Core Objective of Riichi Mahjong

1.1 What Is Riichi Mahjong?

Riichi mahjong (立直麻雀, riichi maajan) is the standard Japanese variant of mahjong, a four-player tile game that combines elements of draw-and-discard mechanics, pattern recognition, probability assessment, and strategic decision-making. Unlike many Western card games, riichi mahjong is fundamentally a four-player game in which every player's decisions affect every other player at all times. The game is played with 136 tiles (or 137 in some formats that use a red blank), and the objective within each hand is to assemble a complete winning pattern of tiles before your opponents do.

However—and this is a critical distinction that many beginners miss—the objective of riichi mahjong is not simply to win individual hands. The game is played over multiple rounds (typically 8 or more hands in a hanchan/半荘, or 4+ in a tonpuusen/東風戦), and the ultimate objective is to finish with the best placement (着順, chakujun) relative to the other three players. A player who wins many small hands but deals into one enormous hand may finish last. A player who wins no hands at all but avoids dealing in while others self-destruct may finish second. The game rewards holistic decision-making across an entire match, not just within individual hands.

1.2 The Dual Objective: Hands and Match

At the hand level, each player is trying to complete a valid winning hand (和了, agari) by drawing and discarding tiles until their 13-tile hand, plus one winning tile, forms the required pattern (usually four groups and one pair). If you complete this pattern and satisfy certain conditions—most importantly, having at least one scoring element called a yaku (役)—you win the hand and receive points from one or more opponents.

At the match level, each player begins with a starting score (typically 25,000 points) and gains or loses points based on hand outcomes. At the end of the match, players are ranked by their final scores. In competitive play and on ranked online platforms, placement is what matters: first place is rewarded, fourth place is punished, and the specific point margin is secondary to the rank itself. This means that riichi mahjong strategy is not about maximizing raw points—it is about maximizing your expected placement across the match.

This dual-layer structure—hand-level tactics plus match-level strategy—is what gives riichi mahjong its extraordinary depth. A decision that is correct in isolation (such as going for a high-value hand) may be incorrect given the match context (such as when you are in first place in the final round and a cheap hand would guarantee your lead). Understanding this interplay is the central intellectual challenge of the game.

1.3 The Three Modes of Play in Each Hand

Within each hand, a player is generally in one of three broad modes:

Attacking (攻撃, kougeki): Actively pursuing a winning hand. This means making discard choices that bring your hand closer to tenpai (ready to win), potentially declaring riichi to increase your reward, and accepting some risk of dealing into an opponent's hand in exchange for the opportunity to win.

Defending (守備, shubi): Abandoning your own hand pursuit to avoid dealing into an opponent's winning hand. This means discarding safe tiles—tiles that cannot or are unlikely to be the tile an opponent needs to win—even if those tiles would have been useful for your own hand. Defense is not passive or cowardly; it is a critical strategic skill that directly affects your placement by preventing point loss.

Balancing (判断, handan): The most common and most difficult mode, where a player must weigh the value of continuing to attack against the risk of dealing in. Most of the strategic depth in riichi mahjong comes from these judgment calls. The push/fold decision (押し引き, oshi-hiki) is widely regarded as the single most important skill in competitive play.

1.4 Points, Placement, and What "Winning" Means

In most competitive and online riichi mahjong, placement bonuses (順位点, jun'iten, commonly called ウマ "uma") are applied after the match. A common uma structure is +30/+10/−10/−30 (thousand-point units), meaning first place gains 30,000 bonus points and fourth place loses 30,000. This makes the difference between third and fourth place worth 20,000 points on top of any actual score difference—a massive incentive to avoid last place even when your score is already low.

This placement structure creates asymmetric incentives. When you are in first place, risky aggressive play is rarely justified because the downside (losing your lead) is much worse than the upside (increasing a lead you already have). When you are in fourth place in the final round, desperate aggressive play may be justified because you have little to lose. Understanding these asymmetries is the foundation of placement strategy (着順戦略), covered in Module 21.

1.5 Riichi Mahjong as an Imperfect-Information Game

Riichi mahjong is a game of imperfect information. Each player can see only their own 13 tiles, the tiles that have been discarded (visible to all), and any tiles that have been called into open melds. The remaining tiles in the wall, and the other players' concealed hands, are unknown. This means that every decision involves probability estimation, risk assessment, and inference from available information.

The imperfect information nature of the game also means that correct decisions can lead to bad outcomes (you make the right fold but the tile was actually safe) and incorrect decisions can lead to good outcomes (you push recklessly but happen to win). Over a small number of hands, luck dominates. Over hundreds or thousands of hands, skill dominates. This is why serious riichi mahjong players focus on decision quality rather than outcome quality, and why review skills (Module 29) and statistical thinking (Module 30) are essential.

1.6 The Role of Luck and Skill

There is an ongoing discussion in the mahjong community about the ratio of luck to skill. In Japanese mahjong literature, it is commonly acknowledged that in a single hand, luck plays a very large role—you cannot control what tiles you draw. However, across a full match (hanchan), and especially across many matches, skilled players consistently outperform weaker ones. Professional players in Japan's M-League and top-ranked Tenhou players demonstrate sustained high performance over thousands of games, which would be statistically impossible if the game were purely luck-based.

