Preface
This article will explain how to improve your performance in online mahjong games such as Jan-Soul and Tenhou.
Since I'm writing this based on my own experiences, I think it might resonate with some people but not others.
I hope you can use any parts that seem useful for improvement as a reference.
I will also introduce any strategy books that I found helpful in improving those aspects.
One of the advantages of online mahjong is that your performance is recorded.
While Mahjong Soul still yields less data than Tenhou, it still provides far more comprehensive data than real-life mahjong.
The ability to collect this kind of data means you can visualize your strengths and weaknesses, so I think online mahjong is a useful tool for people who want to improve their skills in real-life mahjong.
As of January 2022, when this article is being written, my overall record in Mahjong Soul: Jade Room is as shown in the image above.
現在は雀聖3で、特徴としては和了率・放銃率が低めで打点が高く、スピードが遅めの慎重なタイプです。
Please understand that there are many aspects of my approach that differ from that of speed-oriented players with high winning and calling rates.
In the following sections, I will explain how to improve each of the main metrics that can be referenced in Mahjong Soul's performance.
It's written in a way that you can skim through it, so feel free to read only the sections that interest you.
Points to note when analyzing performance
When analyzing your performance in Mahjong Soul, it's important to eliminate as much noise as possible.
The stats displayed within the Mahjong Soul game include the results of all matches played to date.
If scores between bronze and ball are mixed in, it becomes impossible to accurately assess the results.
If you are currently playing mainly in the "ball room",Suzume Tama Book StoreIt is important to first display the results for the ball-playing area and analyze them.
If your statistics include results from areas you don't primarily use for battles, it can make your numbers look better than they are, so you need to be careful.
Winning rate
My impression is that in Mahjong Soul, the winning rate is generally around 20% to 25% for most players.
To improve this, you need to increase your open-hand rate and riichi rate.
そしてこれが高くなっていくと、どうしても放銃率も上がっていきます。
和了率は、常に放銃率とのバランスで考えなければいけません。
Even if you make a lot of money, if you've lost even more, it's hard to say you're doing well.
簡易的な指標として、「和了率ー放銃率」が10以上あるようなら、ある程度まとまっているのではと考えられそうです。
Conversely, if the score is below 10, I think some kind of improvement is needed.
To increase your winning percentage, you need to consider the basics of tile efficiency, how to make calls, and the balance between pushing and folding.
If you're interested in tile efficiency, you should first learn basic hand-building theories such as the "5-block theory," and then move on to learning about frequently encountered shapes like "middle-bulging shapes" and "sub-two-sided sequences."
If those aspects are done to a certain extent, then I think there shouldn't be any problems.
Occasionally, you'll see players who have moved beyond the beginner level who have an extreme aversion to penchan and kanchan, and will pay them off even if it means reducing their number of shanten (shanten) points.
Depending on the turn in the game, it has become clear that even if you have a weak hand that only allows you to declare riichi, a preemptive riichi is quite advantageous, so it is also necessary to have the courage to simply aim for a ready hand (from "New Scientific Mahjong").
If you have even one dora tile, a weak-shaped reach is sufficient, and if you have two, you should definitely aim for a ready hand.
When building a hand using melds, it's important to be aware that if you start calling tiles before you have a pair or block, the number of possible winning hands will be significantly reduced.
For example, when making a pair in a closed hand, you can make a pair if you can match any one of the 13 tiles in your hand.
However, once you've called and have 10 tiles in your hand, you need to choose tiles from among the 10 different types, which makes it simply more difficult to make a hand (this isn't a problem if you're okay with ending up with a single tile wait).
The same applies to blocks. If you have five blocks and it only means you have to discard a follow-up tile by calling, that's fine. However, if you were waiting for them to attach to floating tiles, the chances of forming blocks (or pairs) decrease because you won't be able to hold many floating tiles.
It's not just about eliminating kanchan and penchan through calling tiles; the important thing is whether you can call tiles in a way that increases your winning percentage, taking these points into consideration.
