Tile Theory at One-Shanten (6)
Kuttsuki one-shanten has two major characteristics:
- it has very wide acceptance
- it tends to reach tenpai on bad waits
Example 1
Tsumo 
=> Discard
| Discard | Tenpai chances | Acceptance count |
|---|---|---|
|
   |
31 (11 good-shape tenpai) |
|
    |
33 (11 good-shape tenpai) |
|
|
38 (16 good-shape tenpai) |
Example 1 is the simplest possible kuttsuki one-shanten.
For a plain kuttsuki shape, you still keep tiles according to the normal tile-theory priority:
3-7 > 2/8 > 1/9
So discarding is straightforward here.
Even the apparently inefficient choices, discarding or , still have more than 30 tiles of acceptance.
Compared with a one-shanten made of two ryanmen, which has only 4 types and 16 tiles, this shows just how fast a kuttsuki one-shanten reaches tenpai.
But for a simple kuttsuki shape like Example 1, the number of tiles that actually lead to a good-shaped tenpai is small.
Even the best discard, , has only 16 such tiles.
So for kuttsuki shapes, the count of tiles that reach a good-shaped tenpai should be emphasized.
Once the acceptance to tenpai is around 30 tiles, that is already wide enough.
The difference between 30 and 31 hardly matters in practice.
Kuttsuki shapes are one of the few situations where it is reasonable to care more about the final wait at tenpai, or even the yaku.
Example 2
Tsumo 
=> Discard
Example 2 is a classic case. The tenpai chances are:
| Discard | Tenpai chances | Acceptance count |
|---|---|---|
|
 |
43 (16 good-shape tenpai) |
|
 |
37 (22 good-shape tenpai) |
|
|
40 (22 good-shape tenpai) |
If you follow the principle of “prioritize tenpai chances,” you would discard .
But if you compare the count of good-shaped tenpai, it reverses:
Discard 3s = Discard 3p > Discard 3m
So the correct play here is to discard .
And why is discarding better than discarding ?
First, its total acceptance is already larger.
Second, it has the extra sanmenchan improvement when you draw .
All the candidates in Example 2 are kuttsuki shapes based on the tile 3.
Remember this rule:
if the number tile is the same, the side that already forms a continuous shape is stronger overall than the isolated one.
Example 3
Tsumo
=> Discard
| Discard | Tenpai chances | Acceptance count |
|---|---|---|
|
|
50 (26 good-shape tenpai) |
|
 |
46 (29 good-shape tenpai) |
|
|
38 (24 good-shape tenpai) |
The connected souzu shape is the strongest kind of four-tile run, so it must be kept.
The real close comparison is between discarding and discarding .
If you look only at total acceptance, 50 versus 46 looks like a big gap.
But for a kuttsuki shape, 46 tiles is already more than wide enough.
So what matters more is the count of good-shaped tenpai.
Discarding gives 29 tiles that reach a good tenpai, so that is the recommended play.
The ease with which a middle-bulge shape turns into ryanmen is still very attractive.
The relative strength of continuous shapes has already been explained earlier, so I will not list every pattern again here.
Theory Summary
Because kuttsuki one-shanten is already wide enough in raw acceptance, the most important comparison point is the number of tiles that reach a good-shaped tenpai.
If a kuttsuki shape reaches tenpai on a bad wait, not taking tenpai can also be a strong option.
For example:
Tsumo
That shanpon wait is far too unsatisfying.
Kuttsuki shapes have so many tenpai tiles that even if you break tenpai, there is still a good chance to recover quickly in the next few draws.
So here you should simply tsumogiri and wait for a better draw afterward.
Original Japanese page: http://beginners.biz/pairi/pairi18.html