Loading field map
Yaku theory is the branch of mahjong study concerned with the yaku—the predefined patterns of tiles that a player must assemble to legally win a hand. In most modern mahjong rulesets, a winning hand is not merely a complete set of fourteen tiles; it must also contain at least one yaku, a condition that functions as a scoring category, a strategic objective, and a gatekeeping mechanism all at once. Yaku theory examines what these patterns are, how they are classified, how they interact, and how their design shapes the game's strategy.
To understand yaku theory, one must first understand the basic structure of a mahjong hand. A standard hand consists of fourteen tiles: four sets of three (either triplets or sequences) plus one pair. This structure alone, however, is insufficient to win in most rulesets. The yaku requirement ensures that a player's hand meets some additional criterion of intentionality or difficulty. Common yaku include having no sequences at all (all triplets), having a hand composed entirely of tiles from one suit plus honors, or having a particular terminal or honor tile in every set.
The yaku system serves several functions simultaneously. It prevents wins that are purely accidental, since a random collection of tiles rarely forms a yaku. It creates a hierarchy of hand values, since different yaku carry different point values. And it shapes the strategic landscape: players must decide early in a hand which yaku to pursue, and the availability of certain yaku influences which tiles are discarded, which are kept, and when a player declares a win.
Yaku theory is distinct from the broader study of mahjong strategy in that it focuses on the yaku themselves as a designed system. It asks questions like: What makes a yaku easy or difficult to achieve? How do yaku overlap, and which combinations are mutually exclusive? How does the point value of a yaku relate to its probability of occurrence? These questions are not merely academic; they inform how players evaluate hands, how rulesets are designed, and how the game evolves across regional variants.
The modern yaku system emerged from the codification of mahjong rules in the early twentieth century, particularly in Japan. Mahjong originated in China, where early rulesets had no standardized yaku list; winning conditions varied by region and even by house. The Japanese game, which developed its own distinct ruleset in the 1920s and 1930s, introduced a fixed list of yaku with assigned point values. This codification was a significant departure from earlier practice and is the direct ancestor of most yaku theory as it exists today.
The Japanese system, often called riichi mahjong, established a canonical set of roughly forty yaku, ranging from simple patterns like a single sequence of one suit (which is actually not a yaku in most rulesets) to rare and valuable patterns like a hand containing all four winds and a pair of the player's own wind. The riichi ruleset also introduced the concept of dora, bonus tiles that add points but do not count as yaku. This distinction—between yaku, which are required to win, and dora, which only increase score—is central to yaku theory.
Other regional rulesets have their own yaku systems. Chinese mahjong, in its various standardized forms, uses a different scoring framework that often includes many more patterns and assigns points differently. Western mahjong, popular in the United States and Europe, has its own simplified yaku list. Yaku theory, as a field of study, must therefore account for these variations. The Japanese system is the most thoroughly analyzed, but the theoretical questions—how yaku are defined, valued, and combined—apply across all variants.
A central task of yaku theory is classification. Yaku can be grouped along several axes, and these groupings reveal the underlying logic of the system.
One fundamental distinction is between closed and open yaku. A closed hand is one that has not called any tiles from another player's discard; all sets must be drawn from the wall. Some yaku, like the pure double sequence (two identical sequences in the same suit), are only valid if the hand is closed. Others, like all triplets, can be achieved with an open hand. This distinction matters because calling tiles—declaring a pung, chi, or kan from another player's discard—makes a hand easier to complete but often disqualifies it from certain yaku. The closed/open distinction is a primary axis of yaku classification because it directly affects the risk-reward calculus of every hand.
Another axis is suit-based versus structure-based yaku. Suit-based yaku depend on which tiles are present: a hand with only one suit plus honors, a hand with only terminals and honors, a hand with no honors at all. Structure-based yaku depend on how the tiles are arranged: all triplets, all sequences, a hand with a specific pattern like seven pairs. Many yaku combine both axes, such as the "pure straight" (three consecutive sequences in one suit), which is both suit-based and structure-based.
A third axis is rarity and value. Yaku are assigned point values that roughly track their difficulty, though the correlation is imperfect. The rarest yaku, like the "thirteen orphans" (one of each terminal and honor tile plus one duplicate), carry the highest values. Common yaku, like having a triplet of dragons, carry low values. Yaku theory examines this relationship between probability and payoff, asking whether the point values are well-calibrated and how they influence strategic choices.
