Tenhou Stuff XVII


Reading music: https://www.youtube.com/watch?v=h0NziKrCQ-0

Tenhou screenshots approaching! 3…. 2…. 1….


I think the best plan for this hand would be Pinfu (SSK) Dora 1. However, that scenario is not super likely.

The blocks are:

666 78m 1335p 88p 23s 7s
66 678m 1335p 88p 23s 7s

I’m not sure but I think option no 1 is more favorable. Option no 2 also has 5 blocks + a lone tile but two of the blocks are pairs.  In both options 7s may not seem useful at first glance. However, it’s still a good lone tile and it gives us a chance to use Dora + SSK which is needed here. I think it’s worth keeping. For now, at least.

Im torn between dicarding 1p and 3p. I would say 1p is better because it also moves us towards Tanyao, just in case. On the other hand – 3p is not very helpful. We don’t need that pair desperately and ryankan 135 has a slightly wider ukeire.

Your thoughts?


Sanshoku is tempting here…


35 89m 24 66 789p 578s

It doesn’t take a genius here to see that aka 5s is not helpful at all, tile efficiency wise. However, in case we get 6s FIRST, our Sanshoku will be canceled and by keeping 5s now we secure that one fan we really need.

Penchan 89m, I feel, is much more useful than normally – Sanshoku chance + Dora. That’s 3 Han.

I think I would discard 2p. If we compare kanchan waits – 24 is weaker than 35 because it can develop into ryanmen with only 5p whereas 35 can get either 2 OR 6 to form a double sided wait. On the other hand – improving 35m to 56m would result in an overlapping Dora wait.

Again, I’d welcome any thoughts!


Yaaaaaay, we’re Oya. And we’re tenpai!!! We even have Pinfu Dora 1!!!! Oh, wait…. one side of our ryanmen is dead :/ Toimen seems dangerous (possible Mangan with Toitoi)…. Shimocha is fishy too… Kamicha could be dama by this point as well…. that 2s doesn’t look safe. Remaining Sha are still not visible on the table. Riichi is not an option.  No one likes to fold as Oya but I think it’s the way to play. What do you think?


I think that Chiitoi with one Aka and 8 wait is worth a Riichi. The only thing holding me back is no sign of Dora. Maybe dama is better?

By the way: one thing worth noting is tanki wait with a walled (or almost walled) neighbor. If those 7p were visible to everyone, 8p would be an even better tanki wait because if someone wanted to push a tile, they could try discarding one chance or no chance tiles.

Look at this:

Attentive players will keep in mind that 6p is kabe. That makes 7p a good wait even though one has been called by Toimen. Even more so because we can see 3x8p and 3x9p. That makes 7p very hard to use for our opponents.


Our blocks are:

2 456 8m 3 678p 34 446s

This is back to that famous 24568 shape. It’s so flexible but easy to overlook if we only focus on our immediate situation:

12 456 8
22 456 8
23456 8
2 45678
2 456 88
2 456 89

It’s even cooler here because it gives us some hope to use Dora.

Lone 3p gives us a chance for SSK 234. I think we should discard 6s. That block already includes a ryanmen + pair which is good enough.


Not that I remember every single hand ever played…. but I don’t think I’ve ever had such a hand shape. It’s really interesting!

I think both nobetan shapes should stay because they are super flexible. They can easily produce two blocks, including ryanmen waits. Just look at that:


I think that with 1m dead, keeping 2m as my pair is a good idea. On the other hand, we’re trying to get SSK here so maybe 9m discard would be better?

What do you think?

That is all for today.

May Yakuman be with you!

