Seven pairs or a standard hand?
When a hand can go either way, shanten is only the beginning. Comparing the paths with tile acceptance and expected draws.
Seven pairs or a standard hand?
How should we handle a hand that can develop into chiitoitsu (seven pairs)? A lower shanten number is useful, but the shape of the hand and its useful draws matter too.
I have been practising tile efficiency and my AI reviews show fewer severe mistakes. The standard-hand-versus-chiitoitsu decision is still a regular blind spot, so this post works through one concrete shape.
A useful heuristic
Riichi Book 1, section 5.5.2, gives a good rule of thumb: when a hand contains many side waits, the standard hand is often preferable to seven pairs.
456 56m 45 88p 667s NN
This is 2-shanten standard but 1-shanten chiitoitsu after discarding an isolated tile. The standard hand also has routes to pinfu and iipeiko.
Counting tile acceptance
For chiitoitsu, three single tiles can each become a pair: roughly 3 × 3 = 9 accepted draws. Tenpai has roughly three tiles in its lone pair wait.
For the standard hand:
37m · 36p · 58s → 4 × 6 = 24
8p6s → 2 × 2 = 4
Visible tiles change the exact count. The progression is:
| Stage | Chiitoitsu | Standard hand |
|---|---|---|
| Current hand | 9 | 28 |
| 1-shanten | 9 | 16 |
| Tenpai | 3 | 8 |
The extra standard-hand improvements are omitted, so this comparison is conservative.
Turning acceptance into turns
If N unknown tiles remain and T improve the hand, the expected draws until the first improvement are approximately (N + 1) / (T + 1).
| Path | Useful tiles | Expected draws (121 left) |
|---|---|---|
| Chiitoitsu 1-shanten | 9 | 12.2 |
| Standard 1-shanten | 16 | 7.2 |
| Standard improvement | 28 | 4.2 |
The standard hand reaches an improvement about eight draws sooner here. Ukeire is doing the real work; value, defence, dora, and yaku still belong in the final decision.