Land battle chess 陆战棋

A game you cannot referee yourself

Luzhanqi is played by two people, and it takes three.

The pieces are small upright tiles that stand on the board with their written face turned towards their owner. Each player reads his own; nobody sees the other side's. So when a piece lands on a space held by the opponent, someone has to compare the two ranks to decide which of the two disappears.

This is where the game parts company with everything Europe knows. The rule does not say that both players turn their pieces over: it says that the loser leaves the board, and nothing more. The attacker learns that he has won, that he has lost, or that both pieces fell together. He does not learn what he fought.

That silence is the heart of the game, and two people cannot keep it alone. To compare two ranks you must see them both; if one of the players does the comparing, he must be shown a piece, and the secret collapses at the first fight. So the game needs an outsider who sees both camps, decides, and announces only the outcome. That person is the referee, 裁判 (cáipàn), and without one the game cannot begin.

The comparison with the closest Western game is telling. In Stratego, both pieces are revealed to both players: information leaks, but it leaks both ways at once, and no third party is needed. Luzhanqi made the opposite choice and paid for it in logistics — plenty of people have never played simply because the third person was missing.

This form, the one with hidden pieces, is called 暗棋 (ànqí), the dark game. It is the reference form, the one described in the rest of this page. Two other forms do without a referee by giving up the secret; they are mentioned further down.

Three names, and a homophone

The game goes by several names, all of them in use:

• 陆战棋 (lùzhànqí), literally "land battle chess";
• 陆军棋 (lùjūnqí), "army chess";
• 军棋 (jūnqí), the short form, by far the most common in speech.

The third name hides a coincidence the game seems to have sought out. 军棋 (jūnqí), the game, and 军旗 (jūnqí), the military flag you must capture to win, are perfect homophones: same syllables, same tones. Only the second character differs — 棋 "game piece" against 旗 "flag" — and both are read qí. The name of the game and its objective sound exactly alike.

Two of the board's terms are missing from this site's dictionary, so we give them here without being able to link them: 行营 (xíngyíng), the campsite, and 山界 (shānjiè), the mountain. Every other term used on this page is in it.

Where it comes from: what is documented, and what is merely repeated

This is the point about which the most nonsense is written, and it is worth separating carefully what is documented from what is merely repeated.

What is documented. There is an earlier French game, and its paper trail is clear: on 26 November 1908 Hermance Edan filed a French patent for a "battle game with mobile pieces on a board", granted in 1909 under number 396795. The game was sold from 1910 under the name L'Attaque, and the English publisher H. P. Gibson and Sons bought the rights in 1925. From that game descends Stratego, worked out by the Dutchman Jacques Johan Mogendorff, registered as a trademark in 1942 and published in the Netherlands in 1946. The principles are those of luzhanqi: ranked pieces, hidden, a flag to capture, bombs and mines.

It is also documented that military board games were being made in China in the same era. The Annals of Shanghai's Second Light Industry (上海二轻工业志, 1997) record that in 1927 Lu Xingchu (陆杏初), an employee of the Commercial Press in Shanghai, hired about ten workers in the Hongkou district and opened a games-piece workshop that produced junqi sets.

What is not. The direct line from the French game to the Chinese one is asserted almost everywhere, Chinese encyclopaedias included — they write that "luzhanqi has the French army chess as its ancestor". But the reference they give points to an amateur site that has since vanished and survives only in web archives. In other words: the claim circulates widely and rests, as far as we can tell, on no first-hand source. We report it as such, without adopting it.

What speaks for it still deserves to be said, because it is a line of reasoning rather than a belief. The agronomist Dong Shijin (董时进, 1900-1984), in Two Families (两户人家), a memoir published in Taiwan in 1981, describes a military game sold in his childhood and lists its pieces: the mine, the sapper, the division commander, the brigade commander, the spy, the flag. The spy is a piece from L'Attaque and Stratego; today's Chinese game has none. Conversely, that 1920s game had no bomb yet, today one of luzhanqi's most characteristic pieces. The same encyclopaedias add that a 1954 book describes the game in its modern form, and conclude that the current version — ranks modelled on the People's Liberation Army, railroads, campsites, headquarters — took shape between 1927 and 1954, by an unknown hand. The reasoning seems sound to us; we have not been able to check the two books ourselves.

The honest way to put it, in our view: a game probably imported, then remade in China to the point of becoming something else. This site's dictionary says much the same when it glosses 军棋 as a Chinese board game similar to Stratego.

The board: railroads, campsites, two headquarters

The board is what makes the game original — far more than its hierarchy of pieces.

Each side has thirty stopping points, laid out five columns by six rows, the two camps facing each other across a central band. Of those thirty points:

• twenty-three are 兵站 (bīngzhàn), posts — ordinary spaces;
• five are 行营 (xíngyíng), campsites: a piece standing there cannot be attacked, which makes them safe havens and unassailable observation posts;
• two are 大本营 (dàběnyíng), the headquarters, on the back row. The flag must be placed in one of the two, and a piece that enters a headquarters can never leave it again.

The central band has five spaces: three 前线 (qiánxiàn), the frontlines, through which you cross into enemy territory, and two 山界 (shānjiè), the mountains, which cannot be crossed at all. Sixty-five spaces in total.

Then comes the game's most recognisable trait: the lines joining these spaces are of two kinds. The 公路 (gōnglù), roads, along which a piece advances one space per turn. And the 铁路 (tiělù), railroads, drawn as thick lines, along which a piece runs in a straight line as far as it likes, as long as nothing stands in the way. Three railroads connect the two camps.

