Classic Xiangqi Books: The Great Manuals of Ming and Qing

A shelf of puzzles and openings

Surviving xiangqi writing from before 1911 falls into two broad groups. The larger group consists of compositions, called 排局 (páijú), where a position is set up and one side must force a win or a draw. The smaller group covers openings and whole games. Many older works are lost. The books below survive in several editions, and modern Chinese reprints make them easy to find.

Dates and authors for old manuals are sometimes disputed, because books were reprinted, retitled and occasionally sold under a grander date than they deserved. Where sources disagree, we say so.

Shiqing Yaqu, 1570

Shiqing Yaqu (适情雅趣), often rendered in English as the Elegant Pastime Manual, was compiled by Xu Zhi of Jinling, today's Nanjing, and published in 1570 during the Longqing reign. The standard editions hold 550 positions across ten volumes, most of them middlegame and endgame situations in which Red wins, followed by opening material that includes a sequence known as the Eighteen Variations of the Golden Roc (金鹏十八变).

Its value today lies in the practical endgames. Several of its sections group similar positions together, an early form of the systematic endgame study that modern players take for granted.

Invincible since leaving the cave

Zi Chu Dong Lai Wu Di Shou (自出洞来无敌手) has one of the best titles in the literature, roughly: no rival since I came out of the cave. It is signed by a Daoist calling himself Chunyang Daoren, and its exact date is unknown, though it is generally placed in the late Ming. The title is a line of verse recorded by the Song writer Yao Kuan, and the book uses each of its seven characters to head a chapter, with five games per chapter and 35 in all. Each chapter follows an opening family, from opposite-direction cannons (列手炮) to the palace-heart cannon (窝心炮).

The play is lively rather than deep by modern standards, full of quick kills, which makes it good reading for improving players who want attacking ideas in their simplest form.

Secrets in the Tangerine, 1632

Juzhong Mi (橘中秘) is probably the most famous xiangqi book ever written. It was compiled by Zhu Jinzhen (朱晋桢) and is dated to 1632, late in the Ming dynasty. The title comes from a Tang-era tale in which a man from Ba-Qiong split open two enormous tangerines and found, inside each, a pair of white-haired old men calmly playing a board game. After that story, the pleasure inside the tangerine became a literary way of referring to the game.

The book is known above all for the central cannon. Its opening sections show Red attacking with a cannon on the middle file, often against Black replies that also rely on cannons, while the later scrolls collect refined practical endgames. It also contains a short statement of principles, the Quanzhi (全旨), full of advice you could pin above a club table today. One line says that when you give up a piece you must gain the initiative in return. Another says that when your forces are strong you should look for the win, and when they are weak you should head straight for the draw.

The book reached the West early: the British sinologist W. H. Wilkinson based an English treatise on it in 1893. Later, H. J. R. Murray's A History of Chess passed on Wilkinson's observation that a book nearly three centuries old was still the standard Chinese work on openings.

The Plum Blossom Manual and the screen horse

Meihua Pu (梅花谱), the Plum Blossom Manual, is the other half of the famous pair. Chinese players speak of 橘梅, tangerine and plum, as the two great classics. Its author, Wang Zaiyue (王再越), lived in the Kangxi era of the Qing dynasty; little else is known about him, and the book's exact date is uncertain. It circulated in handwritten copies throughout the Qing and was not printed until a lithographic edition appeared in 1917.

Its big idea was defensive. Earlier books tended to answer the central cannon with more cannons. Wang's games showed the screen horse (屏风马), with both horses developed to guard the centre, not only holding the central cannon but beating it. The eight games of that section are usually considered the heart of the book, and together they argue that patience can overcome force. The contest it started, cannon against horse, has never really ended: the central cannon against the screen horse remains one of the main battlegrounds of modern play.

The full title is sometimes given as Meihua Bisheng Pu, the sure-to-win plum blossom manual. Its success produced a string of later books with plum blossom in their titles, so check that you have Wang's version if you go looking.

Taolüe Yuanji and the rise of the draw

Taolüe Yuanji (韬略元机) survives in editions dated 1707. Its original title used the character 玄, but that character appeared in the personal name of the Kangxi emperor, so it was changed to 元 to avoid the taboo. Sources disagree about who compiled it. Chinese references describe a group of editors, while at least one English account names a single author. The book is mostly compositions, with full games and handicap games at the end.

What makes it important is a change of mood. Many Ming compositions end with Red delivering a brilliant mate. Taolüe Yuanji gives far more weight to the defence, and many of its positions end in a draw after precise resistance. That shift toward hard, accurate defence shaped composition from then on.

Zhuxiangzhai, the composer's workshop

Zhuxiangzhai Xiangxi Pu (竹香斋象戏谱), also written 竹香斋象棋谱, is the high point of Qing composition. It was compiled by Zhang Qiaodong of Jiangsu together with his son Zhang Jingxu. The first two collections were printed in 1804. Zhang Qiaodong died in 1812 before the third was finished, and his son added it to a new printing in 1817. One English-language account says the name comes from a small bamboo retreat where Zhang and his friends met to test and correct positions.

Chinese writers often say that among Qing books the Plum Blossom Manual leads for full games and Zhuxiangzhai leads for compositions. It is usually counted among the four great Qing composition manuals, along with Xinwu Canpian (心武残篇) and Baiju Xiangqi Pu (百局象棋谱). The fourth place is disputed: some writers give it to Taolüe Yuanji, others to Yuanshen Haikuo (渊深海阔). Most of its positions end in draws, and the last collection holds the hardest problems.

How to read the old notation

Modern reprints almost always translate the moves into today's Chinese notation, which is worth learning because it is everywhere. A move names the piece, its file, the direction and the destination or distance. 进 means forward, 退 means back and 平 means sideways along a rank. By convention Red's files are written with Chinese numerals and Black's with Arabic digits, each side counting from its own right. So 炮二平五 is Red's cannon moving from file two to file five, the central cannon, and 马8进7 is Black's horse jumping forward from file eight to file seven. In English notation these are C2=5 and H8+7.

The originals can look very different. The Plum Blossom Manual, for example, recorded moves with a poem of ninety characters, one for each point on the board, so every move is a pair of characters naming a starting point and an ending point. Older texts also use variant characters, such as 包 for the cannon, and at least one early collection has Black moving first. If a move looks impossible, suspect the edition, the transcription or the notation before you suspect the author.

Rules have changed as well. Some old compositions rely on repeated checks or chases that modern rules would not permit, so a line that was a fair win in its day can be illegal now. When a solution looks odd, ask whether it depends on a repetition.

Why they still matter

You do not need these books to enjoy xiangqi, and a modern engine will find holes in many of their lines, yet they repay the effort. The openings you play today, from the central cannon to the screen horse, were argued out in their pages. Their compositions trained generations of players to calculate, and many practical endgames started life there. Reading them also gives you a feel for style: Ming books favour quick, romantic attacks, while later Qing works value patient defence.

The best way to use them is actively. Set up a real board, play through the main line, then go back and try each variation. Use an engine afterwards, not during, and notice where the old analysis breaks down; that is often where modern theory begins.

When a composition from one of these books finally clicks, keep the momentum going with XiangqiPlay's free library of 5,000 engine-proven mate puzzles.