WWYD 2012/05/30

[Shirluban]

At first sight, I'll say .

[Iapetus]

here... I really don't like waiting on those dora-area kanchans, and it does have the least backfires. 8sou enables 347sou to help the hand.

Edit: Oh, the 34 do help without the 8, I misthought. I see the reasoning.

[WaveMaster]

We have a somewhat strange pond here considering our hand. I have to wonder how this hand progressed.

Anyway, there aren't any really excellent choices here. I don't want to get rid of any of the dots right now since they all have the chance of connecting to the dora in an open tanyao hand. The backfire if we cut 3p and later draw 4p would be annoying in this hand.

Cutting 8s backfires if we draw 7s or 8s. Drawing 7s is not any more or less likely than drawing 4p, and the backfire would be worse with 4p because we lose a chance to form a set containing a dora. If we draw another 8s, it's more annoying, but I think it's worth keeping our options around the dora open.

There is no backfire on 3s or 4s when cutting 8s. If the [222-4]s shape evolves into a set and a pair, then the middle wait we would have had with [6-8]s is no more likely to complete than either of the two middle waits that contain the dora tiles, but is less valuable. The hand can proceed to something like:

[Kyuu]

To be honest. I don't know.

At first look. This looks to become a Tanyao + 3 dora hand minimum. Already, that's a nice hefty shape to work with. Add in a pair of ankou of 3-man and 2-sou. Then keeping the pair of 5-pin as the dora. That brings about the possibility of san anko, or at the very least, a fast toi toi attainable just by one call and a couple good draws. Seeing a 2-pin and 3-pin already in the discard.

I'll discard.



Place a greater focus on the 4-6-8 sou Ryankan. For me, this is a new word.

Too bad a 6-sou is in the discard too. Like wise, with 5-pin as the dora. It is within expectation of the other players holding onto tiles around that 5-pin. With a pair of 5-pin in possession, it is not as important to even wait on a 4-6 pin. This also leaves the open option to pon the 5-pin later on.

[Senechal]

I've tried to look at this hand with an openness to all solutions... and I disagree with most of them.

Masayoshi Ara's tool places high scores on 3p, then 7p and 4m, but with three other results with outcomes that aren't bad at all. (They all give 6.5 to 8.0 % chance of winning the hand)

So I'm going to say that my choice is within those other 3 choices. Why 5p?

* Tanyao dora-dora confirmed. Still room for extra points with riichi, or a 2s or 3m kan. While a haneman score would be possible, it isn't probable. No point in playing tight for low gain potential.
* The hand is 2-shanten. The one thing I do not want to get caught with are dora side tiles to finish not one but two proto-groups.
* A hand only needs one pair. Straddle waits already contain implicit pairs. With two already in hand, I can't see any logic behind making an explicit pair in the hand.
* A hand has to have an expectation to get to tenpai, and then an expectation to win. I do not believe cutting any (edit: souzu) or 4m really contribute to either.
* While 8s by itself has the least usefulness, I'm pretty sure the goal of the souzu tiles in the hand are to create a six tile end result. If 3s is picked, there's a chance that it becomes a 5 tile arrangement, but to bet on it is ludicrous. While the shape is good for a 7 tile arrangement if 7s is picked before cutting other souzu, it's not needed because of the uncentered straddle wait compared to the centered straddle in manzu. Totally disagreeing with WaveMaster who produced a 3-shanten solution for a 2-shanten hand.
* Seven draws left doesn't give you any time to play fancy. Other protip: playing fancy is -EV.

As for this specific metagame, it's East 2, 5200 or 6400 is more than sufficient to take the lead, whether open or closed (silent, btw!)

tl;dr:

other precision: 3900 is a possible floor. I was aware of it at the moment of writing the text, but the chances and methods for gaining fu points are clearly available. 5200 without a kan is still possible... if my reasoning is not to your liking, listen to xkime. Fully endorse his reasoning for everything.

[WaveMaster]

Totally disagreeing with WaveMaster who produced a 3-shanten solution for a 2-shanten hand.