A Chinese board game crossed by a railway network is obviously nothing ancient, and that is the best clue to its age: luzhanqi is a twentieth-century game, depicting a war in which divisions move by rail. It descends neither from xiangqi nor from go, whose boards have not changed in centuries.

The twenty-five pieces

Each player fields twenty-five pieces: nineteen ranked units, three landmines, two bombs and one flag.

The basic rule fits on one line: the higher rank removes the lower, and two pieces of equal rank fall together.

The list is a vocabulary lesson in itself. These are not ranks in the Western sense but commands: each name is the unit level followed by 长 (zhǎng), "the head of". And the unit levels run in exactly the descending order of the Chinese military: 军 the corps, 师 the division, 旅 the brigade, 团 the regiment, 营 the battalion, 连 the company, 排 the platoon. Learn the luzhanqi board and you have learnt the Chinese army's order of battle — a rare return on a board game.

One rule has it that the fall of the field marshal forces its owner to reveal where his flag is. It is the game's finest idea: losing your best piece does not just cost a piece, it costs the secret everything else rested on.

The landmine, the bomb and the flag

Three pieces sit outside the hierarchy, and they are what gives the game its tension.

The landmine never moves. It is laid at the back of the formation and stays there for the whole game. It destroys more or less everything that touches it, and you have three.

The bomb moves like an ordinary piece, but any fight it starts against a unit, or suffers, destroys both pieces at once, whatever the rank across the board. It is the weapon of despair and the field marshal's nightmare: the highest piece in the game falls to a bomb exactly like the lowest. You have two, and they may not be placed on the first row at the start.

The flag does not move either. It sits in one of the two headquarters, beats nothing at all, and its capture ends the game. The whole game consists of protecting a perfectly useless object.

Now the real difficulty of luzhanqi comes into view. The flag is at the back, inside a headquarters, behind three mines that kill whatever touches them. Bombs can blow mines up, but they die doing it, and there are only two of them for three mines. In other words: force alone will not get you through.

The engineer, the most interesting piece in the game

Enter the 工兵 (gōngbīng), the engineer — the piece you must understand to understand the game.

The engineer is the weakest of all the units: bottom rank, it loses to absolutely everyone. And it holds the only two powers that decide games.

First: it is the only piece that can turn corners on the railroads. Every other piece runs in a straight line; the engineer follows the network, takes the bends, changes lines, and can reach in a single move any point of the rail that nothing blocks. The weakest piece in the game is also, by far, the most mobile.

Second: it is the only piece that clears mines without dying. Where any other piece destroys itself on a mine, the engineer lifts it and stays on the board.

The result is a complete inversion of instinct. Your two or three biggest pieces win the fights in midfield, but they will never take the flag: three mines stand in front of it. It is your three little sappers — the ones the opponent does not bother to intercept — that will open the way. And since there are only three, each one counts: a player who has lost all three engineers has lost the ability to win by force, even if he dominates everywhere else. He is left hoping for the other road to victory, the one where the opponent no longer has a single piece able to move.

Protecting three worthless pieces that decide the game, while making your opponent believe they are somewhere else: that is what playing luzhanqi is.

What the rules leave open

It has to be said plainly, because most introductions gloss over it: luzhanqi has no single rulebook. We have found no reference ruleset comparable to what go or chess enjoy, and usage varies from region to region and publisher to publisher. The rules described here are the most widespread; these are the points where you will most often see something else.

• The mine attacked by an ordinary piece. Depending on the rule in force, either it destroys the attacker and stays in place, or both disappear together. That is no small detail: in the first case three mines are a wall, in the second they are worth three pieces and no more.
• Where mines may be laid. Some rules confine them to the last two rows; others allow them anywhere.
• The run along the rail. Many tables cap it at three spaces for every piece but the engineer, instead of letting them run without limit.
• The mountain. In some regions you may enter it and shelter there as in a campsite; elsewhere it is simply impassable; elsewhere again only the engineer may cross.
• The field marshal. Local rules forbid it to take the flag, or declare the game drawn if one side loses its field marshal while the other's still stands.

Before a first game with someone, it is worth spending three minutes settling those five points. That is the custom in China, and nobody there takes it for a weakness of the game.

Two other forms are commonly played, and they need no referee. 明棋 (míngqí), the open game: each player sets up in secret, then everything is turned face up and the game is played with full information, which makes it pure calculation. And 翻棋 (fānqí), the flipping game: the pieces are shuffled face down and turned over as play goes on, which brings in a good deal of luck. Both are easier to organise, and markedly shallower than the dark game.

The referee became a program

A game that requires a third person is a game that does not get played often. That is what held luzhanqi back for so long, and why it almost never left China: the games that travel are the ones two people can take out of a box, any evening, without recruiting anyone.

The computer removed the obstacle at a stroke, without changing a single rule. A program sees both camps, compares the ranks and announces only the outcome: that is the definition of the referee. It does not cheat, does not tire, and is always available. Chinese gaming platforms did it early, and luzhanqi found a huge audience there in a four-player variant, 四国军棋 (sìguó jūnqí), "the war chess of the four countries", which appeared in mainland China in the 1980s: two teams of two facing each other, each player ignorant even of his own partner's pieces.

It is a rare and rather satisfying case: computing did not replace a human player, it replaced a role nobody wanted. Luzhanqi never needed a machine to play against. It needed a referee.