ChessMachina

How much thinking does it take to beat you? ChessMachina answers this in two ways: with reconstructed historical chess machines and with its own modern local engine whose compute budget you control yourself.

Play against nine reconstructed stations of computer chess history, from 1912 to 1997. The machines are algorithmically rebuilt from available historical documentation, not with original code and not by emulating the original hardware.

Compute Mode uses its own modern local reference engine. You control how much computation the engine gets per move through reproducible work or node budgets. ChessMachina does not display CPU percentages, energy consumption, or an exact Elo rating.

Depending on the machine, ChessMachina shows how a decision was reached: rules, candidate moves, minimax, alpha-beta, search depth, singular extensions, or historical search extensions. Not every machine has the same metrics, since historical and modern algorithms work differently.

Games you play land in a local game history together with the opponent and engine version. A saved game can be reconstructed as a replay from its starting position and move sequence and watched again.

ChessMachina uses clean-room reconstructions based on available historical documentation. Depending on the machine, algorithm fidelity, parameter fidelity, hardware fidelity, data asset fidelity, and behavioral fidelity differ. Missing original parameters or historical data are never presented as original.