A brief history of computer chess

Since the 18th century, people have been fascinated by the idea of machines that could play chess against humans. With the advent of the digital electronic computer in the mid-20th century, that dream finally became feasible. What followed was six decades of intense development in the field of computer chess, from research projects to commercial products. Over that period, computers grew from playing only a limited subset of chess to beating the World Chess Champion in a six-game match. Today, a worldwide computer network (the Internet) facilitates play between humans all over the world.

Clearly, computer chess has come a long way. Let’s take a compressed look at some key moments in its evolution over the years.

In 1770, Hungarian inventor Wolfgang von Kempelen unveiled an invention that astounded the world: a machine that could seemingly play (and beat) a human at chess without any assistance through the hands of a robot that resembled a Turkish gentleman. If that sounds too good to be true for 1770, it was: The Turk turned out to be an elaborate hoax, with mechanical parts just for show. For the Turk to function, a diminutive person had to hide inside its case and guide the Turk’s arm. Still, it ignited an interest to create a real chess-playing machine that would not be fulfilled for nearly two centuries.

In the mid-1940s, British mathematician Alan Turing (pictured) began theorizing ways that a computer could play chess against a human. Across the Atlantic, in 1949, Bell Labs researcher Claude Shannon published a seminal paper describing a potential program to do exactly that. The following year, Turing created the first computer chess playing algorithm, and lacking a suitable machine to run it on, Turing himself took the role of the computer, manually calculating moves via his algorithm (which took roughly 30 minutes per move) in a game against his friend. His program lost, but history was made, and Turing published his ideas in a 1953 paper as part of the book Faster than Thought.

In 1951, German inventor Dietrich Prinz created the first chess program of any type to run on an actual electronic computer. He did so on the Ferranti Mark 1 (the first commercially available general purpose electronic computer) at the University of Manchester in England. The Mark 1 lacked the power to play a full game of chess, so Prinz created a limited program that could find the best move in a chess game only if the move was two moves away from checkmate.

(Photo: Computer History Museum)

IBM researcher Alex Bernstein (seen here standing) created the world’s first complete chess-playing computer program to run on an actual computer in 1957. The program ran on an IBM 704 mainframe and was developed in conjunction with MIT. It was distinguished from Prinz’s 1951 effort because it could play a full game of chess against a human from start to finish, taking roughly eight minutes per move.

(Photo: Computer History Museum)

Up until the late 1960s, computer chess programs displayed their moves in either written chess notation (i.e. “e4 e5”) or through a visual diagram of a chess board printed on paper. The earliest known electronic graphical user interface (GUI) for chess was created in 1968 for a DEC 340 display attached to a PDP-6 mainframe (seen here lower-right). t displayed moves from the Greenblatt Chess Program, which also happened to be the first chess program to play humans in a tournament.

In 1970, NASA researcher Chris Daly created a rich graphical interface for his chess program Daly CP using an IDIIOM CAD system (upper left). With it, one could use a light pen to select moves.

Progress in computer chess on mainframes continued to advance steadily in the 1970s. During the middle of that decade, a new, smaller form of machine emerged: the personal computer. At the heart of nearly every personal computer lay a microprocessor, hence the term “microcomputer.” In 1976, Peter Jennings (not the ABC news anchor) created the first commercial chess program for microcomputers, MicroChess. It ran on the KIM-1, a single-board machine powered by a 6502 CPU. It was the first of many to come.

(Photos: Peter Jennings)

With the price of microprocessors rapidly falling in the 1970s, it was only a matter of time before some company delivered a dedicated electronic chess computer product. The first company to do so was Fidelity Electronics of Chicago in 1977 with their Chess Challenger. To play verses the primitive machine (which used a 2MHz 8080 CPU), players had to enter moves in notation on the unit’s built-in keypad. The computer, with only one difficulty setting, would display its move on an segmented LED display. Over the next few decades, many chess computers of increasing capability would follow.

(Photo: My Chess Computers)

In 1978, a new chess program called Sargon, programmed by Dan and Kathe Spracklen, won the world’s first computer chess tournament for microcomputers. During those early days of personal computers, few users had access to standardized removable data storage, so Hayden Books published the code for Sargon in book form so users could type it in themselves. A few years later, the game received a popular port to the Apple II (right), which spawned a successful series of Sargon products throughout the 1980s.

