More Thoughts About Menzen Tsumo

[HotelFSR]

This is in some sense a continuation of the thread I started around a month ago about whether the Han bonus for fully concealed self drawn wins is good for the game.

Since then I have played a lot more and improved my game considerably, although I am far from being an expert. I have come to accept this Menzen bonus as part of the game, but I can\'t say that I like it or that it makes much sense to me intellectually. I find the game most interesting when it comes to the business of choosing/defending the right waits, and I think a Menzen Tsumo bonus undermines the risk/reward balance in a variety of common gameplay choices.

Normally you want to go out on the largest number of winning tiles, such as with a two or three-sided wait, but sometimes you might want to bluff or trick your opponents by choosing a single/edge/closed wait that looks safe based on your discard pile. This adds strategy and variety to the game, but is undermined by the concealed self draw Han bonus.

When it comes to self drawn wins, you always want to maximize your number of possible winners- single waits are statistically harder to self draw, so when you are rewarded more for Menzen Tsumo there is less incentive to mix up your waits- and therefore less variation. This creates a bias against the more strategic and interesting parts of the game. Why give an additional bonus to drawing your own winner when it is already such a great bonus in itself? Self drawing a win is the least skill oriented part of the game and is not based on player interaction- there is no type of defense game, however tight, which can specifically predict or prevent a tsumo. Why reward it even further than normal?

Furthermore, this Menzen bias indirectly weakens hands like Chanta, Sanshoku and Itsuu. These hands tend to need a specific tile to complete, and so do not lend themselves to self drawn wins. This means no likely self draw bonus, and effectively makes hands like Pinfu just as attractive- if not more so. Considering that Pinfu makes up around 20% of hands, and Menzen another 20% or so, whereas hands like Chanta/Itsuu/Sanshoku all tend to be 5% or lower, why is this 4x difference in rarity not compensated? Consider than Menzen is more likely to stack with simple hands like Tanyao and Pinfu, so the reward discrepancy is actually far more than 4x!

In my live group we play without Menzen Tsumo, but instead give +1 Han to hands like Sanshoku, Chanta, Jun Chan and Itsuu, etc. This basically means that we boost all the Kui-Sagari hands by +1, except for the single suit flush hands (Chinitsu/Honitsu), because they do not tend to have such restrictive waits and are not nearly as rare. This is feels much more rewarding than the official rules when you actually win with a rare hand! It also adds more strategic variation to the game, by adding further incentives other than making the simplest hands with the best waits.

Why is an open Chanta-Sanshoku normally worth only 2 Han!? That\'s essentially the same as a Reach-Pinfu or Tanyao-Dora! Think of how common these are. Does anyone else feel that this is an unequal reward? Add the possibility of a Menzen bonus and the rare hands start to look even worse. Having a Menzen bonus restricts the effective play-style choices available, by undermining more interesting aspects of the game. The use of Red Fives compounds this problem because they create further inflation and can be most easily worked into simple hands.

I think this take is a more pertinent and pressing angle on the Menzen debate than the variance issue brought up in the previous thread.

Comments anyone?


P.S. before anyone mentions the issue of Menzen being used as a defensive hand (i.e. with no other yaku)- the average number of hands with just Menzen is only around 4%! The issues discussed here have a massively greater impact on the game.

[Shirluban]

Normally you want to go out on the largest number of winning tiles, [...] but sometimes you might want to bluff or trick your opponents by choosing a single/edge/closed wait that looks safe based on your discard pile. [...]
when you are rewarded more for Menzen Tsumo there is less incentive to mix up your waits- and therefore less variation.

So, it\'s that incentive to go for the maximum number of winning tiles who enable you to put such traps.
With more variation in waits, your opponents would not assume you\'re going for 2/3-side waits and would be carefull about single waits... nullifying your trap.

Furthermore, this Menzen bias indirectly weakens hands like Chanta, Sanshoku and Itsuu. These hands tend to need a specific tile to complete, and so do not lend themselves to self drawn wins.

W... Weakens !!??!! :ohmy:
These hands trend to end with a two-side wait with only one side giving the yaku.
Menzen tsumo enable you to go out with the other side.

Let\'s take en exemple:
3-bam 4-bam 5-bam 3-crak 4-crak 5-crak 4-dot 5-dot 7-bam 7-bam 7-bam 9-crak 9-crak

It exist two winning tiles, 3-dot and 6-dot.
In the case you self-draw of one them :

- If you allow menzentsumo :
3-dot : san shoku + menzen tsumo = 3 han
6-dot : menzen tsumo = 1 han

- If you forbid menzentusmo :
3-dot : san shoku = 2 han
6-dot : nothing = nice with furiten or weep?

