Reachmahjong.com

I would like to make sure that I haven't got myself into a rut (due to habit) on what to discard in these very common situations where there seems to be more than one decent possibility, but maybe not. Is there always one tile that is better than all of the others or can two (or more) have equal discard potential?
I am trying to play a faster game but occasionally. I just get stopped cold and have to think (and naturally, run out of time, if there's a limit).
When I win the hand, I tend to think I made the right choice, and if I lose, I tend to think I made the wrong one, but that's not necessarily true. Something called luck. How would you approach these situations?

(1) (2) (3)

1: Without having the Tenhou calculator at my disposal, it would be either the 4m or 7s. The Tenhou calculator does confirm 4s as being better than any other choice, but 3s or 6s are better than 7s. What you can't do is drop 5m or any pin tiles.

By remembering that every winning hand has only one explicit pair, by chucking 4m, I see three suits of 2 + 5 + 6. Considering that you are dealing, and your lead is small versus South, and that 4--7m tiles are to be very dangerous the longer you wait, I'd just go for a silent wait if you can pick up the tiles yourself, or failing that, if 4p comes out of North's hand, call it and wait on 2--5--8s.

There are a lot of what-ifs that can be raised, but versus a hinted flush hand, get rid of the 4m now, and if you get 7m dora later, sit on it. The rest is secondary.

2: I'd chuck 9s myself, but that's wrong. A hand only needs one pair: best choices to discard are 4/7p. 4p is already down once, so that is that.

3: 9m is your pair, 9s is a set, one open-ended wait and two gut waits. One is dora, so pop the other.

Thanks. It was helpful. A few more thoughts.

(1) I remember deciding to use the 7s as the pair so I chose the 3s as my discard. As an alternative discard, I strongly considered tossing the 5m instead of the 4m (in case I draw, or can call, the 3m or 6m) but you state this should not be done. Not sure why not. It seems more efficient. It's not???

(2) I did chuck the 9s, and eventually the other 9s too. Was thinking tanyao, or riichi pinfu, or both. But you're right. Discard 4p/7p. Forget tanyao. Got it.

(3) If by the other, you mean the 5p,7p, that's what I did.

Thanks again. By the way, what's a tenhou calculator?

http://tenhou.net/2/?q=567m233778p45699s

As for your feedback for #1, your hand has everything connected and "able to be used". Or 6 potential taatsu and other incomplete groups (protogroups). For everything in pin and sou, you are one tile from solidifying two groups, although if you get 2p/6p, then your hand is moving sideways rather than forward.

Back to the protogroups: if you have the potential for 6 protogroups (i.e. no gaps anywhere), then you have to take one apart. If you get something like 2/5/7/8s, you solidify three protogroups into two actual groups. That leaves two more groups and a pair to have and three protogroups in hand. Good, now which one to force as a pair? Picking up 2/6p with that hand gives 1--4p or 4--7p options to go out. Picking up 4p naturally closes the groups.

That's why the 455m group is the least useful: it's a really good shape to have to build a hand, but when the rest of the hand can compete with it with multiple tiles and options, then it isn't that hot anymore.

Understood. Good explanation. Thanks. I guess I was too quick to decide on the 7s for the pair.

That calculator thing is a pretty handy tool. You plug in a hand and it gives you the possible discards in order of preference and the number of good tiles you can draw, I think.
As Will Smith said in the movie Independence Day, "I HAVE GOT TO GET ME ONE OF THESE!!"
(after I get it translated so I can better understand exactly how to use it)
I wonder, is that how some of those detailed answers for those really tough WWYD's arise?
Seems like cheating, a little, I guess....sort of like using a word descrambler in those online Scrabble games. Well, maybe not as bad as that. You still have to understand and be able to explain what you're talking about.

Which is why if I would do something wrong, I would tell you about it honestly, exposing proper insight into the decision-making process rather than BS you with "Senechal knows best"-type answers.

If the goal is to rationalize specific situations to a general rule enabling you to discard within 3 seconds on tough hands, it is only with an explanation or your own discovery that it will happen.

1.

Note that discarding 6s would leave a chance for 345 sanshoku. It's not worth the loss of speed though.

