Application of Block Theory – Part 2
So you’ve got five blocks now, the right number of blocks to form a winning hand, but you’re still not tenpai yet. What do you do? Does it even matter? It certainly does – read on to find out more.
1. Followup tiles
These are tiles that strengthen one of your blocks, such as the in
The shape waits on to form a complete block.
Some additional things to note:
If you have a choice between different followup tiles, keep the one that strengthens your weakest block. For example, with the following hand
you should discard the rather than . This is because waits on six tiles as opposed to four for , which is a 50% increase in speed. On the other hand, waits on ten tiles as opposed to eight for , which is only a 25% increase.
Remember that your weakest block will become the bottleneck of your hand.
Consider whether or not you can incorporate dora or akadora. For example, with the following hand
you should discard instead of to ensure that you can make use of the .
2a. Safe tiles
If you can’t strengthen one of your blocks, the next best thing to do is to hold onto tiles that will be safe to discard later. The safer the better – ideally this would be a tile that all three other players have already discarded, or more commonly an honor tile that has been discarded at least once or twice. If neither is available, prioritise holding onto tiles that are at least safe against the dealer, since dealing into them hurts the most.
Holding a safe tile provides you with several benefits. The most obvious benefit is that if another player happens to declare riichi before you, you can safely avoid their ippatsu and buy some time to decide whether you should defend or if you can keep attacking. The second benefit is that, if you get to tenpai first and declare riichi by discarding the safe tile, you can avoid giving useful information to your opponents who may be wary of suji traps or matagi suji.
A caveat – notice that safe tiles are ranked 2a. If your hand already has good shapes, with your blocks clearly defined and unlikely to change, or if it’s getting late in the round and other players are likely to be nearing tenpai (i.e. about the tenth turn and onward) then you should keep safe tiles. If not…
2b. Floating tiles
If it’s still early in the round and you have two or more weak blocks, then you should consider holding onto floating tiles in the hope that you can switch out one of your weak blocks for a stronger one. For example, with this hand
Naturally, the same ranking for floating tiles described in part one of this series still applies.
3. Tiles that are both useless and dangerous
This shouldn’t need any explanation. If you’re not going to use them and they’re not safe to hold onto, then just get rid of these ticking timebombs as soon as you can. This isn’t quite sakigiri, but when people talk about sakigiri being useful, they’re probably talking about this.
It’s time to test your knowledge! Take a look at the following hands and pick the best discards according to the rules set out in these articles. Post your answers in the comments!
1st one is easy haku. Dora is dora but a good shape is a good shape. With high chance of pinfu and one akadora we’re loving life, we might even get a red 5m to boost our value.
2nd one is probably 1p. 122p isn’t that weak a block because of the 6 ukeire (2p, 3p) but all the 3p are overlapping so you pretty much only have 2p as ukeire, which you also retain after discarding 1p. Furthermore you keep a floating tile in the souzu area in case you can improve a wait (an 8m also helps us improve at which point we’d probably drop 8s)
for 3rd 113 block is decently strong with 1m and 2m ukeire and could in theory turn ryanmen with a 4m as could 2p with a 3p. So it’s a choice between 5s and 7z. Now if we’re really far behind and need to keep our hopes alive for ittsuu and the extra points (since it’s E4 we could already be in trouble or if it’s tonpuusen that’s even worse) we’d cut 7z here, else it’s probably better to cut the 5s fast before it gets too hot. All its ukeire for improving to ryanmen are overlapping after all.
4th is a classic situation of “I want to change that penchan”. So we need to cut the worst floating tile which is the 2s, half of whose ukeire turn it into another penchan and eventual death X_X
5th we cut 2s. Even in the case we split the 2445 block into two (24 and 45) we have too many blocks and we need to cut the weakest one which would still be the kanchan 24s with only 4 ukeire compared to 67m with 7 ukeire 45s with 8 and 688p with 6.
6th the nakabukure of 8s isn’t good and will in most likelyhood not improve the hand in any imaginable way so it’s the best thing to cut. The first cut here is easy in that aspect but the future of this hand is more interesting. If the floating 6p improves we look at the manzu as one block of 677m and 4m goes but if we improve with 5m then we have the option of looking at 77m and 99p as two blocks instead. Depending on point situtation and improvement on the souzu tiles towards ittsuu again, we might end up breaking a pair (specifically 7m which would be more dangerous for later)
7th is tricky because the hand can push towards chiitoitsu which can be a guaranteed mangan on riichi/tsumo with tanyao as opposed to riichi pinfu tanyao which will fall shorter, especially on ron. So depending on the point situation perhaps 3p or 7p can be one of the best choices if we’re going chiitoi. If we’re pushing normally we should break one of the pairs in the pinzu where our hand is almighty (probably 6p to keep the sanmenchan shape that can take more tanyao tiles though we’ll end up discarding 2p eventually anyway in all likelyhood) and hoping to improve our shape on the souzu. Riichi on the kanchan for iipeikou might not be a bad move since riichi iipeikou tanyao is stronger than riichi pinfu tanyao and almost mangan on a tsumo.
8th is pretty simple. We’re in south so the winds are strong and they can stay. Including that we have too many blocks. The weakest goes. Since we have ryanmen everywhere we cut 12s starting with 2s.
