From Turing's Paper Program to Deep Blue
In 1948, Alan Turing and David Champernowne designed "Turochamp," one of the first chess-playing algorithms ever conceived, years before any computer existed that could actually run it. Turing formalised the rules in a 1953 paper and, in the summer of 1952, played an actual game using it against colleague Alick Glennie, hand-simulating every calculation himself, move by move, since no machine of the era could execute the program. It took nearly fifty more years of steady progress to reach Deep Blue's 1997 win over Garry Kasparov (see our full account of that match).
A program with no machine to run it
Turochamp is remarkable specifically because of the gap between its design and its execution, Turing and Champernowne worked out real decision-making rules a chess-playing machine could theoretically follow, then had to abandon actually running it on the available 1951 Ferranti Mark 1 computer because the algorithm was too complex for the hardware of the day. Turing played it out by hand instead, a human executing a machine's logic rather than the other way around.
The slow middle decades
Meaningful computer chess progress through the 1950s to 1970s came in increments, early programs on room-sized mainframes could barely see a few moves ahead, and it was not until specialised hardware and better search techniques emerged in the 1980s that computers began reaching genuinely strong club-level play, then master level. Our piece on Deep Thought's 1988 win over Bent Larsen covers the moment a computer first beat a grandmaster under real tournament conditions, the clearest marker that the decades of incremental progress had crossed a genuine threshold.
Deep Blue, and the finish line
Deep Blue, Deep Thought's direct successor, beat Kasparov in a 1997 rematch after losing their first encounter in 1996, closing the gap Turing had only been able to describe on paper in 1948. Less than fifty years separates a chess algorithm too complex for any existing computer to run from a machine beating the strongest player in the world.
What actually changed across that timeline
The core idea, evaluate a position, search ahead through possible moves, pick the best outcome, was already present in Turing's original design. What changed over five decades was almost entirely a matter of scale: how many positions per second a machine could actually search, not any fundamentally new insight into how a computer should think about chess.
Why this history still matters today
Every engine analysis a player runs after a game today is built on the same foundational idea Turing described before any computer could execute it. The distance between "I have a theory for how a machine could play chess" and "this machine plays chess better than any human alive" turned out to be almost entirely a question of raw computing power, not a missing idea.