Men capture forward only
Frisian FrisianBoard 10×10 Yes Required (by value) Diagonal + orthogonal captures
Russian RussianBoard 8×8 Yes Not required Mid-capture promotion
Brazilian BrazilianBoard 8×8 Yes Required International rules on 8×8
Antidraughts AntidraughtsBoard 10×10 Yes Required Lose all pieces (or get blocked) to win
Breakthrough BreakthroughBoard 10×10 Yes Required First player to make a king wins
Frysk! FryskBoard 10×10 Yes Required (by value) Frisian rules with 5 men per side
>>> from draughts import (
...     StandardBoard,
...     AmericanBoard,
...     FrisianBoard,
...     RussianBoard,
...     BrazilianBoard,
...     AntidraughtsBoard,
...     BreakthroughBoard,
...     FryskBoard,
... )

>>> board = AmericanBoard()
>>> board
 . b . b . b . b
 b . b . b . b .
 . b . b . b . b
 . . . . . . . .
 . . . . . . . .
 w . w . w . w .
 . w . w . w . w
 w . w . w . w .

SVG Rendering

>>> import draughts
>>> board = draughts.Board()
>>> draughts.svg.board(board, size=400)  # Returns SVG string
Standard draughts board
>>> board = draughts.Board.from_fen("W:WK10,K20:BK35,K45")
>>> draughts.svg.board(board, size=400)
Board with kings

Engine

Built-in alpha-beta engine with transposition tables and iterative deepening:

>>> from draughts import Board, AlphaBetaEngine

>>> board = Board()
>>> engine = AlphaBetaEngine(depth_limit=5)

>>> move = engine.get_best_move(board)
>>> move
Move: 32->28

>>> move, score = engine.get_best_move(board, with_evaluation=True)
>>> score
0.15

External Engines (Hub Protocol)

Use external engines like Scan via the Hub protocol:

>>> from draughts import Board, HubEngine

>>> with HubEngine("path/to/scan.exe", time_limit=1.0) as engine:
...     board = Board()
...     move, score = engine.get_best_move(board, with_evaluation=True)
...     print(f"Best: {move}, Score: {score}")
Best: 32-28, Score: 0.15

Compatible engines:

  • Scan - World champion level, supports 10x10
  • Kingsrow - Multiple variants, endgame databases
  • Any engine implementing the Hub protocol

Engine Benchmarking

Compare engines against each other with comprehensive statistics:

>>> from draughts import Benchmark, AlphaBetaEngine

>>> stats = Benchmark(
...     AlphaBetaEngine(depth_limit=4),
...     AlphaBetaEngine(depth_limit=6),
...     games=20
... ).run()

>>> print(stats)
============================================================
  BENCHMARK: AlphaBetaEngine (d=4) vs AlphaBetaEngine (d=6)
============================================================
  RESULTS: 2-12-6 (W-L-D)
  AlphaBetaEngine (d=4) win rate: 25.0%
  Elo difference: -191
  ...

Writing Your Own AI

Build custom agents with neural networks, MCTS, or any algorithm:

>>> from draughts import Board, BaseAgent, AgentEngine, Benchmark

>>> class GreedyAgent(BaseAgent):
...     def select_move(self, board):
...         return max(board.legal_moves, key=lambda m: len(m.captured_list))

>>> board = Board()
>>> agent = GreedyAgent()
>>> move = agent.select_move(board)

# Use with Benchmark
>>> stats = Benchmark(agent.as_engine(), AlphaBetaEngine(depth_limit=4), games=10).run()

ML-ready features:

>>> board.to_tensor()        # (4, 50) tensor for neural networks
>>> board.legal_moves_mask() # Boolean mask for policy outputs
>>> board.features()         # Material, mobility, game phase
>>> clone = board.copy()     # Fast cloning for tree search

Server

Interactive web interface for playing and engine testing:

from draughts import Board, Server, AlphaBetaEngine, HubEngine

server = Server(
    board=Board(),
    white_engine=AlphaBetaEngine(depth_limit=6),
    black_engine=HubEngine("path/to/scan.exe", time_limit=1.0)
)
server.run() # Open http://localhost:8000
image

Performance

Legal moves generation in ~10-30 microseconds:

Operation py-draughts pydraughts Speedup
Board init 3.30 µs 579.45 µs 176x faster
FEN parse 27.40 µs 295.10 µs 11x faster
Legal moves 21.35 µs 5.18 ms 243x faster
Make move 1.20 µs 552.50 µs 460x faster
Speed Comparison Chart

Engine search at various depths:

Depth Time Nodes
5 274 ms 3,263
6 619 ms 7,330
7 2.20 s 21,642
8 6.55 s 98,987
Engine Benchmark

Testing

Comprehensive test suite with 260+ tests covering all variants and edge cases:

pytest test/ -v

Tests include:

  • Move generation - Push/pop roundtrips, legal move validation
  • Real game replays - PDN games from Lidraughts for all variants
  • Edge cases - Complex king captures, promotion mid-capture, draw rules
  • Engine correctness - Hash stability, transposition tables, board immutability
  • All 8 variants - Standard, American, Frisian, Russian, Brazilian, Antidraughts, Breakthrough, Frysk!

Contributing

Contributions welcome! Please open an issue or submit a pull request.

License

py-draughts is licensed under the GPL 3.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

py_draughts-1.8.1.tar.gz (153.7 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

py_draughts-1.8.1-py3-none-any.whl (166.0 kB view details)

Uploaded Python 3

File details

Details for the file py_draughts-1.8.1.tar.gz.

File metadata

  • Download URL: py_draughts-1.8.1.tar.gz
  • Upload date:
  • Size: 153.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for py_draughts-1.8.1.tar.gz
Algorithm Hash digest
SHA256 00ffacede6b80deca73fe9a104bcf78c0266cf4ddb7c574283c61c216027a21b
MD5 763085aea0184b6048d0c7af6057fa53
BLAKE2b-256 c57d87e1b8bc7c106b8f8bd8169bc9e559aae0403fd3b02af1a34c5299f107e0

See more details on using hashes here.

Provenance

The following attestation bundles were made for py_draughts-1.8.1.tar.gz:

Publisher: python-publish.yml on miskibin/py-draughts

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file py_draughts-1.8.1-py3-none-any.whl.

File metadata

  • Download URL: py_draughts-1.8.1-py3-none-any.whl
  • Upload date:
  • Size: 166.0 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for py_draughts-1.8.1-py3-none-any.whl
Algorithm Hash digest
SHA256 12865ebde5950cdd327bd18720c0c3cf189e3cca3a6876c429e31e2564b9108e
MD5 e04909d7838cb4998b93d7fe9d2704db
BLAKE2b-256 72013f322c29b88e97c83b15ff0f3a865d6ce6b6f462cd5e5a97e91273d1f48a

See more details on using hashes here.

Provenance

The following attestation bundles were made for py_draughts-1.8.1-py3-none-any.whl:

Publisher: python-publish.yml on miskibin/py-draughts

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page