Yaku rarely occur in isolation. A single hand can satisfy multiple yaku simultaneously, and the total score is the sum of their values (with multipliers in some rulesets). This stacking property is a key subject of yaku theory. Some yaku naturally combine—for example, a hand that is all triplets and also contains no honors satisfies two yaku at once. Others are mutually exclusive by definition: a hand cannot be both all sequences and all triplets.
The interaction of yaku creates a combinatorial landscape that players must navigate. A hand that satisfies one yaku may be one tile away from satisfying a second, and the choice of which tile to pursue can dramatically change the hand's value. Yaku theory analyzes these adjacency relationships: which yaku are "near" each other in tile space, and which combinations are worth pursuing versus which are traps that force a player into a low-value hand.
A particularly important interaction is between yaku and furiten, the rule that a player cannot win on a tile they have previously discarded. This rule, central to riichi mahjong, interacts with yaku in subtle ways. A player pursuing a specific yaku may be forced to discard a tile that later becomes their winning tile, and the yaku system determines whether an alternative win condition is available. Yaku theory thus connects to the broader strategic framework of the game, including defense and risk management.
Yaku theory is not a single unified discipline but rather a set of approaches that ask different questions and use different methods.
The taxonomic approach is the most basic: it catalogs yaku, defines their conditions precisely, and organizes them into classification schemes. This approach is foundational because it establishes the vocabulary and the boundaries of the system. Taxonomic work is largely descriptive, but it is not trivial. The precise conditions of a yaku—for example, whether a "pure double sequence" requires the two sequences to be in the same suit or merely the same rank—can be subtle, and different rulesets define the same yaku differently. Taxonomic yaku theory clarifies these definitions and exposes ambiguities.
The probabilistic approach uses combinatorial mathematics to calculate the likelihood of achieving various yaku. This approach asks: given a starting hand and a sequence of draws, what is the probability of completing a particular yaku? These calculations are complex because they must account for the actions of other players, who are simultaneously discarding tiles and potentially blocking the hand. Probabilistic yaku theory provides a quantitative foundation for evaluating hand values, though its models are necessarily simplified. The results are used to calibrate point values and to inform strategic decisions about which yaku to pursue.
The strategic approach treats yaku as a decision-making framework. It asks: given the yaku available, how should a player choose which hand to build? This approach integrates yaku theory with the broader game theory of mahjong, including hand evaluation, tile efficiency, and defense. Strategic yaku theory is the most practical and the most widely discussed in player communities. It analyzes trade-offs—for example, whether to pursue a high-value yaku that is difficult to complete or a low-value yaku that is nearly guaranteed—and develops heuristics for hand selection.
The design approach examines the yaku system as a designed artifact. It asks: what makes a good yaku system? Are the point values well-calibrated? Do the yaku encourage interesting strategic choices, or do they create degenerate strategies? This approach is comparative, examining how different rulesets handle the same design problems. It is also historical, tracing how yaku have been added, removed, or modified over time. Design-oriented yaku theory is less formalized than the other approaches but is increasingly relevant as new mahjong variants are created and as the game is adapted for digital platforms.
These approaches are not mutually exclusive. A complete understanding of yaku theory requires all four. The taxonomic approach provides the raw material; the probabilistic approach quantifies it; the strategic approach applies it; the design approach evaluates it. In practice, most serious analysis of yaku combines elements of all four.
Yaku theory today is a mature but still-evolving field. The Japanese riichi system remains the most thoroughly analyzed, with a large body of strategic literature and sophisticated probabilistic models. Chinese and Western systems have received less formal attention, though they are studied within their respective player communities.
The rise of online mahjong platforms has had a significant impact on yaku theory. These platforms collect vast amounts of game data, enabling empirical analysis of yaku frequencies and win rates. This data-driven approach complements the theoretical probabilistic models and has sometimes revealed that the assumed difficulty of certain yaku does not match their actual occurrence rates. The availability of large datasets has also made it possible to test strategic heuristics rigorously, leading to refinements in hand evaluation.
Another development is the growth of mahjong as a competitive sport, particularly in Japan and China. Competitive play has driven demand for more precise yaku theory, as players seek any edge in high-stakes matches. This has led to more formalized strategic frameworks and a deeper analysis of the interactions between yaku and other rules, such as furiten and the scoring system.
Yaku theory remains a niche field, studied primarily by serious players rather than academics. It has no formal institutions or journals, and much of its literature exists in player forums, strategy guides, and video content. Nevertheless, it is a coherent body of knowledge with its own questions, methods, and standards of evidence. For the educated newcomer, the key to understanding yaku theory is to recognize that it is not a single doctrine but a set of complementary approaches to a shared problem: how the yaku system structures the game of mahjong, and how players can navigate that structure to win.