Reachmahjong.com

First off, I would like to strongly thank xKime for providing me some help with his expertise in Mah Jong. I would like the members of the community to look at some of my stats and maybe criticize some of my recent games.
I am currently at 3 Kyuu with a rating of 1752. I have logged around 50 games in this account.

My Tenhou Stats:
http://arcturus.su/tenhou/ranking/ranki ... e=NoSatsui

My Most Recent 4 Games

http://tenhou.net/0/?log=2011122706gm-0 ... ee5a9&tw=1
http://tenhou.net/0/?log=2011122706gm-0 ... 9b1f0&tw=0
http://tenhou.net/0/?log=2011122706gm-0 ... f5fdf&tw=3
http://tenhou.net/0/?log=2011122705gm-0 ... 5c019&tw=0

I accidentally made a furi-ten riichi and luckily got my draw. Then there was the time where I switched 8 bamboo for 4 sou when 5 sou was dora. A few errors, here and there. Then there was when I folded my near ideal tao-pin (possibly 3 colour straight) - (Not exactly sure whether I should or should not have folded it). Also, I ended up making a few riichis where I only had Riichi only or maybe 2 yaku.

I appreciate the criticism. Thanks!

Thank you so much for the appreciation. But I don't profess anything other than what other people have investigated. Like most (pseudo-)digital players, I can't conduct a research of my own for optimal data, just use that of researchers and try to apply it in-game. That's the difficult part, but it will also be what determines strength afterwards, too.

In just a few games, there is a lot of variance. So don't draw too many conclusions just yet; stats start meaning something after 500-1000 games. Thanks to variance, ridiculous things can happen. You can get to 3rd dan in 99 games, or to R1800 when you're still 4kyu, or get like 35 games without a last place:




What matters is what you get in the long run.

There is a text-book'ish situation right in the beginning of the first game. East One, 9th Draw. Dora 6p.



East is a live tile. You discarded .
When you discard 8m, you fixate on 7m as your pair and reduce your tenpai chance to 3m 1s 4s. 1s doesn't really make you any happy either. The worst thing that could happen is getting 1s around the 12th draw just to deal in with the East.

The correct answer is discard

Sure, it could get pon'd by someone, but you cross that bridge once you get to it. Also, the chances for it being pon'd and not ron'd at that point are good enough, and also (though it may not look it at first sight) your hand is good enough. Of course you still get to tenpai with 3m and 1-4s, but you have also the chance to draw 2m and 4m for perfect one away of mentanpin sanshoku dora 1.



It has many chances to be a valuable hand. (The "worst" thing that could happen at that point is drawing 7m or 2m and winning on 1s, but we're pretty happy about that, too!)

When you break up a ryanmen-toitsu shape to fixate on a pair while you still have other shapes that need improvement (like penchan or kanchan) you're losing quite a few tiles for hand evolution.

Remember: The basic rule of mahjong is creating ryanmen. The basic rule of a ramen restaurant is creating ramen.

More: East at that point is not a safe nor a danger tile. Not even a useful tile for tenpai. It doesn't meet any of the three requirements, so discard it away! Your tenpai speed is what matters the most when there is no attack sign from the others, so you must obediently attack strongly when there are openings like this. If afterwards you draw 3m and you happen to deal in with 8m, that's just unavoidable, mere coincidence, as it was the best choice at that time, what happens afterwards matters little. Only the long run matters.

Here's my two cents!

As a general matter, I think you trend to keep an isolated honor too long.
Even an isolated suit tile is more useful to build your hand, and a live honor becomes more and more dangerous as the game goes. If you can't use it for your hand, and can't use it as a safe tile, discard it before it becomes dangerous.


1st hand:
E1:
Same remark that xKime, you had kept that East too long.

E4, 1 continuance:
Dora is , you've discarded the from .
Your hand was (red)
If you're not going for ittsu, you'd better discard or now and keep the chance for the dora.

2nd hand:
E3:
I see no reason to keep the red dragon so long.
The 5m and 7p you've discarded before were more interesting for building your hand.

3rd hand:
E2:
7th discard:
is more interesting than .
can only be improved with a , while can be improved by any of .
Plus, a will give you iipeikou, and prevent you to have tanyao.

14th discard:
You've discarded , breaking your iishanten hand, while was safe.
At this point, you still had a chance to be tenpai/win without taking any risk (and you would have been tenpai, if you had discarded 3m and 9m instead of 8m and 9m).

E3:
Dora was
I can understand you've discarded to keep an isolated , since you didn't have any yaku and gives a change to get sanshoku.
But why discarding this for a ? Two South are already gone, so you're certainly not going to do anything with it.
Plus, it moves you away from your other possible yaku: tanyao.