9th Chiitoitsu this time is much closer with this hand so we’re pushing for that. Given that all free tiles except for the dora are of equal value discarding either of them is an alright choice. However setting a suji trap on 8p is pretty useful so an early 5p discard might net you the ron you need later.
1oth and last is also very tricky. We have two dora already so if we can get speed that would be ideal. Now let’s calculate our options. for maximum effeciency/ ukeire 2 or 5m go. That might bring the hand into a very bad tenpai shape however which might not be ideal. We also have 4 pairs so we could sacrifice a bit of effeciency with 6p to keep chiitoitsu alive. which, with 2 dora is guaranteed mangan and haneman on tsumo. But there’s also another option. Cutting 9s in order to retain the ability to open up for a possible tanyao. We lose chiitoitsu and the 9s ukeire but we can improve our other shapes and call in multiple scenarios to advance. I think the choices in this one depend mostly on one’s playstyle and on the board/ opponents.
Haku. Dora is nice, but it cuts your hand’s potential tiles too much.
8s. While the potential for a 3-sided wait is nice, the chance of drawing the 7s is simply too low to drop anything else in the hand. Dropping the 1p decreases tile acceptance in the event you draw a 3p (and removes pinfu from that particular equation) as well as removing a fairly safe tile that you could discard later if someone makes a move.
5s if chun have been discarded, chun otherwise. Chun is likely to become a thorn in your side if none are out yet, and the 5s could either lead to ittsu potential or an extra dora. If chun have already been discarded, the 5s becomes notably less appealing since yes, it is a center tile that is also pretty likely to become a time bomb in short order.
6s. It’s E1 (early E1 at that), pushing for value over speed here could pay off considerably in the long run. Discarding 2s would remove sanshoku potential; 3m would remove iipeikou and an extra dora potential.
2s. Everything else is too well-connected.
6p. With three pairs, chiitoi is a very real possibility (especially given what the dora is).
Tough call, but I think 7p is the loser here. 5 pairs makes chiitoi highly likely to say the least, and the 7p makes it easier to get back into tenpai in the event the sou develops improbably.
2s. Most dangerous tile of the weakest block of a six-block hand. Not much more to say here.
8p. This hand is very likely going to be chiitoi, which makes it a tough call, but in the event that it isn’t (and you somehow still need a set after that), the 8p overlaps with the 5p slightly, whereas the 8s does not.
2m. Given the dora, and the four pairs, chiitoi is a very likely possibility. This said, cutting anything else would weaken the not-chiitoi potential of this hand too much to consider. 6p removes the ability of the pin block(s) to complete on 2345 (or 6, taking into account chiitoi). 8s should be obvious (it removes four tiles from the six that would complete that block.
Hopefully I’ve gotten them in order.
I’m assuming we are playing on Tenhou against 7d opponents for my answers:
1. 9s. We want pinfu or haku there for value. Sacrificing 7s acceptance for haku is easily worth it.
2. 8s. We can keep sanshoku, tanyao, and pinfu potential without the 8s.
3. 5s. It doesn’t help much, and it’s getting late.
4. 2s. Sanshoku is kinda far away. Riichi dora 1 is an acceptable hand.
5. 2s/4s. Only need one set from the souzu.
6. 8s, because of tile acceptance overlap between 135s and 8s.
7. 2p. Chiitoi looks tempting but it’s much slower than regular hand here.
8. 5m. It’s a bit of a sakigiri. We have three pairs so the loss is just two tiles here. If we draw 3s we get pinfu.
9. 8p, because of tile acceptance overlap between 5p and 8p.
10. 9s. This is a really slow 2-shanten hand. 9s is an attempt to move toward tanyao. The 2345m shape can easily improve the hand. Even with bad shapes for regular hand, chiitoi is still too slow here.
Haku and dora even it is maybe to late if nobody already discard one.
1p for tanyao. And with224p, 1p is useless. Of course 4p can also be use with the 5p
Chun or 5s. I hesitate. I will discard chun. The 5s can always be use with 5s red.
2s. I think sanshoku is two far. Between 2s and 6s, 6 is far better.
2s or 4s. But I want to keep the 5 red and later 4s will be more dangerous to discard, so I discard 4s
I discard 8s. It is not a good block to split in 2, it will create a penchan.
Keep 4m and 6p is good for the missing block.
This one is very difficult for me.
If I want to push chiitoisu, I should keep the 7p because it will be the best wait for the last pair.
But! I don’t know that else to discard.
For normal yaku and not 7 pairs, I may discard 2s because to much pairs.
But I think in the end, I may discard 7p and try to have sanankou .
6 blocks. The weakest is 1 and 2s. So I discard 2s
Chiitoisu may be the solution. It is better to keep tiles which are easily discarded by opponent, like 8
I will discard 5p to eventually prepare suji trap and because later I will be a dangerous tile.
I’m not sure between 2m and 9s to push tanyao.
As we are Est3, I think I will discard 2m
After writing my point of view, I check the answers of the three others. It seems that it is not easy. We have quite different way of playing. That may be the reason why your works are so useful. Thank you for this.