MODULE 27

Efficiency versus Value

27.1 The Fundamental Tradeoff

Tile efficiency (牌効率) maximizes the probability of reaching tenpai quickly. Hand value maximizes the points received when winning. These two goals frequently conflict: the most efficient discard (highest ukeire) may lead to a lower-value hand, while a discard that preserves a high-value hand shape may sacrifice efficiency. Navigating this tradeoff is one of the most important skills in intermediate-to-advanced play.

27.2 When Efficiency Dominates

Early hand (high shanten): At 4+ shanten, tenpai is far away and hand value is hard to predict. Prioritize efficiency to progress toward tenpai. Value considerations are secondary because your hand shape will change many times before tenpai.

Low-value situations: When your hand has no clear path to high value anyway, maximize speed. A 1-han tenpai hand is better than a 0-han non-tenpai hand.

Speed-critical scenarios: When you need to win quickly (ending opponents renchan, reaching tenpai for noten penalties, etc.), efficiency dominates.

27.3 When Value Dominates

Near tenpai with value options: At 1-shanten, you can often see whether a particular discard preserves or destroys a high-value yaku. Sacrificing 1-2 tiles of ukeire to maintain mangan potential is often correct.

Dora-rich hands: Hands with dora are inherently more valuable; small efficiency sacrifices to keep dora in the hand are justified.

Placement-critical situations: When you need a specific score threshold (e.g., mangan to move from 3rd to 1st), preserving value becomes essential even at significant efficiency cost.

27.4 Practical Guidelines

1-tile ukeire difference is small. If two discards differ by only 1-2 tiles of ukeire but one preserves pinfu/tanyao potential while the other destroys it, prefer the value-preserving option. The efficiency cost is marginal; the value difference may be significant.

4+ tile ukeire difference is large. Sacrificing 4+ tiles of ukeire for value is rarely justified except in extreme cases (preserving a near-certain yakuman, or needing an exact score in all-last).

The "2x rule" heuristic: Some Japanese strategy sources suggest that you should sacrifice efficiency for value when the value difference is at least double the base hand. For example, if pure efficiency gives a 2-han hand but a slightly less efficient path gives 4-han, the value path is likely correct.

QUIZ — Question 27.1

At 1-shanten: Discard A gives 12 ukeire and leads to a 1-han tenpai. Discard B gives 10 ukeire and leads to a 3-han tenpai. Which is generally better?

  • A. Discard A — always maximize ukeire.
  • B. Discard B — the 2-tile ukeire sacrifice is small, and the value jump from 1 to 3 han roughly quadruples the hand worth.
  • C. They are equivalent.
  • D. Cannot determine without knowing the specific tiles.

Answer: B. A 2-tile ukeire difference (12 vs 10) is relatively small—roughly 17% less acceptance. But the value jump from 1 han to 3 han is enormous (approximately 4× the score). The expected value calculation (win probability × hand value) strongly favors Discard B. This is a textbook case where value should override efficiency.

27.5 The Decision Framework — Worked Example

Your hand at 1-shanten:

2m 3m 4m 5p 5p 6p 7p 4s 5s 6s 7s 8s 1z

You draw 8p. Your hand is now 14 tiles. What to discard?

Option A: Discard 1z. Keeps the pinzu block flexible. Your tenpai possibilities include ryanmen waits in pinzu. The hand heads toward tanyao + pinfu potentially (if you discard the 1z and avoid terminals). Ukeire: high, as many pinzu and souzu tiles improve you.

Option B: Discard 8p. Keeps the 1z in hand. This makes no sense tile-efficiency-wise — 1z is strictly less useful than 8p for building sequences. But perhaps you have a reason to hold 1z (it is a guest wind and you want defensive options).

In this case, pure efficiency and value both point to discarding 1z: it is isolated, has no sequence potential, and blocks tanyao. The 8p connects to the 5p-6p-7p area. Discard 1z — this is a case where efficiency and value align.

But consider the modified scenario: the dora indicator is 7z (→ dora is 1z!). Now suddenly that isolated 1z is worth 1 han. Keeping it in your hand adds significant value. Is it worth sacrificing some efficiency (discarding a better-connected tile instead) to keep the dora? In this case, yes — the value of 1 dora han typically outweighs a small efficiency loss. This is the quintessential efficiency-vs-value decision, and the dora changes the calculus completely. Kobayashi Gō (小林剛) has emphasized in his M-League commentary that "ドラは1牌効率に勝つ" (dora wa ichi hai kouritsu ni katsu, roughly "one dora outweighs one tile of efficiency").

Source notes: Content validated against standard Japanese riichi mahjong references and strategy literature. Strategic concepts reflect consensus from Japanese professional commentary and analytical sources.