E4:
7th discard:
The West pair is not much interesting since it can't become better than a pair, but it's still far more useful than a lonely White dragon. My point is not against discarding the pair by itself, but about discarding it in favor of an unusable tile.
I would have no problem if you had waited one more turn (i.e. when you drawn the 7sou) before breaking the West pair.

Thanks once again, everyone : P

I think that I have some problems with honor discards and when to discard them. I obviously understand the principles to not discard a live dragon or fanpai during mid-game when someone's showing signs of a half flush, but sometimes I'm unsure how whether to keep it or not. My strategy against some players who are "pon-happy" and want quick wins is to stack up a fast tenpai, discard the wind or dragon tile when I am nearly ready or ready, and declare riichi the round or two after. Since they have already exposed their hand, they'll have less safe tiles and ultimately may end up dealing into my riichi more frequently considering that they don't completely fold / have enough safe tiles to last them the entire round. I'll keep dealing east in mind considering that as you've said, I could've got a great tsumo/ron with dama ten mentanpin sanshoku dora 1.

Hmm, sometimes I deal closed waits because I feel as though they may slow me down. I should've definitely dealt 1 pin in that situation though. I was probably scared of dealing 9 pin because drawing a 8 pin would backfire. For the other hands, I probably just ended up making mistakes all around.

Also, one thing I've noticed myself doing is keeping a safe wind for a turn while dealing out a tile that may be potentially useful. For example...

Draw: (2 have already been dealt)

Sometimes, I may deal if east for example, has already pon'd on a bonus point fanpai and appears to be close to honitsu tenpai. I feel as though sometimes this is effective, but other times very bad to do considering that it'd backfire the moment I draw 1 sou to 5 sou for example.

Also, another question: Would you pursue a strong hand even when you are in the lead? For example...
East 3rd Round; You are at 42 000, south is at 16 000, north is at 22 000, and west is at 20 000. You are given this hand:
Draw:
Would you go for Dai San Gen/Shou-San-Gen by ponning Hatsu and Chun when they come out, and maybe waiting to draw a haku? Or would you simply pon one dragon when you can (the logical choice of course).

Ponning Both Dragons can cause higher # of points, potential intimidation to cause opponents to fold, and more danger since more of your hand has been exposed which means less safe tiles available for drawing.
Ponning just one dragon would be a least number of points, no intimidation, and less danger obviously.

What would you choose? Thanks for the help you've already done. I strongly appreciate it. (The hands I'm coming up are just from the peak of my head, just to note.)

"Other players are pon-happy" is the oldest excuse in the book to tamper with the timing for discarding honors. No matter what the old people at renmei may tell you, most of the strongest players I know -are- "pon-happy." And they should be! Unless you're playing in a parlor where the importance of menzen has a huge impact on your average winnings, the speedy "naki mahjong" way is the modern way to go. (I've played against and alongside Nihon Pro Mahjong Kyoukai's Kawamura Akihiro, and he calls from very odd shapes but he wins around 30% of the hands before the 10th draw while only dealing in 10% of the time. He is the definition of "pon-happy" but strong as hell! It gets almost impossible to win a hand against him. I'm not taking importance out of menzen, ippatsu and ura-dora are pretty strong so you should go menzen when you can, but because of that it is extra-important to speed up your hand by calling when it is evident you won't be getting there with a menzen hand yourself. A vast majority of mahjong hands have at least one or more pon declarations.)

More than that, rather than worrying about the calling habits of people in the internet that you don't know, worry about your own hand. Most of the "dealing in" doesn't happen from keeping a lone 3s flying free in your hand. It happens because deep in the game, you're not making the right push-pull judgement or even worse, you're not doing either of them (uchi-mawashi). "Oh, dayum, I dealt with that 3s. I should have kept the south an discard it when it was safe." No. If you wanted to attack, that 3s is a vital part for evolution in your hand; or otherwise, if you wanted to defend, you don't deal a dangerous 3-7 tile. If you do things in a half-assed way, you get half-assed results. If the correct fight-or-flight judgement is "fight," if it's "flight" then flight. If you kept pushing a pinfu-nomi ii-shan-ten han in late game and you happened to end up dealing in with that 3s, the problem wasn't "not keeping south in the 9th draw," the problem was that you kept pushing too much in a situation that no longer called for it.



As long as there aren't any special conditions (riichi on the table, dealer with three calls, last draw, whatever), just discard south instead of 3s. Because the chance for 2s or 4s is very important (sanshoku included). If you really feel like it's a round to "stop to smell the roses" and lay-low, then you may consider 89s. But it's a huge setback, tenpai chance-wise.

Of course, if the right judgement is to fold (like you commented about the dealer's movement), don't do either, start a proper betaori (fold) discard order.

