Chess History, Rules, and Computers — Chaturanga to AlphaZero
Chess history from India's chaturanga through Persia, the Islamic world, and Europe, plus the basic rules and computer chess from Deep Blue to AlphaZero.
Chess is the world's best-known two-player strategy board game. Each side places 16 pieces on an 8×8 board, and you win by trapping the opponent's king in a position where it cannot escape capture. The rules can be explained in a few pages, but the calculation that unfolds on the board has captivated people for nearly 1,500 years.
In this post, we look in turn at where chess came from, when its modern rules were finalized, and how computers came to surpass humans. We also cover how janggi, the Korean chess familiar to many Koreans, and chess are actually cousins that branched off from the same ancestor.
The Ancestor of Chess: India's Chaturanga
The oldest known ancestor of chess is India's chaturanga, first attested in records from India around the 7th century. In Sanskrit it means "four parts," referring to the infantry, cavalry, elephant corps, and chariot corps that made up the ancient Indian army. These four branches are thought to have become today's pawn, knight, bishop, and rook, respectively.
Chaturanga was played on an uncolored 8×8 board called the ashtapada. The fact that today's chessboard is 8×8 is a tradition handed down from it.
Among chess historians, the mainstream view is that chaturanga is the common ancestor of the following games:
| Game | Region |
|---|---|
| Chess | Europe → worldwide |
| Xiangqi (象棋) | China |
| Janggi | Korea |
| Shogi (将棋) | Japan |
| Makruk | Thailand |
| Sittuyin | Myanmar |
In other words, janggi, which Koreans play during holidays, and Western chess are games that branched east and west from the same root. Common traces remain in both games, to the extent that researchers have studied the relationship between how janggi's elephant (象) moves and the elephant piece in chaturanga.
Persia's Shatranj and the Origin of "Checkmate"
Chaturanga traveled from India to Persia. Around 600 AD, in Sasanian Persia, the game was called chatrang and considered part of a noble's education. Later, after the Arab conquest of Persia, sounds not found in Arabic were changed, and the name became shatranj.
Words we still use today were born in this period. When attacking the opponent's king, players would call out "Shāh!" (King!) in Persian, and when the king had nowhere to escape, "Shāh māt!" This is the origin of the English words check and checkmate.
Shatranj spread throughout the Islamic world and entered Europe via North Africa, the Iberian Peninsula (al-Andalus), and Sicily. The Spanish "ajedrez" and Portuguese "xadrez" come from shatranj, while the Italian "scacchi" and French "échecs" come from the Persian "shāh" (king).
Modern Rules Take Shape in Europe
After chess arrived in medieval Europe, it became so popular among the upper classes that it was counted among the accomplishments a knight should master. However, chess back then was much slower than today. The piece in today's queen position was originally the "vizier," a weak piece that moved only one square diagonally. The bishop, too, could only jump two squares diagonally.
The big change came in the late 15th century in the Valencia region of Spain.
- Around 1475, the Catalan poem Scachs d'amor, written in Valencia, contains the first record of the queen moving long distances as it does today.
- As the modern moves of the queen and bishop spread between 1475 and 1500, chess came close to its present form by around 1500.
- In 1497, Spanish cleric Luis Ramírez de Lucena published a chess book in Salamanca, the oldest surviving printed chess book.
As the queen grew stronger, the game sped up, and at the time this was even called "the chess of the mad queen." The rule allowing a pawn to advance two squares on its first move, the resulting en passant, and castling, which developed from the custom of moving the king to safety, also became established around this time.
Basic Rules of Chess at a Glance
The rules of modern chess are governed by the International Chess Federation (FIDE), founded in Paris in 1924. Here are the essentials:
| Piece | Movement |
|---|---|
| King | One square in any direction |
| Queen | Any number of squares horizontally, vertically, or diagonally |
| Rook | Any number of squares horizontally or vertically |
| Bishop | Any number of squares diagonally |
| Knight | Moves in an L shape and can jump over other pieces |
| Pawn | One square forward (two on its first move); captures diagonally forward |
- White moves first. You cannot skip a turn.