Menzentsumo REENFORCE hands like san shoku, chanta and itsu.

By opposition, a pin-fû hand can end whatever the winning tile you draw, with or without menzentsumo.

[HotelFSR]

Hmm...

I definitely take your point. Is that really the whole story though? I\'m not sure yet, but I could be wrong.


e.g.


a) \"without the bias for two-sided waits, traps would be nullified\"

Is this true? I think you would still pick multi-tile waits extremely often even without Menzen Tsumo. You will still want a higher chance of winning by self draw or available discards (just as you would with an open hand where there is no self draw bonus).

b) \"menzen STRENGTHENS hands like chanta, sanshoku, etc.\"

Is this always true in practice? It still does weaken these hands relative to hands like Pinfu and Tanyao which will work with tiles on both sides of the wait. The rare hands are also more likely to require closed and edge waits.


This is also why when we play without Menzen, we add +1 Han to these rare hands, to compensate for the mentioned effect without boosting the common hands. Also, in pracitice hands like Chanta are very often open anyway, and do seem to be scored too low in those cases irrespective of the Menzen issue.

i.e. open Chanta Sanshoku = only 2 Han

[zzo38]

I would prefer to make menzen tsumo a virtual yaku, meaning it counts 1 han only for purposes of requirement but not for purpose of scoring.

[gartheee]

You make a very persuasive argument against Concealed Self Draw (CSD, Menzen Tsumo), but in the end, I’m not persuaded.

First of all, if we’re going to complain about yaku (hand points) that are just based on luck, let’s get rid of the Instant Win hands (tenho and chiho). I know they have infinitely less impact on the game than the subject at hand as I have yet to have ever witnessed either in action. Still, the fact that they’re lurking out there, ready to pounce on the day that I finally make it to the final table of some tournament, blech, it’s lost me many a good night’s sleep.

But on to the subject at hand.

I’m willing to concede that just because we can think up some weird situation or combination of tiles, doesn’t mean that we should assign hand points and give it a name. That would be American Mahjong, to which, as I understand we can give thanks for All Green (Ryu-ii-sou) which I think is as silly and luck-based as a hand could get. We could eliminate it and almost nobody would be any the wiser. And I certainly won’t argue with you that the point assignations for the different hands are not always perfectly reflective of their relative probability. Still, eliminating the hand point for Concealed Self Draw is not going in the right direction.

First of all, Reach Mahjong (Japanese) tends to reward play that keeps hands concealed. CSD is another in the list of rewards for not stealing tiles. And it should be fairly obvious that the probability of drawing all the tiles one needs to finish a hand is lower than the probability of getting some of them from other players, thus its due reward.

Indeed, the frequency of Peace/Simple hands is probably higher in comparison to open Mixed Outside/3 Colored Runs hands. However, I’d say that has more to do with the reward than the actual probability comparison. As you point out, not only does CSD stack nicelier there, but so does Reach and hence, Hidden Lucky Dragons. Of course people are going to aim more for Peace/Simple more than cheaper open hands. Take away the rewards for keeping hands closed and you’ll have the willy nilly tile stealing/exchanging that goes on in Chinese Mahjong. Why wait to draw what you need when someone’s just pooped it out on the table there in front of you. They can’t wait to open up their hands. “Look! So far, I have this, and this, and this! How do you like that!” Blech.

Also, there are times when for one reason or another, you may decide to pass on a winning tile from another player in order that you might draw it yourself. That sort of tight play needs to be rewarded, hence the extra hand point.

And finally, are you honestly suggesting that if a player gets a hand to ready but has no hand point in it, he must Reach or just give up the hand? That you can’t defend against it unless that player is forced to Reach with that otherwise \"pointless\" hand? This harkens back to the conundrum of whether the best offense is a good defense, or the best defense is a good offense. Well, in the end it’s probably a little of both, as usual. And if the fact that someone might be putting his hand together faster than you is what finally puts a little fire under your chair, maybe that’s a good thing after all.

So I would recommend figuring out how to work Concealed Self Draw into your game rather than campaigning against it. Increase the points for rarer hands, eliminate the ridiculous ones, but CSD has earned its point on the bottom rung of the ladder.

