A history of shogi

Arrival from the continent

Every game in the chess family appears to descend from a single ancestor that took shape in northern India in the sixth century or thereabouts, and then spread west into Persia and east across Asia. Shogi is the Japanese branch of that eastern spread. What nobody can say is which road it came by.

Two routes are usually proposed. One runs through China, where a chess game with its own long history was played on the lines rather than in the squares, and across to Japan by way of Korea. The other runs up through Southeast Asia, where board games with a similar army of short-range men were played, and reaches Japan by sea. Arguments have been made for both, and for a mixture of the two, on the strength of piece names, board shapes and trade contacts. None of them is decisive. Shogi keeps its men in the squares rather than on the intersections, which is a point against a simple descent from the Chinese game, and it has a pawn that captures straight ahead rather than diagonally, which is a point in favour. The honest summary is that the route is unknown and the date of arrival is not fixed within a century or two.

The Japanese record

The first hard evidence is archaeological. Excavations at a temple site in Nara turned up small wooden pieces, five-sided in the modern shape, with piece names written on them in ink, deposited in the eleventh century. They are the oldest shogi men anyone has found. Written references from around the same period mention a game by name without describing how it was played, which is the usual frustration of early game history: the sources tell you the game existed and almost nothing about its rules.

For several centuries afterwards the game seems to have been played in more than one size. There were larger boards with larger armies, some of them very large indeed, with fabulous beasts among the pieces and games that must have run for hours. A middle-sized version had a following among the educated for a long time and has never entirely died out. The smaller game — nine by nine, with the army of twenty that is standard today — coexisted with these for a long period before winning out. Why it won is not documented. The obvious guess, that a shorter game with fewer men is simply a better game to sit down and play, is only a guess.

One casualty of the shift is worth naming. The small game once included a piece called the drunk elephant, which moved like a king but without the straight backward step, and which promoted into a second king — a crown prince, so that capturing one king did not end the game. It is absent from the modern set. Tradition credits its removal to a sixteenth-century emperor. The tradition is late, the documentation is thin, and it is best treated as a story about the change rather than as evidence of who made it.

The drop rule

No other member of the chess family returns captured men to the board. It is the one genuinely Japanese contribution to the family, and its origin is the thinnest part of the record.

What can be said is roughly this. The oldest known Japanese boards, and the larger historical versions, show no sign of it. By the seventeenth century it is unmistakably part of the game, described plainly in the literature that survives from the beginning of the Edo period. Somewhere between those two points, most likely during the sixteenth century, the rule was adopted. That is as precise as the evidence allows, and writers who give a firmer date are going beyond it.

The reason often given — that soldiers in a period of civil war changed sides, and the game reflected it — is a pleasant story with nothing behind it. A duller explanation fits the game better: the small board with only twenty men a side empties out fast, and a game that returns captured men keeps the position full and the fight alive to the last move. Whatever the motive, the effect was to turn a slow game of short-range pieces into a game with almost no drawn results and no safe simplification, which is what it remains.

Schools, titles and the modern game

Under the Tokugawa shogunate, shogi and go were both put on a formal footing. Hereditary houses were established, supported by stipends, and the headship of the shogi houses carried the title of meijin. Players from these families composed and published collections of mating problems — some of them extraordinarily difficult — and played an annual game before the shogunate that gave the modern professional calendar its ancestry. Because the title passed within families rather than being won at the board, the strongest player of an era was not always the titleholder, and the record of that period has to be read with that in mind.

The old arrangement did not survive the end of the shogunate, and the modern professional structure was rebuilt in the twentieth century around a national association of players, which became the Nihon Shogi Renmei, known in English as the Japan Shogi Association. The hereditary meijin was abolished in the 1930s and replaced with a title decided by play. The association today runs the professional ranks, the major title matches, the newspaper-sponsored tournaments that have supported professional shogi for a century, and the published tournament rules that this site follows.

The last thirty years have added computers to the story. Shogi resisted strong engine play for longer than western chess, largely because the drop rule makes the number of legal moves in a middlegame position very large, and because a captured man is never off the table. That gap closed in the 2010s. Engine analysis is now a normal part of professional preparation, exactly as it is elsewhere in the chess family, and one consequence is that openings which had been considered settled for decades are being reopened.

Sources

Everything on this page is written from scratch. These are the works it leans on, and what each one settles.

• Nihon Shogi Renmei (Japan Shogi Association) rules and published record

  The governing body of the professional game. Its rules fix the modern form described in the closing section — the drop restrictions, repetition and impasse — and its own institutional record is the source for the replacement of the hereditary meijin by a title decided at the board.

• H. J. R. Murray, A History of Chess (1913)

  The standard nineteenth-and-early-twentieth-century survey of the whole chess family, and still the reference point for how the eastern branches relate to one another. It is used here for the descent of shogi from a common ancestor, for the larger historical Japanese boards, and above all for the caution that the route of transmission and the date of the drop rule are not settled by the surviving evidence.