Early-Hand Tile Theory (2)
Last time, the examples were hands where the answer could be found just by comparing floating tiles.
This page looks at two different patterns:
- cases where you compare taatsu against taatsu
- cases where you compare isolated tiles against taatsu
1. Six Taatsu and One Head
In this pattern, you simply compare the taatsu.
Example 1
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
Tsumo
=> Cut ![]()
Following the standard rule for taatsu comparison, you remove the penchan first.
If you later draw
, then even though it creates furiten,
dismantling
![]()
would still be better than cutting from the inside first, so discarding from the outside is slightly more favorable.
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
Tsumo ![]()
Example 2
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
Tsumo
=> Cut
or ![]()
All the blocks here are ryanmen taatsu, so you should remove the souzu taatsu that has become a nido-uke shape.
You do not drop ![]()
because of the Sanshoku possibility, and also because as a final wait,
-
is superior to
-
.
What you must notice here is that cutting
would send the hand back to four-shanten.
If you break your only head, the shanten count will definitely increase.
So even if Tanyao crosses your mind, cutting
from Example 2 is still a bad move.
Example 3
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
Tsumo
=> Cut ![]()
In Example 3, all three suits contain a nido-uke taatsu.
If you ask yourself:
“When I cut this tile, which draws become a loss?”
the right answer should narrow itself down.
The manzu block is a ryanmen taatsu, so of course you keep it.
That leaves pinzu and souzu to compare. Since keeping the
acceptance is better, the correct play is to tsumogiri
.
Theory Summary
2. Five Taatsu Plus a Floating Tile
In principle, you simply compare the remaining isolated floating tiles.
Example 4
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
Tsumo
=> Cut ![]()
If you treat the manzu section as a block that must still produce two complete sets, then
has almost no reason to remain in the hand.
is worth keeping because it still has follow-up value on draws of
![]()
.
If you choose to cut
here, being accused of “Tanyao disease” would be fair enough.
That choice gives up two lower-value ways to keep Tanyao available, and trying to rush into an open Tanyao hand when you do not even have a completed set yet is also quite dangerous.
Example 5
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
Tsumo
=> Cut ![]()
By simple part count, the hand has five blocks, but the penchan-kanchan nido-uke shape around
is just too painful.
When there is a taatsu this bad, keeping the isolated floating tile instead is entirely possible.
Example 6
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
Tsumo
=> Cut
or ![]()
In Example 6, the pinzu block is triply blocked by
![]()
![]()
,
so the penchan has extremely little value.
In a case like this, keeping the floating tile often leads to smoother development afterward.
Trying to force Ittsuu in pinzu is not realistic, because the useful draws are far too restricted.
Theory Summary
Original Japanese page: http://beginners.biz/pairi/pairi09.html