Until 1980, users of chess computers had to move the computer’s pieces by hand. The introduction of the Boris Handroid, an extremely rare unit with a robotic arm, changed that. More notable was the 1982 Novag Robot Adversary (left), which used a robotic arm to pick up and move the computer’s pieces automatically. Only 2500 units were produced, however, with high failure rates limiting sales.

A more successful robotic design was the Milton Bradley Grandmaster (1983, right), which used an electromagnet on a movable arm under the chess board to move the computer’s pieces.

(Photos: Novag, Computer History Museum)

The best selling chess program franchise in history began in 1986 with The Chessmaster 2000, originally released by The Software Toolworks for several home computer platforms including Commodore 64, Amiga, and MS-DOS. Chessmaster used an engine created by Dave Kittinger that was renowned as being particularly strong for a microcomputer chess engine at the time. Chessmaster, as a brand, dominated sales of chess video and computer games for the next two decades.

Personal computer chess programs took an entertaining leap forward in 1988 with Interplay’s Battle Chess (first for Amiga, later ported to many other platforms), which enhanced a typical chess game with amusing richly-illustrated animations. While the ruleset was the same as standard chess, players could watch as pawns stabbed with spears, knights severed limbs, and rooks (which turned into golems) devoured their enemies upon every capture, making chess fun even for those who would not normally play the game.

In 1992, the Internet Chess Club launched the Internet Chess Server (ICS), the world’s first means for people to come together on the Internet and play chess. Initially, members of the Server logged into the service for free via the ancient but reliable text-based Telnet protocol. Games were displayed as notation or depicted as an ASCII-drawing of a chess board; later, users developed their own graphical front-end clients to facilitate play.

After the ICS began charging for memberships in 1995, many members flocked to free alternatives like the Free Internet Chess Server (FICS), which remains popular to this day.

The influence of the Internet on the chess world continued to be felt throughout the 1990s. In 1996, a group of chess fans of varying skill (aka “The World”) challenged several-time World Chess Champion Anatoly Karpov to a game. Through the Internet, the group cooperated and voted on the best moves to make against the grandmaster. Ultimately, their combined knowledge of chess wasn’t enough, and Karpov won the game in 33 moves.

Many “The World” games have been played since then. In 2007, The World scored its first victory vs. grandmaster Arno Nickel in 41 moves.

(Photos: Stefan64, NASA)

Since the beginning of computer chess, programmers had dreamed of creating a program that could defeat even the strongest human players in the game. That goal proved elusive in a real tournament setting until 1996 when IBM’s Deep Blue computer defeated World Chess Champion Garry Kasparov in a single game. Kasparov won the multi-game match, however, and the two met again in 1997 for a rematch. In a stunning upset that excited the world news media, Deep Blue won that six-game match with three victories and three mutual draws. It was the computer chess equivalent of landing on the moon.

(Photo: Computer History Museum)

After the victory of Deep Blue over Kasparov, chess engine programmers hunkered down with the goal of producing the world’s strongest engine—typically for play verses other chess engines.

In 2005, a strong leader emerged from Czech-American programmer Vasik Rajlich in the form of Rybka, which soon after began to win repeated computer chess tournaments, including the World Computer Chess Championship (WCCC) from 2007 to 2010. In 2011, accusations of plagiarism began to emerge against Rybka, prompting the governing body of the WCCC to disqualify Rybka from past and future events. Despite that, Rybka remains one of the world’s strongest chess engines.

(Photos: Vasik Rajlich)

Throughout 1990s and 2000s, consumer chess programs continued to gain in playing strength while adding teaching features and enhanced graphics. One such typical program from the mid-2000s was Chessmaster XI, which at present is still the most recent product in the Chessmaster line. Another popular consumer chess engine is the Fritz series, which is based upon the world-class chess engine of the same name.

Since that first Internet Chess Server in 1992, the Internet has grown from a curiosity to a ubiquitous fact of life, providing connections to an endless stream of living chess opponents any time of the day. That fact has diminished the popularity of standalone computer chess products like Chessmaster among casual players; many prefer to play online against other humans on websites such as Chess.com or dozens of others.

So it’s slightly ironic that computer technology has now brought us full circle—beyond the time when men dreamt of programming human-vanquishing AI, and back to a time when two real people played chess together. They just might be 2000 miles apart.