5 thoughts on “Tenhou Stuff XVII

  1. Congratulation for reaching 4dan! Kudos to your good attitude and effort in improving yourself this much. It is really a big difference since I last see your play like, 2 years ago? In terms of attitude at least, you ought to be a role model for many players.
    Your blogpost now focus more on stretching the basic-intermediate concept which can be a good learning/discssion place for most intermediate players. I’ll drop by more often. ^^
    1. I tend to favour pair+kanchan shape over ryankan unless i’m doing a pinfu hand. Its 2 tiles efficiency difference but gives you the flexibility to pon (which is a lot faster than chi) so I argue that in terms of speed, they are the same. If you focus on only menzen speed, like riichi-nomi, I would still prefer this shape because I rather get a 2nd ankou and increase hand value via kan if possible. In terms of defense, it is also usually better than having ryankan since you risk fewer tiles in general (discarding pair/ankou). Yes some players might still go for ryankan regardless in terms of tile efficiency but I believe in the end, it depends what you prefer. Both give very different pros & cons and can vary in terms of hand shape and situation.
    Back to your scenario. I will discard 1p hands down. Tanyao is not exactly a strong point but also don’t estimate its flexibility for pon or ankou 3p. Its slim but there’s still chance of 3ankou. For example, if you throw 6s, you will get rid of 78m, fixing your 6m ankou. At this scenario you has the flexibility to chi 4s if you see a need. And if you draw 7p (after drawing 6s/8s), your 3p becomes your pair. So for these benefits, I will argue for 1p discard over keeping the extra 2 tiles efficiency (8 vs 6). An important skill in planning your hand to note that, do prepare the scenario of your 7s forming a sequence, because that’s where the dora is. For this hand, it is the only set/block that increase the value of your hand easiest so prepare your hand as though as you know your block around 7s will be formed. When you draw 68s, you will eventually find the ryankan hinders you the most.

    2. 2p or 3m is debatable and I couldn’t come out with a good conclusion on which is better. But I will definitely force aka 5s as 1 block unless drawing in 9s first.
    In terms of my bias, I will discard 3m and force 589m as 1 block because I don’t like overlapping dora wait. With the dora suit being manzu, I just detest having 2 1sided wait blocks of manzu and will try to move away around it so that it is easier to play my hand.

    3. 4s shouldn’t be discard. It should be 3p and 2p when you had 4s. When you had 4s, you should already decide whether you are going to discard 2s or not. Don’t wait till you come to this stage. Even if the 1p is pon-ed and not kan-ed, the decision should be the same. When toimen has those melds, your priority is to simply remain tenpai till ryuukyoku or go for open tanyao but never push any dangerous tile. But to answer your question, yes, fold here.

    4. There are few ways of looking at this. 8m discard can be a defensive one because of toimen. Why would oya rush into meld 2 ryanmen? Plus, the manzu set is from 566. He must have a big hand. Ton pair or ton ankou. With 2 sets of different suit out, you must not discard pinzu anymore. I’m not sure his 7p is a lone tile pushing for ssk or not, but if is not than 8p is deadly.
    So my plan is discard 8m, damaten. Draw ankou of any pinzu or honour tile, you can consider go for honitsu depending on the situation.

    6. I think I will choose 3m discard. 9m is needed for ankou 9m or 8999m. Your hand is weak for pinfu, doraless and dora is nan so I’ll be careful here and ensure that if my hand is not big in value enough, I will not play this hand. If it end up pinfu, I will damaten. 7m comes, I will just chi. As you are approaching mid-game you have to plan for 7 turns ahead. This means expecting nan to arrives (stupid to discard dora nan for a potentially riichi-nomi or pinfu-nomi hand) so if you draw nan, you can even move towards chanta ssk dora dora. Start planning for an open hand now since this is a weak riichi hand.

    Liked by 1 person

    1. Thank you SO MUCH for you kind words! I’ve improved a lot since I read Daina Chiba’s book & started to play fast Tonpuusen. My attitude had to change as well. I’m much more confident, disciplined and willing to view losses as lessons rather than something to bring me down (which happened often in the past…).

      I have to set a good example because I teach Mahjong to people and I don’t want them to form bad habits. Good attitude is of course healthier, too (both for the mind & body).

      Teaching people is a great opportunity to review & rethink the basics!

      I’m glad you enjoy reading my posts more and more! Yeah, I’m currently trying to make my posts a sort of a follow-up to the concepts presented in Riichi Book 1. That includes stuff I’m still working on myself so I will often be wrong in my analyses but I will also learn more that way – especially if more strong players start to visit my blog & leave comments!