East 3rd Round; You are at 42 000, south is at 16 000, north is at 22 000, and west is at 20 000. You are given this hand:



I'd still go for the fast win if possible. It looks fast, but it's also kind of heavy. While obediently discarding guest winds, expect the yakuhai pon and go for the win until a riichi (or mid-game dora pon, or we get into late game and the hand isn't moving, or whatever other obvious 'it's time to give up' situations) comes in then fold. I don't think I would call the 23s ryanmen right away (if we were on a different point situation, I'd do it right away for 1s anyway). That's about it.

About the "intimidation" part, try not to depend on off-game things like that. Just think it rationally from everyone's standpoint. But more importantly, your own.

Thanks once again for your clarification on my rash tendencies. I think there may be a problem with how I'm playing if you are typically expected to be pon-happy. I go for menzen 80% of the time and occasionally pon when I am already in the lead, or have a fast hand to offer. I've always felt the Naki mahjong is extremely dangerous because of the threat of riichi's. Do you actually have some theory on hands and when to pon/chi them to maximize speed? The player you played, Kawamura Akihiro sounds incredible if he can win 30% of his hands before the 10th draw haha.

GGnores wrote:Thanks once again for your clarification on my rash tendencies. I think there may be a problem with how I'm playing if you are typically expected to be pon-happy. I go for menzen 80% of the time and occasionally pon when I am already in the lead, or have a fast hand to offer. I've always felt the Naki mahjong is extremely dangerous because of the threat of riichi's. Do you actually have some theory on hands and when to pon/chi them to maximize speed? The player you played, Kawamura Akihiro sounds incredible if he can win 30% of his hands before the 10th draw haha.

80% of menzen means 20% of naki. It's kind of low. What is your hand win % (average won hand value too) and your deal in % (average deal in value as well) in relation to that? I think you may be playing overtly passive.

Well, I would have to write a 100 page book with examples to tell you in detail (which I am doing!) but in short:

A one away starting hand, a hand that is both cheap and slow even if you open it, a hand that has no yaku or a difficult yaku if you open it, a hard that you need to squeeze a lot of value from (like, haneman distance/mangan distance in South Round from third place): Assume menzen.

A two/three away hand with one or two dora respectively, a hand that is close to tenpai but will be cheap and have a bad wait when it gets to tenpai even if menzen, a hand that you need to win no matter its value, a one away hand in mid-late'ish-late game, a hand that can be high-value even if open (because it will face hurdles if you assume menzen, like hon itsu with a yakuhai and dora, or hon itsu with two yakuhai), assume naki.

A hand more than four away, if you don't really need the win, concentrate on future defense. If you kinda need the win, concentrate on future defense while aiming for chii toi. If you need the win, discard tenari for the meanwhile and obediently build your hand. If you need a -lot- of points, chii toitsu or kokushi musou with all your might.

You cannot compress mahjong strategy in a few lines like this, but the point I'm trying to bring upon you is that:

Your hand (uke-ire and how many away), the point difference, the dora and the table (including riichi deposits and whatnot) and more importantly the rules you are playing with; that's where your information towards what to discard and what to call is. Not whether your opponents are pon-happy, or you "feel like" or "uneasy with..." The judgement comes from that information. Not the last tile you drew, or whether you won the last hand or not, or whatever other occult "nagare" or "poker player theory."

Even when I used to play equally to how I play online at a parlor, I was mostly losing money instead of gaining rate. Can you guess the reason? I wasn't playing according to the rules in a way. I was expected to make more use of menzen (for the ippatsu and ura dora chips) and all in all be more aggressive (especially when you have red 5's in your hand).

I see the opposite pattern happen quite often as well. Parlor staff going for hands not worth of it, and passing on calling vital tiles for speed online. Then they don't get a stable rank, and they blame it on "internet mahjong being just a game" or "the other players aren't serious" or "this server is rigged to give out good hands to everyone." Or lately "Everyone is so cheap!"
Funny how in mahjong it's always something else's fault when you lose.

Being "sucky" at mahjong, basically means that the income you want (be it money, rate or rank) is not coming to you. Improving, is finding the reason behind it and altering it. If you realize you're dealing in too much into the dealer, fold a little more to him; if you are not winning enough hands, try to fight them better. You study the game and yourself in order to do that.

The important thing about mahjong is that there is no "immediate improvement." There is no "I just learnt this cool thing, so now I'll win ALL the games!" It is a long run mission. And for a casual player, this results in a very, very long journey.

I am unsure about finding my average hand win % and deal in %, so I just uploaded an edited version of my current stats on paint and uploaded it.

http://s1209.photobucket.com/albums/cc3 ... ustats.jpg

So depending on the information given (as you've listed: riichi deposits, the rules being played with, point difference, # of dora in hand, etc.) would determine whether I should go for naki or not. It's generally situational play. It's just based on how we all play then, whether we are in fourth and need a comeback (to get the desperately needed dealer position) or even first just to maintain the lead. What is the ideal percentage of going Naki under your opinion? I assume somewhere between 50% - 75% then?

