Pair Theory (Part 2)

We continue this week with xKime’s translation of Puyo’s blogs. We also have professional input from JPML’s Garthe Nelson, adding to the depth of the conversation. Hopefully some of these issues will bring some good conversation to the forums as well.From: http://blog.jpmahjong.net/read.php/517.htm

The discussion is whether we should go for All Pairs (Chii Toi Tsu) or whether we have excessive pairs in our hand. Generally speaking, the borderline is “two pairs”, with any more than that the hand is likely to become bad shape. If you have too many pairs in your hand, there are a few choices:

a. Discarding them directly.
b. Combining those pairs with existing tiles in the hand to form a better shape.
c. Going for an All Pairs hand.

Many times beginners are reluctant to throw away excessive pairs, for example:
3c4c7c8c9c1d1d3d3d7d3b7bwestwest
(west is a non-value wind tile (otakaze))
Discarding the lone 7 of dots, 3 of bamboos or 7 of bamboos are all choices commonly seen in beginners, but I think it’s better to man up and discard 1 or 3 of dots.

As far as a loss is respected, discarding 3 of dots only becomes a loss if you draw one of the other two remaining 3 of dots. But if you discard 3 of bamboos, the losses are 1, 2 and 4 of bamboos; 12 tiles total. In a situation where you don’t have enough incomplete groups (taatsu), the possibilities for pairs to become groups or turning into a winning hand shape are not very good.

In situations where you have 2-3 pairs, if it’s still early in the hand, it’s very possible to think about discarding one of them, leaving inside the hand possibilities for isolated tiles, value tiles, other hand points (yaku), good shape or to increase the possibility of completing groups (mentsu)… This is not only a mistake beginners make, even more adept mahjong players fail to deal with this blind spot from time to time.

As for hands with 4 pairs or more, generally speaking it is possible to keep All Pairs within your field of vision.

4c4c9c9c1b2b7b8b9b3d7b3dwest8c
draw: west9b
dora: west3d

This is from an actual MFC hand history, sitting North. If you draw 9 of bamboos in the 4th turn, what would you discard?
For a Group Hand(i.e. 4 groups +1 pair, the standard winning shape) the most efficient discard would naturally be 9 of bamboos.
In appearance, since this hand has many pairs it looks like it should be easy to form a group hand, but it isn’t. At the time we only have the 789 completed run (shuntsu). We still need one of the remaining two 4 and 9 of cracks (manzu), 3 of dots (pinzu) and West Wind (shaa) to have a chance at winning the hand. There are only two tiles remaining of each, so the tile efficiency here is really low. The most important thing is, we need the pairs to form a good shape, but that’s not easy either.

Therefore in this hand we must go for All Pairs; in this actual game, I discarded 2 of bamboos (souzu).

————————————————————————————————————————————————–

Comments from Garthe
Ah yes, another age old conundrum, 7 Pairs or normal shape. I remember running into it a few years ago in my last game at the OUI tournament. I made the wrong decision, at least regarding what I eventually drew. I opted for 7 Pairs as I tend to do, the tiles I drew afterward would have completed a Simples, Peace, Dora hand.

Distraught that I had knocked myself out of the tournament I looked for support so I could blame the loss on bad luck. I wasn’t so lucky. Both Rumi (who was at the same table that game) and Jenn were surprised that I chose that road, and Jenn’s reason for why the normal shape was so obviously the right choice has stuck with me since that time: if the waits are good, take the easy road, go normal!

I have to admit, as I play more and more, one of the things I’ve noticed on days that I can’t seem to win a game is that I’m often going for a lot of 7 Pairs hands. Sometimes I remember Jenn’s advice at a crucial moment and manage to go normal instead of taking the hard way. But more often than not, I keep my pairs.

I won’t go into a long explanation of why the approach may or may not be justified, especially as it might just be unjustifiable. But perhaps I can still provide a little method to my madness.

The article says that the only loss in discarding 3-dots in the first example is drawing one of the other two 3-dots. First of all, I would argue that that is in fact a pretty big loss because that would have completed a group moving us to one away from ready. 1,2, 4 of bamboos team up to only form a nice potential shape for a group, and we’re still two away from ready.

Second of all, that reasoning also assumes that we’re really no longer aiming for anything but Peace. I would argue that if 7 Pairs is still on the table, a draw which pairs ANYthing else in our hand is also a loss, seven singletons at 3 remaining each gives a loss of 21 tiles. And finally, if we’re ok with just the fastest route to ready then completing the other triples is even something of a loss as we’d still be two away from ready rather than one, four more losses.

I tend to go that route because I dream of hitting monsters. 7 Pairs can become a potential monster with just the addition of one Dora, haneman if self drawn after reaching. If the pairs start tripling up 3 Concealed Triples is usually at least mangan because Reach and self draw are in there as well. And of course if the triples keep coming to the end, yakuman (limit hand)!! 4 Concealed Triples is my favorite hand.

I can appreciate the reasoning for breaking the pair there, but I would keep it and throw one of the bamboos.

One last note, many people thought Kosho Tsuchida was perhaps the best player in Japan until he left our association to start his own and we had to start thinking someone else must be the best player in Japan. Anyway, he was famous for leaving pairs in his hand when no one else would have. What does he know that we don’t?

Well, time to get lucky…

RM on Social
error