      Thanks for your thoughts and suggestions! Very insightful!

      See you around 🙂


  2. I read your post with Feng’s comments and there are some things he already covered in detail so I can elaborate less, but there might be some parts that I disagree as well.

    1) Discard 1p definitely. Possible tanyao/sanankou/sanshoku, less chance to give opponents to meld with it, more flexibility in late game to meld if necessary. Definitely leave some space to use dora with 7s.

    2) This was a really difficult scenario for me to explain. I narrowly choose to discard 3m over 2p because of future development of the hand. Note that it is too early to fix 6p as the pair of the hand. You definitely want to try to incorporate aka 5s into your hand. If you draw 4s/6s, it is easy. But what if you draw 5s? You can keep an overlapping pinzu set while fixing 5s as your pair. If you draw 5s at this point, you will be left with 89m 2466789p 5578s. Let’s say you draw 7m or 9s first. You can also keep this shape: 789m 2466789p 578s and then when you pair up the 5s, discard 2p for maximum tile efficiency shape. if you had kept 35m instead, you will have 6 blocks: where you have to break either 35m/66p/55s, which is not efficient. I am sorry I couldn’t be clearer because tile efficiency intuition will come from playing more hands and visualising them yourself. One more thing – the difference between a 24 set and a 35 set is negligible since only one side can be extended to a ryanmen.

    3) Agree with Feng that you should have discarded 23p earlier, and then you will end up with a 3s kanchan wait in this situation. If a tile like 2s is dangerous, try to incorporate it into your hand instead of isolating it, so you don’t have to give up completely. It’s pretty difficult to salvage this hand now, unless you somehow get to draw 2s/5s.

    4) Ok right here I don’t agree that oya must have a big hand, because tenhou players prioritise clumsy renchans all the time. But 8p might flow out naturally if you stay in silent tenpai since you can see 3 7p. There is no need to riichi since you could use this hand to end toimen’s oya quickly, and things might get disastrous if you draw a ton later or someone gets to pon ton. i don’t find 8p highly dangerous in this situation – if I draw a good honour tile, I will discard 8p and riichi.

    5) I think 678 is the most likely ssk with this hand. Dropping 6s only considers the immediate tile efficiency (tenhou.net/2/), which means that you only considered the least number of tiles to decrease shanten by 1, but did not consider what happens when at the next stage, or the eventual wait. For example, tenhou.net/2/ might tell you that a hand like this: 22389m 45677s 5789p, dropping 5p is the most efficient, but you are dependent on a penchan, instead of the more flexible 5p which can extend on both sides. I think I used this example to illustrate how useful 6s can become in this situation, even with a possible 678s ssk happening. Considering kamicha’s discards, I will discard 3p first because if it gets stranded too long, it might become dangerous later. You don’t want to get tenpai, and then discard 3p and furikomi to oya and end up in the same situation as shimocha.

    6) I was prepared to discard 6s to commit to junchan ssk since many suozi around 6s have been discarded. Then I realised that 1m is completely dead, making the 23m ryanmen useless. So discarding 3m would be a better choice. Right now the shape can accommodate both 678 and 789 ssk depending on if you draw 6m or 9m first. If you turn 6s or 6p into a new taatsu, you can also fix 789 ssk and discard your redundant set. It is possible that the hand might evolve in an unpredictable way and some how you get 678 ssk with tanyao. Adapt to the situation and draws and do not commit the hand to a weak riichi nomi.

    Liked by 1 person

    1. Thanks for another comment! Especially to situation no 2. Good explanation.

      Visualizing my potential shapes and draws is still something I don’t do intuitively. I really need to focus to do it capably & I miss a lot of stuff. I will keep working on that, though!

      Situation no 5 – I still sometimes keep lone tiles instead of those not-so-useful-at-the-moment tiles in shapes which can actually evolve into something great. Gotta change that!

      See you in the future posts! 🙂


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s