The skill elements include: tile efficiency in building hands, accurate defense, reading opponents' hands from their discards, push/fold judgment, placement-aware strategy, and emotional discipline. Each of these is covered in dedicated modules throughout this academy.

1.7 Summary of Core Objectives

1.8 Beginner Game Walkthrough — Understanding the Objective in Practice

To make the core objective concrete, consider a typical beginner's first hanchan. The four players start at 25,000 points each. In the first hand (East 1), you are dealt 13 tiles. You look at your hand and see scattered tiles across all three suits with a pair of South wind. You are at approximately 5-shanten—very far from tenpai. Over the next several turns, you draw tiles, discard the least useful ones (isolated terminals and guest winds first), and gradually your hand improves to 3-shanten, then 2-shanten. On turn 12, an opponent declares riichi.

This is the first critical decision point. Your hand is at 2-shanten with no dora and modest value potential—perhaps 1-2 han if you complete it. The opponent is already tenpai and threatening. The correct play, which most beginners fail to make, is to abandon your hand immediately and switch to pure defense (betaori). Your probability of winning this hand is very low (you need at minimum 2 more perfect draws to reach tenpai, during which time you must discard potentially dangerous tiles), while the probability of dealing into the opponent's riichi is significant and the cost could be 3,000-12,000 points. The expected value of pushing is strongly negative.

The beginner who understands the core objective—maximizing placement across the entire match, not winning every hand—will fold here without hesitation. The beginner who only thinks about "completing my hand" will push forward, deal in, and lose thousands of points unnecessarily. This single conceptual shift—from "I want to win this hand" to "I want to make the decision with the highest expected value for my placement"—is the foundation everything else builds upon.

1.9 How a Hanchan Actually Unfolds — Typical Score Trajectories

In a typical hanchan among intermediate players, the following patterns are common. During the East round (hands 1-4 minimum), scores fluctuate as players win hands and deal in. It is common for the lead to change hands multiple times. One player may jump to 35,000+ with an early mangan win, while another drops to 18,000 after dealing into a riichi hand. By the start of the South round, a rough hierarchy has usually emerged, though comebacks remain very possible.

During the South round, awareness of placement becomes increasingly important. A player in first place with a comfortable lead should shift toward conservative play—folding more often, avoiding unnecessary risks, and accepting small wins or draws rather than pursuing expensive hands. A player in fourth place should become more aggressive, seeking high-value hands and accepting risks they would normally avoid, because the marginal cost of dropping further into fourth is much smaller than the benefit of climbing to third or second.

The final hand (all-last/オーラス) is where placement strategy reaches its peak intensity. Every decision must be made with explicit reference to point gaps and target scores. A player who needs exactly 3,900 points to overtake the player above them should build toward exactly that threshold—not more, not less. Overshooting wastes effort; undershooting wastes the opportunity. This calculation-driven approach to the endgame is a hallmark of strong play, extensively discussed in the works of Kobayashi Gō (小林剛), a professional player known for analytical precision.

1.10 The Four Outcomes of Every Hand

Every hand ends in one of four ways from your perspective, and understanding the expected value of each is fundamental:

You win (agari): You gain points. The amount depends on your hand value (han + fu), whether you won by tsumo or ron, whether you are dealer, and the honba count. Winning is good, but the value varies enormously—a 1-han 30-fu hand (1,000 points by ron as non-dealer) is worth 32 times less than a yakuman (32,000 points).

You deal in (houjuu): You lose points. The amount depends on the winner's hand value. Deal-ins are costly not only for the direct point loss but for the placement implications—a mangan deal-in (8,000 points) can easily drop you two placements. Minimizing deal-in rate is the single most impactful statistical improvement most players can make. Data from Tenhou's ranked games shows that the deal-in rate of top-ranked players (鳳凰卓, Houou table) averages approximately 10-12%, compared to 14-16% for lower-ranked players. This 3-5% difference in deal-in rate accounts for a large portion of the skill gap.

An opponent wins from another opponent (yokozuke/lateral movement): You are unaffected except by tsumo payments (if the win is by tsumo). Lateral wins between other players can be strategically favorable if they damage the player directly above you in placement.

Exhaustive draw (ryuukyoku): If you are tenpai, you receive a share of the 3,000-point noten penalty pool. If you are noten, you pay into it. Being tenpai at a draw is a modest gain; being noten is a modest loss. The dealer retains their position if tenpai, which can be significant.

1.11 Comparative Difficulty and Learning Curve

Riichi mahjong is sometimes compared to poker as a game of imperfect information with betting-like risk management. However, several structural differences make riichi mahjong uniquely challenging. The four-player structure means threats come from three directions simultaneously, not one. The furiten rule creates a deterministic defense system absent in poker. The scoring system's non-linearity (a hand can be worth 1,000 or 32,000 depending on composition) creates dramatic variance. And the match structure (placement over multiple hands) requires sustained strategic consistency rather than the hand-by-hand independence of most poker formats.

The learning curve in riichi mahjong, as described by Fukuchi Makoto (福地誠) in his instructional writings, follows a characteristic pattern: rapid initial improvement as basic rules and tile efficiency are learned, a plateau as defense and push/fold concepts are introduced (these feel counterintuitive to new players who want to "play their hand"), and then gradual, sustained improvement as reading skills, placement strategy, and emotional discipline develop over thousands of games.