- Checkmate: If your king is under attack and has no way to escape, you lose.
- Stalemate: If it's your turn and you're not in check but have no legal moves, the game is a draw.
- Special rules: Castling (moving the king and rook together in one move), en passant (capturing a pawn that advanced two squares as if it had moved only one), and promotion (turning a pawn that reaches the last rank into another piece).
- Draw rules: A draw can be claimed if the same position occurs three times (threefold repetition), or if both sides make 50 moves without moving a pawn or capturing a piece (the fifty-move rule).
Compared with janggi, the differences are clear. In janggi, the general moves only within its palace and the cannon jumps over pieces, whereas in chess the king can roam the entire board and a pawn that reaches the end can become a queen. This shows how games from the same ancestor grew into completely different games according to each region's tastes.
Modern Chess and the Era of World Champions
From the late 18th century to the 1880s, "Romantic chess," which prized fast, flashy attacks, was in fashion. America's Paul Morphy defeated the leading players of his day during a short career from 1857 to 1863, and Prague-born Wilhelm Steinitz codified the principles of "positional play": avoid creating weaknesses and exploit your opponent's.
In 1886, the first official World Chess Championship was held. In the 20th century, the title passed through Cuba's Capablanca, Alekhine, and others, and after Botvinnik in 1948, Soviet players held the top spot for a long time. In 1972, America's Bobby Fischer became champion, leaving one of the iconic scenes of the Cold War era. After Karpov, Kasparov, Kramnik, and Anand, Norway's Magnus Carlsen became champion in 2013.
Computer Chess: From Shannon's Paper to Deep Blue
Chess has long been the testing ground for whether machines can think. In 1950, American mathematician Claude Shannon published a paper on chess search, proposing two strategies: expanding all possible moves into a tree structure and calculating them, and reading only selected promising moves. This was before any chess-playing program existed.
The first chess programs of the 1950s vacuum-tube computer era were so weak they could lose even to beginners. But about 40 years later, things were completely different.
| Year | Event |
|---|---|
| 1996 | IBM's Deep Blue won one game against Kasparov; Kasparov won the match 4–2 |
| 1997 | Deep Blue won the rematch 3½–2½ |
Game 1 of the first match, held in Philadelphia in 1996, was the first game in which a computer defeated a reigning world champion under standard tournament conditions. In the 1997 rematch in New York, Deep Blue won the entire match, recorded as the first time a reigning world champion lost to a computer under tournament conditions.
AlphaZero and Today's Chess Engines
After Deep Blue, chess engines reached beyond-grandmaster strength even on ordinary PCs. The most prominent open-source engine is Stockfish.
In December 2017, Google DeepMind unveiled AlphaZero. AlphaZero learned solely by playing against itself, without opening books or endgame databases. According to DeepMind, after about nine hours of training, it played 100 games against Stockfish 8, recording 28 wins, 72 draws, and 0 losses. However, there was also criticism that Stockfish's settings at the time put it at a disadvantage, and results from games under revised conditions were later published in the journal Science in 2018.
AlphaZero's influence was significant. The neural-network-based open-source engine Leela Chess Zero emerged, and Stockfish also adopted neural network evaluation. Today, even engines running on smartphones can consistently beat the world's best players, so professional players use engine analysis as a standard part of their preparation. Online chess, which began in the mid-1990s, also became hugely popular in the 21st century.
If You're Learning Chess for the First Time
If you're new to chess, we recommend this order:
- Start by learning how the pieces move and the concept of checkmate.
- Practice the basic endgames of driving the opponent's king into checkmate with king and queen, and king and rook.
- In the opening, following just three principles is enough: control the center, develop your knights and bishops first, and protect your king by castling.
- Review your games with an engine to see where things went wrong.
Playing chess and janggi together lets you feel firsthand how differently two games from the same ancestor grew. KEKE STUDIO's Action Janggi brings Korean janggi into 3D, so if you're familiar with chess, it's an easy way to try out chaturanga's eastern descendant.