I initally was going to discard 7s. That's because instead of giving the hand an extra 4 tiles to a bad wait tenpai, I'd like to use the 4m for a chance to improve the shape. For example, with a hand like
,
the 4m lets me draw 35p to replace the 4p wait with a ryanmen on 36m. But in this hand that's meaningless, since adding 3 or 5 to the 33455 gives a good shape without any extra tiles necessary, plus there's less motivation to deconstruct the iipeiko.

For the first one, I would definitely go for as the outright correct choice. (there is a three sided wait which is more ideal so there's a clear preference for to (and to discard) ). Sometimes you have to ignore tenpai efficiency in the short term to increase your chance of winning in the long term and this is one case I definitely recommend doing so. The key to me thinking so is the shape, you definitely want to avoid a kan- wait. when I look at this hand, I look at where I want a jyantou (pair) and where I want my mentsu (sequence or triplet) and is a very good jyantou + mentsu shape and that's where i prefer to have my jyantou and by doing that avoid the kan- wait. The clear advantage of over (jyantou kotei) is the flexibility in approach and the likelihood you'll end up with a good wait at tenpai. After the discard you end up with , from this if you draw you get a very good tanyao pinfu 1-shanten. the only draw back to this is that you lose the tenpai from or draw to get a wait. Comparing the two scenarios ( and ), with there are 22 tiles that give you a very good tanyao guaranteed 1-shanten for you hand

or with discard, 4 more tiles that you can get tenpai with or 8 tiles ( (not tanyao guaranteed) and ) to give you a double sided wait on the shape. for both cases gives you the same tenpai regardless of choice.

As you can see, by seeking for the highest tenpai probability in the short term, your missing out on alot of tiles that can give you a very good 1-shanten and probably a better winning probability in the long term. The only case I recommend seeking for the highest tenpai probability is when it's very late into the game and you need tenpai for the draw at the end or you think that you haven't got enough draws left to take advantage of the better 1-shanten.

Second one is definitely or , preferred due to one already being out and it being more dangerous as is dora so you dont deal in to shapes like +pair or / ( also toimen has a fast discard which 'usually' means higher probability for a or wait.) discard can also be argued because is east's safe tile. discarding and going for tanyao is slowing your hand down as well as lowering your points potential, riichi pinfu's average score is close to around 3200 and kuitan currently is 1000 points guaranteed.

Third one I will throw then , the order is specific because if you get a red next draw it's better to get a pair with red then a again with a red .

sorry for the long explanation and if you have any questions, please tell me and I will try to come back and answer them

Understood. Thanks. Appreciate you taking the time.
The only thing I'm not exactly sure of is what you mean by ..."kan- wait"...
If you mean relying solely on the 4 for iipeiko, I got it. Have not seen the word kan used that way before.

カン(kan) = カンチャン待ち(kanchan machi) ( wait)

when talking about a wait that is single wait between two tiles ie ( wait) or ( wait). in Japanese you say it as kan - (wait) machi.

others you may already know are

( wait) , ( wait) are ペンチャン待ち penchan machi and when specifying the wait, similar to before you can abbreviate to pen (wait) machi

I think I confused you with kan as in four tile kan.

Once again, almost a year later, I would like to reaffirm that I haven't got myself into a rut (due to habit) on what to discard in these very common situations with multiple possibilities. How would YOU approach these situations as to a good discard, what to aim for, and whether to open it up or stay concealed? I would think there would be a variety of opinions here, maybe.....

Here are my suggestions:

1st screen - I'd go with Sha. None of them have been discarded so it could be that Toimen is waiting for his Yakuhai. I'd just wait to Pon Chun. I'd say it might be a hard one to win as all players seem to be collecting Souzu Chun could also not come out at all. I wonder if anyone would go for Chiitoi-Honitsu?
2nd screen - You're overloaded with pairs. I'd go with 6m/2m.
3rd screen - hmm I'd probably discard 4p to aim for Tanyao-Ippeiko. I don't see much benefit in opening this hand with such point distribution.

#1 I went with the South wind as a discard since there was one out already, aiming for honitsu and a dora, obviously, open or closed, with the intention of ponning the red dragon if it was discarded.

#2 I didn't see much difference here between the so I tossed the 8. Since there are terminals that would benefit this hand, my aim was riichi pinfu tsumo.

#3 Had my doubts on this one with 4 pairs and a pung staring at me, but because of the point situation, decided to toss the and see what happens next. The dora would be a good draw.

Glad I didn't have any time restrictions on these. Hope I made decent choices.