Tenpai, Shanten, and Ukeire

7.1 Tenpai (聴牌) — Ready to Win

A hand is in tenpai (聴牌) when it needs exactly one more tile to become a complete winning hand. In other words, tenpai is the state where you are "one tile away" from agari (winning). Being in tenpai is significant for several reasons: you can declare riichi if your hand is closed, you receive points from noten players at an exhaustive draw, and you are actively threatening opponents with the possibility of winning on their discards.

The tiles that would complete a tenpai hand are called machi (待ち, "waits") or agari hai (和了牌, "winning tiles"). The number and type of waits determine both the likelihood of winning and, to some extent, the hand's score. Wait types are covered in detail in Module 09.

7.2 Shanten (向聴) — Measuring Hand Progress

Shanten (向聴, also written 向聴数 shanten-suu) is the measure of how many tiles away from tenpai your hand is. A hand at 0-shanten is tenpai (one tile from winning). A hand at 1-shanten (iishanten) needs to draw one useful tile and discard one tile to reach tenpai. A hand at 2-shanten needs two improvements, and so on.

Your starting hand after the deal is typically 4-6 shanten, meaning you need several good draws to reach tenpai. Reducing shanten with each discard is the fundamental mechanical skill of hand progression. A discard that reduces shanten is an effective discard; a discard that increases shanten is a backward step (though sometimes strategically justified, e.g., for defense).

Calculating Shanten

Calculating shanten accurately by hand requires practice. The basic idea is: count the number of complete mentsu (groups) and the number of partial mentsu (pairs, connected tiles like 4-5, or close tiles like 4-6) in your hand, then use the formula:

shanten = (4 - complete_mentsu) - partial_mentsu_count (simplified; actual algorithm is more nuanced for edge cases)

The precise algorithm must also account for the pair, chiitoitsu structure, and kokushi structure, choosing whichever gives the lowest shanten. Online tools and software calculate shanten automatically, but understanding the concept is essential for making good discard decisions.

7.3 Ukeire (受入) — Tile Acceptance

Ukeire (受入, "acceptance" or "acceptance count") measures how many tiles in the remaining unseen tiles would improve your hand (typically, how many tiles would reduce your shanten by one). This is the most important metric for tile efficiency.

For a tenpai hand, ukeire equals the number of remaining copies of your winning tiles. If you are waiting on 3m and 6m, and two copies of 3m and three copies of 6m remain unseen, your ukeire is 5. Higher ukeire means a greater chance of winning.

For a hand at 1-shanten or higher, ukeire counts all tiles that would bring you one step closer to tenpai. A discard choice with higher ukeire is generally superior from a pure speed perspective—it gives you the best chance of progressing toward tenpai.

7.4 Practical Shanten Counting Tips

Step 1: Sort your hand by suit. Group tiles that form or are close to forming mentsu.

Step 2: Count complete mentsu (three-tile groups). Count partial mentsu (pairs, sequential pairs like 4-5, and gaps like 4-6). Count isolated tiles.

Step 3: Remember you need 4 complete mentsu + 1 pair. Each missing mentsu is roughly one shanten. Partial mentsu reduce the count because they can become complete with one tile.

7.5 Relationship Between Shanten, Ukeire, and Strategy

The strategic significance of these concepts is profound. Early in the hand (high shanten), your primary goal is typically to reduce shanten as fast as possible—tile efficiency dominates. As you approach tenpai (low shanten), other considerations like hand value, wait quality, and defense become more important. A hand at 1-shanten with excellent ukeire is very close to winning and should usually be prioritized. A hand at 3-shanten may not reach tenpai before the hand ends, and the player might consider defensive options instead.

The interplay between shanten/ukeire and hand value is the subject of Module 27 (Efficiency vs. Value). For now, understand that shanten tells you how far from winning, and ukeire tells you how likely your next step is.

7.6 Shanten in Practice — Beginner Hand Analysis

Here is a typical beginner starting hand that we will analyze step by step:

Sorting and analysis: Manzu: 1m (isolated terminal), 5m (isolated middle), 8m (isolated). Pinzu: 2p-4p (kanchan partial, needs 3p), 7p-9p (kanchan partial, needs 8p). Souzu: 3s-3s (pair), 6s (isolated). Honors: 1z (isolated guest wind), 3z (isolated guest wind), 5z (haku, yakuhai potential).

Counting: 0 complete mentsu, 2 kanchan partials (2p-4p, 7p-9p), 1 pair (3s-3s), several isolated tiles. Using the approximate formula: shanten ≈ (4 − 0) − 2 partials − 1 pair-related ≈ 4 − 3 = roughly 3-shanten considering the pair partially fills the jantai requirement. More precisely, this hand is about 4-shanten — quite far from tenpai.

What should the first discard be? Following the opening strategy principles from Module 16: discard 1z or 3z (isolated guest winds with zero sequence potential and no yakuhai value). Keep 5z (haku — yakuhai potential if you draw a second). The isolated manzu terminals and middle tiles are also weak, but they have some suited connectivity. Guest wind honors go first.

7.7 Ukeire Calculation — Worked Example

Advanced ukeire counting considers not just which tiles improve your hand but how many copies of each remain unseen. This is where tile counting (Module 2, Section 2.10) intersects with efficiency.

Ukeire calculation tools are available online. Tenhou's replay analyzer computes optimal ukeire for every discard choice. The Japanese website "牌効率論" (Hai Kouritsu Ron, "Tile Efficiency Theory"), maintained by analytical mahjong enthusiasts, provides interactive ukeire calculators. Fukuchi Makoto's (福地誠) book series includes printed ukeire drill sets for offline practice.

7.8 The Relationship Between Shanten and Win Probability

Data from large Tenhou game samples, as analyzed by the Japanese mahjong data community, shows the following approximate relationships between shanten at the time an opponent declares riichi and the probability of completing your hand:

These numbers, while approximate, demonstrate why shanten is the single most important input to push/fold decisions. The dropoff between 1-shanten and 2-shanten is enormous — this is the key threshold. As noted in Kawada Jiro's (川田浩之) statistical mahjong analyses, the tenpai probability at 2-shanten is roughly half that at 1-shanten, making the expected value of pushing dramatically worse.