案件 • Board Game Arena

所有案件

日本麻將的案件

#46510: "First impression suggestions"

implemented: 此建議已獲實施

這是關於哪方面的案件？

日本麻將 #193940433

發生什麼事? 請從下方選擇

建議：依我所見，有些調整將大幅增進遊戲完成度

發生什麼事? 請從下方選擇

建議：依我所見，有些調整將大幅增進遊戲完成度

請檢查是否已有同課題案件

若肯定，請「投票」給這樁案件。最高票的案件將「優先」處理！

#	Status	Votes	Game	Type	Title	Last update

細節描述

案件沿革

2021年Aug13日 15:04 • ufm • 此建議尚未被開發者分析過:

I don't know I'll playtest this further (I play mainly on tenhou and other Japanese platforms, but not regularly nowadays), but tried to summarize my suggestions.

2021年Aug13日 15:06 • ufm • 此建議尚未被開發者分析過:

Also, any chance of variant rules other than red doras and game end condition?

2021年Aug14日 18:17 • gilthe • 此建議尚未被開發者分析過:

1) Individual setting: agreed.

2) Autopassing instantly is giving information away. This is a problem with Mahjong in relation to the BGA framework because the timer stops when the player becomes inactive. At a real table, you do not have the information that player X cannot make a call, here you would. The timeout options are for this purpose.

3) Well, a game can easily have 1000-2000 moves. Turn-based is definitely not recommended Not sure how to solve this elegantly. Autopass again gives away information.

4) I don't agree on that one.

5) In the meta-information, on How to Play or Strategy Tips I agree. I couldn't add those yet.

6) What other rule variants do you have in mind?

ps:
Some diplomacy if I may ask well, I created the worst mahjong tiles you have ever seen myself ;-) The point was of course to have a playable, beginner-readable set, irrespective of how aesthetically pleasing they are ...

2021年Aug14日 21:32 • ufm • 此建議尚未被開發者分析過:

2) Turn player does not need any confirmation. Riichi should be declared before discarding, so current implementation (which asks to declare riichi or not after discarding) is somewhat incorrect.

3) Perhaps allowing to choose what tiles to call preemptively? Or create a table option to enable autopass.

6) Abortive draws, counted yakuman, nagashi mangan, head bump, etc.
There are plenty of other variants (disallowing concealed kong robs even with 13 orphans, local yakus, and continues) but not really interested in those stuff.

ps: Sorry for badmouthing. I understand why that set is added, but it needs some refinement.

2021年Aug22日 02:05 • gilthe • 此建議尚未被開發者分析過:

Hi ufm, a clarification on 2)
Technically, I don't think it makes a difference, and in my understanding there is nothing substantially incorrect. The Riichi is indicated by turning the tile sideways, but the audible call is made before. However, because Riichi can be intercepted by Ron (and voided in the process), the full resolution of whether there is a valid Riichi call is never clear until after the discard is made and visible to all players. The EMA rules do not specify the call to be made before the discard in 3.3.14 but clearly recommend doing so in 5.3. I have also read instructions that recommend waiting with calling/turning the tile for possible Ron.
In my current implementation, the Riichi call is always visible to the others, even it is voided. This is the normal situation in real life as well and consistent with 5.3. The only difference to real life is that some quick-minded players might make a conditional decision whether to call Pon/Chi/Kan depending on the Riichi call. This is a general problem with call synchrony and reversal in real life.

3) See other issue (46572)

6) Sure, agreed, there are hundreds of variants. Double Yakuman are great but extremely rare (roughly 1/100000 wins on Tenhou). Nagashi Mangan is very rare, especially in 4-player Mahjong. The only one I had in mind from the beginning were abortive draws, so I'll look into that. The 2-Yaku minimum, possibly in bonus-dependent fashion might be an option to consider as well to mix up styles of play a bit more.

2021年Aug22日 03:03 • ufm • 此建議尚未被開發者分析過:

2) Actually, other players should notice turn player's riichi announcement before discard.
Riichi itself may affect calling decision (eg. calling to eliminate ippatsu, skipping 1000pt ron to change hand for more score if nobody declares riichi).
Current implementation does not declare riichi before discard.

6) Counted yakuman is counting 13+ fans as a single yakuman, not sanbaiman. 26+ fans do not count as a double yakuman, just in case.

btw, are you implying double yakuman is not implemented yet?
Personally I think counting a single combination (big four winds, 13-wait 13 orphans, 9-wait 9 gates, single-wait 4 concealed triplets) as a double yakuman is too much, but awarding multiple yakuman for multiple combinations (eg. big three dragons + all honours) makes sense.

Nagashi mangan is indeed rare, but as three color triples and stealing a kong.

2-yaku mimimum is an obsolete rule, really. No popular implementation employs that.

2021年Aug22日 06:18 • ufm • 此建議尚未被開發者分析過:

Currently all implementations (at least to my knowledge) declare riichi before discard.
irl mahjong goes the same. In digital games, the system prevents riichi when there is no valid discard.

2021年Aug22日 06:24 • ufm • 此建議尚未被開發者分析過:

Also, irl whether riichi is valid or not won't be revealed until riichi player declares win or the game ends in a draw. At that time if the riich was not valid it is subject to chombo (noten riichi).

btw, I found out in EMA rules do not allow agariyame. Surely EMA is quite different from popular Japanese rules.