Oh, and of course, Mahjong is slowly built on. I'm more than willing to play and improve slowly.

Oh, you rarely even deal in, and you win a fair amount of hands with a lot of riichi, so it's not bad! You just like to lay low while everyone is doing their thing. But I think that with that skill, you can do a lot better if you try taking more initiative!

Most of the time, it's not about calculating EV like a machine, but simple observation. If it is pretty obvious that you're not likely to win the hand no matter how much you reach your hand into the wall, calling is the next step for speed.
Some people take this way to heart and go through horrid hurdles to win a bad shape 1000 point hand with three calls, but that's not digital mahjong either. Especially not in East Wind only games. Rushing towards cheap hands from the start only hurt the game there; it's too much risk for too little gain, and all you accomplish is to have everyone close in points in East 4 which is the worst...

Somewhere between 300-400 naki (average win around 3 han or a little less) is the standard for strong players in tenhou. Some people like me go way overboard and cross the 40%. But that's just because I -continue- to call once I called once. I need more patience on that aspect; the fact I called once doesn't necessarily mean I should be calling more and more.

It's a thing of balance. I just like to keep in mind that being tenpai is very, very important in mahjong. Between being tenpai for a 2000 point hand with good shape, and being one away from a mangan hand... well, tenpai is (usually) better than any one away shape.

Also, to clear that up, Kawamura doesn't win before the 9th draw 30% of the time. He wins 30% of the hands; and it's in average around the 9th draw. There's a difference. w

Originally, I enjoyed laying low and dealing almost all safe tiles unless I was going to be at 1 shan ten / tenpai with a good shape if I were to deal a tile that isn't in my opponent's pond. I usually ended up getting 2nd/3rd for this, and a few times, 4th because I was dealer and it ended up costing me the most for the points payout. I tried incorporating Naki mahjong with some good results today. I've actually gotten 1st in 3-4 of my games today and have luckily avoided some riichi's. Sometimes, I wonder whether I should just fold to the a possibly dangerous riichi from another player, or whether I should deal a somewhat dangerous tile so that I could continue being in tenpai. Riichi Mahjong's payouts can obviously differ significantly if somebody suddenly hits 2 dora for their pair and gets an ippatsu. Sure, I've been lucky today, but I'm not actually certain whether I'll be as successful in the future. Most of my end shapes were generally good (ryanmens) and I did win a couple of rounds while only being a few thousand points ahead of second place. I've also pretty sure that I've made no more than 2 calls in a game. Thanks for everything once again. I look forward to reading your 100 page book on Naki Mahjong Theory.

Oh, and on a side-note, could you tell me what all the kanji means on the the Tenhou Stats box? All I understand is the percentage in first, second, third, fourth and the rating. I don't understanding what the other kanjis tell you about your own stats.

I figured it couldn't hurt to do this myself, and maybe this'll become a "post your stats" thread? This is a new account: I lost the ID for the last one (3xbreak, break^3 after that, I basically ran out of ways to type my handle in English on Tenhou) and figured, fine, I'll do it from scratch again. Here's what the stats look like, having just made it to first dan.



Likewise, here are my last four matches for better and for worse.

http://tenhou.net/0/?log=2012011207gm-0 ... be514&tw=2
http://tenhou.net/0/?log=2012011206gm-0 ... 8289e&tw=1
http://tenhou.net/0/?log=2012011205gm-0 ... 71e0d&tw=0
http://tenhou.net/0/?log=2012011115gm-0 ... ff519&tw=1

One position I noticed was in the first East round of the second game you posted, third discard where your hand looked like this:

drew:
and you cut the . I think if you had cut the it would give you more possible tiles to improve your hand.

------------------------------------------------------------------------------------------------
In the second second East round the player to yor left reached on the third turn on a , you discarded a even though you had a . As the dealer you are 3 tiles away from ready, why not discard the and be safe? You wouldn't need to suffer that painful extra "first turn win" yaku, and no-one else can win off that tile (even though they could pon or chi it). The player to your right has ponned a set of so he is already going to have more trouble safely discarding around the player to your left.

------------------------------------------------------------------------------------------------

on the thrid East round of that game, the dealer reached. You were 3 shanten (I think, I'm still not good at counting it out) so I would have thought it would be better to play it safe. You tossed out the dealer's winning tile even though you had two safe tiles (for the dealer) in your hand. Granted there were tiles in the pond that pointed to that tile being possibly safe, but you had two for sure safe tiles to use up first.
------------------------------------------------------------------------------------------------

That's really all I can contribute: try to play a little safer in situations where you aren't ready and someone reaches. It might lower your ron rate just a little bit, although I don't know what a good value for that should be.