The hand was 2-shanten before drawing the red 5p, and it will still be 2-shanten no matter what tile we discard. There was nothing 3-shanten about the possible solution I described. You can see for yourself that the only tiles I added to complete the hand are 3s, 4p, and 6p. Therefore, only two tiles were necessary to reach tenpai.

[xKime]

8s is too optimistic. The realistic discard to win this hand is

At first sight, 8s looks a little bit wider, but then again you have to look at the -real- number of available tiles and their position (especially the ones near the dora). Discarding 3s would be the, in reality, wider discard and takes away a lot of the stress you put into either a 2p that's as well as dead, or the dora indicator 4p which only slows you down. Not to mention that with any other discard you'd also stress the kanchan 6p as well. A more natural and winnable solution is to discard 3p and see what you will draw next. (If 1p and 2p were at least alive, and the dora was 2s or something else, it'd be a completely different story)

3p is quite certainly the worse tile in this hand. As to where to go afterwards, it depends on your next draw. The one I'd like the least, but which would still be pretty good, is drawing 9p or 6m.

Also, what's there not to like about a hand with three dora? You should try to win it the easier and fastest way, not to stylize it or putting stress into areas that are not worth it, or destroying your only good shapes.

If anything, I'd say those 2p/3p he discarded became a waste...
Not like he could know what he'd draw red 5p next, but he could at least have been ii shan ten earlier.

I wouldn't blame occult players, they can select anything by instinct or "read of the flow" or whatever, but digitally I can only see 3p.

[Senechal]

edit: confused Wavemaster's example final result with a hand as it would properly develop. Can you space out a 14th tile or something when you do that ?

Still find it ludicrous. You're still planning on filling three closed waits to finish your hand.
A shanpon wait with a dora pair makes more sense than that. Barely.

[xKime]

I know, Sen. He seems to believe that we need to turn the two dora tiles into two mentsu to use them, when in fact even using those dora as a pair (which this hand shape technically lacked) would be far more efficient. He seems scared that we would have to discard a dora at some point, but the only situation we'd do that if we discard 3p, is at the time we draw 6p for tenpai (if we draw it before tenpai, we can choose not to discard 5p), but even if that happened we can make up for it by reaching.

8p doesn't even make it easier to call tiles, nor does it make it easier for menzen, nor does it make it easier to actually keep both dora.

Sen, I actually like your reasoning for 5p. It does make for a pretty good alternative discard, except at the time you have to discard either 3p or 7p for tenpai, because the efficiency of that tenpai is lowered as much as the value was lowered, too. Also, correcting an inconsistency in your original post, by the time you open this hand, it will be worth 3900 and not 5200, and you'd need to tsumo that kanchan if you want it to be 5200. I don't know if you intended it that way, but it kinda sounded like you'd call for a 5200+ tenpai, when in fact the value drops to half a mangan at least 3 out of 4 times (if you plan to kan tiles, now that's a different story! w).

Still find it ludicrous. You're still planning on filling three closed waits to finish your hand.

Ludicrous is the right word!

[WaveMaster]

edit: confused Wavemaster's example final result with a hand as it would properly develop. Can you space out a 14th tile or something when you do that ?

Still find it ludicrous. You're still planning on filling three closed waits to finish your hand.
A shanpon wait with a dora pair makes more sense than that. Barely.

You're going off on some weird assumption. I was responding to the post directly above mine which claimed that discarding 8s would backfire if you later drew 3s or 4s. This is not the case, as the poster later noted in an edit. If you discard 8s and then draw 3s or 4s, your hand increases from 2-shanten to 1-shanten (by discarding either 4m or 6s). I demonstrated this by showing one possible hand where 8s was discarded, 3s was added, and the hand was completed with only 2 additional tiles (4p and 6p). A discussion about the shanten count of a hand has nothing to do with likelihoods or whether you draw the tiles yourself or not. I showed a complete hand without a drawn tile because I was worried readers might otherwise misunderstand.