2021年Aug22日 06:26 • ufm • 此建議尚未被開發者分析過:

Anyway, current mahjong implemenations display riichi button only when valid riichi declaration is possible, then do not allow invalid discards (to prevent chombo).

2021年Sep5日 00:28 • gilthe • 此建議尚未被開發者分析過:

On the issue of Riichi timing:
I am updating this discussion here:
As a compromise I have an in-between notification in the updated version. I want to emphasize again that logically the discard and calling Riichi are a joint action. In nearly all cases, the Riichi takes it legitimacy from the discard. As is written by ufm above, what many online clients do is to analyze your hand for you before a discard, then offer you Riichi, then filter the discardable tiles so you do not commit a penalty (Chombo). At a real table, in contrast, it is completely up to you to recognize Riichi and to make sure to discard tiles that create or maintain a Tenpai hand. This requires much more recognition skills from the player. As a result, I do not like this automation of hand recognition. If we accept this position, I identified 3 options in addition to my original implementation:

a) Mimic real-life: you can easily commit a Chombo by an invalid declaration. This would also be a nice way to play online but not beginner-friendly at all. Most players would most likely reject it even though it is closest to the real game. Major restructuring would be required.

b) Insert a second step in which only the discarding player is asked: discard -> optional Riichi call and announcement of call to other players -> discard becomes visible and claimable by others. I agree that this would be the best but it requires some more involved restructuring for very little gain.

c) Pull in the second step. discard -> optional Riichi call and discard becomes visible -> as soon as Riichi button is pressed, the other players are notified immediately -> other players can make simultaneous calls. This will work well in turn-based mode because the discarding player will get the option right away, so will usually call first. I hope this is a reasonable compromise.

Let me emphasize again that the order discard/Riichi Riichi/discard is not relevant per se if you know that you want to discard a specific tile to create a Tenpai hand and call Riichi. The only minor relevance in timing comes from the other players: specifically, from the rare cases where other players make a conditional decision on claiming that tile for a meld that depends on whether Riichi is called or not.

@ufm EMA rules
EMA vs Japanese rules: they are not that different (see your own link). I think competition rules have been converging for quite some time. In my opinion, Tenhou is actually more of an outlier than EMA.

@ufm point 6:
On double/stacked/counted Yakuman hands: as I said, these are extremely rare, they are not part of EMA rules, and not implemented. The reference of 1/100000 was to any multi-Yakuman hand.

2021年Sep5日 01:28 • ufm • 此建議尚未被開發者分析過:

About Riichi, I think b option would be the best. Trust me, a option will be outright rejected by most players. I've seen that already in Janryumon. But if you wish to stick with c option that would be the second best compromise. You are doing the work, so you have the final say.

For rules, yes they are different. You are completely ignoring rules used in jansous(mahjong parlors).
WRC, EMA, and various so-called 'official' organizations employ competitive rules which are designed for leagues they operate. They tend to reduce the weight of luck. For example, they do not use red 5, they count points based on their actual game score (including uma and oka), do not allow double ron, and some of them even removes uradora or kandora.
However almost all online implemetations such as tenhou, majsoul, MJ, and others do not because most of players don't play leagues or championships at all and they are more familiar with parlor rules. Players' ranks are calculated as rank points based on the game result only, not referring the scores.
Thus, I suggest creating some additional variant options (like Hand & Foot implementation) or two or three rule sets as a package (for example, 'EMA rules' / 'Parlor rules').

2021年Sep5日 01:40 • ufm • 此建議尚未被開發者分析過:

In short, there is reasonable logic for existing online implementations to adopt rules which contradicts rules from competitive organizations, and implementing more casual user-friendly rules definitely worths it.

2021年Sep5日 01:43 • ufm • 此建議尚未被開發者分析過:

For probability problem, no, it cannot be a valid defense in any case. Think about heavenly hand. big three dragons + all honours is actually more frequent than that.

2021年Sep5日 02:19 • ufm • 此建議尚未被開發者分析過:

Totally unrelated, but I play only no rate games in jansou. I'm against gambling.

2021年Sep25日 01:28 • gilthe • 此建議已獲實施:

I have solved the Riichi issue 1, I have addressed issues 2-3, and for 6 I have added abortive draws and variable starting points. For issue 5, I posted the basic documentation but somebody else had also posted some (partially inapplicable) rules, of which I kept what I could (under "Additional information"). I'll migrate some of the text to "Strategy Tips".

For multi-Yakuman hands I am not sure I follow your point, ufm. The probabilities are known and they are in fact extremely small. In any case, these extreme scores are not part of EMA rules, and they make more sense in tournament settings than in single games ...

If there's anything else, you can always reopen a new suggestion.

為本案件添加內容

任何可能重現這項錯誤或了解你的建議之相關資訊，都請在此填寫:

其他同樣狀況的桌號／步數
按 F5 是否解決了這個問題？
問題是否發生了好幾次？每次都發生？時好時壞？
建議將此錯誤的螢幕截圖上傳到 Imgur.com 並轉貼連結。

取消更新案情

#46510: "First impression suggestions"

這是關於哪方面的案件？

發生什麼事? 請從下方選擇

發生什麼事? 請從下方選擇

請檢查是否已有同課題案件

細節描述

案件沿革

為本案件添加內容

更改案件狀態為：

說明

貢獻

導覽

追蹤我們