(Incidentally, if you REALLY want to take more luck out of the game , play the competition rules where you eliminate the hidden lucky dragons and the First Turn Wins. That will do much more to increase the value of the more obscure chanta, sanshoku, etc hands than eliminating Tsumo.)

[zzo38]

First of all, if we’re going to complain about yaku (hand points) that are just based on luck, let’s get rid of the Instant Win hands (tenho and chiho). I know they have infinitely less impact on the game than the subject at hand as I have yet to have ever witnessed either in action. Still, the fact that they’re lurking out there, ready to pounce on the day that I finally make it to the final table of some tournament, blech, it’s lost me many a good night’s sleep.

I already do prefer to get rid of tenho and chiho (and renho). And I have made a first-turn win hand once.

I’m willing to concede that just because we can think up some weird situation or combination of tiles, doesn’t mean that we should assign hand points and give it a name. That would be American Mahjong, to which, as I understand we can give thanks for All Green (Ryu-ii-sou) which I think is as silly and luck-based as a hand could get. We could eliminate it and almost nobody would be any the wiser. And I certainly won’t argue with you that the point assignations for the different hands are not always perfectly reflective of their relative probability. Still, eliminating the hand point for Concealed Self Draw is not going in the right direction.

I also prefer to get rid of all green and big wheels hands (both of them are already worth many han anyways and don\'t need to be made worth more). I make menzen tsumo a virtual yaku (1 han for purposes of requirement, 0 han for purposes of scoring).

(Incidentally, if you REALLY want to take more luck out of the game , play the competition rules where you eliminate the hidden lucky dragons and the First Turn Wins. That will do much more to increase the value of the more obscure chanta, sanshoku, etc hands than eliminating Tsumo.)

I usually remove the yaku of first turn win but keep kandora and uradora (but eliminate kanuradora). And in Extended Washizu (my own ruleset, I will post it later when it is more complete), uradora works differently and depends on discarded tiles (as well as a few other things).

[Senechal]

And in Extended Washizu (my own ruleset, I will post it later when it is more complete), uradora works differently and depends on discarded tiles (as well as a few other things).

I don\'t think we\'re playing the same game anymore. I can understand the fanboyism of wanting to say \"I play Usagi style\" but talking about extensions of a style that isn\'t playable without 3/4 transparent tiles and then editing stuff leads to irrelevance, at the very least to the topic of discussion, and probably moreso down the line. Besides, assuming a standard Washizu style, wouldn\'t you want Menzen/CSD counted when you could have your winning tile fed to you almost at will ?

And to sum up Garthe...

Take away the rewards for keeping hands closed and you’ll have the willy nilly tile stealing/exchanging that goes on in Chinese Mahjong.

QED - setsumei shuuryou. (System doesn\'t like kanji input)

[zzo38]

And in Extended Washizu (my own ruleset, I will post it later when it is more complete), uradora works differently and depends on discarded tiles (as well as a few other things).

I don\'t think we\'re playing the same game anymore. I can understand the fanboyism of wanting to say "I play Usagi style" but talking about extensions of a style that isn\'t playable without 3/4 transparent tiles and then editing stuff leads to irrelevance, at the very least to the topic of discussion, and probably moreso down the line.

The comment about Extended Washizu only applied to that sentence (the part about uradora), not to the rest of my posted message.

The things about removing tenho,chiho,renho,all green,big wheels, making menzen tsumo virtual, and removing kanuradora, and making ippatsu count only for ron (not tsumo), I prefer to apply even when playing *normal* mahjong (although the game I currently play on the computer (in the linked screenshot) only allows removing big wheels and kanuradora so those are the only ones I remove).

And I am not taking away most of the rewards for keeping hands closed, I am only making menzen tsumo virtual (and I still use most of the other things that reward keeping hands closed, according to their normal rules).

[gartheee]

More incidentally, I have yet to play at a Mahjong parlor which recognized Big Wheels as a limit hand, and both it and the All Green hand are so rare as to have essentially no impact on the game anyway.

There\'s certainly nothing wrong with changing/adopting special rules to play with your friends. Pretty much every time I play with my friends, we spend the first 5-10 minutes making sure we all agree to the same rules. Likewise any Mahjong parlor will do a quick 5-10 minute explanation of the rule set they use before allowing someone to jump into any games there. Everyone has rules that they like/dislike so such rule confirmation is pretty standard. You should just be aware that you\'ll probably have a hard time getting any Japanese players to agree to make Tsumo worth 0 points. The only thing that might be harder to get rid of than Tsumo would probably be Reach itself.