Loading field map
For centuries, players of xiangqi have faced a fundamental question: how can one evaluate a position when no immediate tactical blow is available? The answer lies in positional strategy—the study of long-term factors such as piece coordination, space control, pawn structure, and the balance between material and initiative. Unlike tactics, which deal with concrete sequences, positional judgment relies on principles that must be tested, refined, and sometimes overturned. The history of positional thinking in xiangqi can be divided into three broad schools, each defined by its sources of authority, its methods of analysis, and the specific positional concepts it emphasized.
The earliest systematic approach to positional strategy emerged from centuries of practice and was transmitted through manuals, game collections, and oral tradition. Classical masters such as those who compiled the Xiangqi Manual of the Ming dynasty and later Qing-era commentators developed a body of maxims that guided positional play. These principles were not derived from formal analysis but from accumulated experience and the authority of past champions.
Key classical concepts included the importance of controlling the river—the central dividing line of the board—as a staging ground for attacks and a barrier against enemy incursions. Pawn structure was treated with care: advancing pawns too early could weaken one's own position, while well-placed pawns could restrict the opponent's pieces. Piece coordination was prized above individual piece activity; a common maxim held that "horses should be linked, cannons should have platforms, and chariots should occupy open files." The classical school also valued the initiative (xian shou) but understood it primarily as the right to attack, not as a measurable quantity.
These maxims were taught through rote memorization of model games and annotated positions. The authority of the classical school rested on the prestige of its sources rather than on systematic verification. As a result, many principles were accepted without question, and positional debates were rare. The classical school provided a stable foundation, but its reliance on tradition left little room for innovation or for testing its claims against a large body of competitive play.
The rise of organized tournament play in the early twentieth century created a new kind of evidence: thousands of recorded games played under standardized conditions. This shift allowed analysts to move beyond received wisdom and begin testing classical maxims against actual results. The Modern Analytical School, which dominated the twentieth century, treated tournament games as its primary data source and sought to refine positional concepts through statistical and comparative methods.
Modern analysts did not reject the classical school outright; instead, they reinterpreted its principles in light of concrete evidence. For example, the classical emphasis on controlling the river was refined into a more nuanced understanding of central squares. Analysts like the Chinese grandmaster Hu Ronghua and later theorists such as Liu Dahua demonstrated that control of the central files (the 4th and 5th ranks) often outweighed control of the river itself. Pawn structure received deeper scrutiny: the modern school identified specific pawn formations—such as the "linked pawns" and "isolated pawns"—and evaluated their strengths and weaknesses in different contexts.
A major contribution of the modern school was the concept of positional compensation. Classical maxims had treated material advantage as paramount, but modern analysts showed that a material deficit could be offset by superior piece activity, better pawn structure, or control of key squares. This idea was formalized in the notion of "initiative" as a measurable resource that could be traded for material. The modern school also introduced systematic opening theory, which integrated positional principles into specific opening systems like the Central Cannon and the Screen Horse Defense.
Despite its advances, the modern school remained limited by human cognitive capacity. Analysts could only examine a finite number of variations, and their evaluations were inevitably colored by intuition and experience. The school's methods were rigorous by human standards, but they could not keep pace with the explosive growth of competitive play in the late twentieth century.
The arrival of strong xiangqi engines in the early 2000s—programs like Xiangqi Wizard and later Deep Xiangqi—transformed positional analysis. These engines evaluate positions using brute-force search combined with sophisticated evaluation functions that assign numerical scores to features such as piece activity, pawn structure, king safety, and space control. The Engine-Driven Positional School treats engine evaluations as the primary arbiter of positional truth, though human analysts still interpret and contextualize the results.
Engines have overturned several long-held beliefs of the modern school. For instance, the modern school had considered certain pawn structures—like the "double pawn" on the same file—as inherently weak, but engines showed that in many positions the structural weakness was outweighed by dynamic compensation. Similarly, the classical and modern schools had both emphasized the importance of controlling the center, but engines revealed that in some opening systems, flank attacks could be equally effective if the center was properly restrained.
The engine-driven school has not replaced human positional judgment; rather, it has forced it to become more precise. Modern players use engines to test their own evaluations, to discover new positional ideas, and to refine their understanding of classical principles. The school's methods are empirical and data-driven, but they still require human insight to ask the right questions and to apply engine findings to practical play.
All three schools agree on the fundamental importance of space control, piece coordination, and the initiative. They also share the view that positional factors are interconnected: a weakness in one area can be compensated by strength in another. However, they disagree on the stability of positional principles. The classical school treated its maxims as timeless truths; the modern school saw them as hypotheses to be tested; the engine-driven school treats them as approximations that can be refined or discarded based on computational evidence.
A key area of disagreement concerns the role of human intuition. The classical and modern schools relied heavily on the judgment of experienced players, while the engine-driven school tends to privilege objective evaluation over subjective feel. This has led to a living disagreement about whether certain positional concepts—such as "good bishop" versus "bad bishop" in xiangqi's palace-centered play—are genuinely useful or merely heuristic shortcuts.
Today, positional strategy in xiangqi is a hybrid discipline. Top players routinely use engines to analyze their games and to prepare openings, but they also draw on the accumulated wisdom of the classical and modern schools. The classical maxims survive as a pedagogical tool for beginners, while the modern school's analytical methods remain essential for understanding game records. The engine-driven school provides the most precise evaluations, but its findings are constantly being integrated into human-oriented frameworks. The result is a richer, more nuanced understanding of positional play than any single school could have achieved alone.