If you're going to learn from strategy books, I think you'll be fine with Nemata's "Modern Mahjong Techniques for Winning More" for the basics, and G. Uzaku's "301 Selections of Standard Discards" for practical problems.
In addition, I found Makoto Fukuchi's "Handmade and Push-and-Pull Tactics" to be quite easy to understand, as it was given ample space for explanation.
Also, even if you don't buy a bookYouTube channel of mahjong creator Genki HirasawaYou can also learn in an easy-to-understand way by watching the videos in the "Tile Efficiency Course" category.
Dealing into a winning hand
放銃率は、雀魂では10％～13％くらいが適正範囲という印象です。
I occasionally see people with over 16% support, but I think that's definitely going too far.
放銃率を低くするには、ベタオリの基礎知識を学ぶこと、押し引き・状況判断の見直し、副露率が高過ぎないかの検討などが必要となってきます。
First, it's essential to understand the order of danger levels for the tiles you discard when playing defensively.
Those who are learning about defensive play will probably understand this to some extent, but it's reassuring to also learn the order of the levels of danger.
For example, between a 1 or 9 tile that is in a safe sequence and a discarded wind tile that has not been discarded, the latter is more dangerous. If you discard these tiles without thinking, you will end up making more unnecessary mistakes.
If you constantly try to visualize "what kind of hand your opponent has if they were to deal into your hand with this tile" by working backward from the number of tiles you can see, it will be very helpful in estimating the level of danger.
If you've mostly mastered the basics of folding, then the next crucial factor is whether you're effectively managing your pushes and pulls.
Make sure you're not pushing too hard on anything from just one step away from a winning hand, or if you are pushing, are you aggressively pushing with unrelated 4, 5, and 6 tiles, or are you making too many reckless chasing riichi calls?
副露率が高すぎる場合も、手が短くなるため放銃率も高めになりがちです。
If your winning rate increases along with your calling rate, that's great. But if not, you might be calling too many tiles.
Also, you might be pushing too hard because you don't have any safe options.
Regarding Betaori, the basic points areYouTube channel of mahjong creator Genki HirasawaLearning "push/pull/fold" is sufficient.
Furthermore, if you're looking to learn from strategy books, I found Makoto Fukuchi's "The Textbook of Push and Pull" and "Handmade and Ironclad Tactics of Push and Pull" to be helpful.
Totsugeki Tohoku's "New Scientific Mahjong" is useful for referring to push-or-fold criteria, but because the content is quite detailed, it is recommended for intermediate players and above.
secondary dew rate
My impression is that the rate of players using melds in Mahjong Soul is mostly between 30% and 40%.
A high rate of calling tiles isn't necessarily better, but I think it's a good indicator of a player's style.
For those who prioritize playing without calling tiles, the figure is often less than 30%, while for those who prefer calling tiles, it's around 40%.
副露率が上がれば和了率と放銃率は上昇し、打点は下がっていきます。
I think it's important to explore what numerical values suit your own style best.
However, there are three patterns for determining when to vocalize: "when vocalization is absolutely necessary," "when vocalization is optional (the difference is slight)," and "when vocalization is not necessary."
Considering this, I think the percentage will probably settle around 32% to 38%, depending on how often they chirp, whether they chirp or not.
I tend to be a conservative player when it comes to opening tiles, so I can't really speak with much authority on this topic.
However, there are several points to keep in mind regarding openings.
Generally, in the early stages of the game, there are many situations where it's okay to pass up the opportunity to call a tile, but as the game progresses, the situations where it's better to call a tile increase.
Furthermore, if the hand has many unfavorable shapes, and even if it's a closed hand, it's only good for declaring riichi anyway, then players tend to call tiles earlier.
In such cases, the strategy becomes cheap and far away, so there is a risk of facing difficulties if another player declares a reach midway through.
However, I've come to realize that even if I cry, I can still get away with it, so lately I've stopped being afraid of my arms getting shorter.
If you continue to defend your hand tenaciously after calling a tile, you increase the chances of other players making a mistake.
In Totsugeki Tohoku's book "New Scientific Mahjong," the boundary line of judgment regarding whether or not to call a tile to achieve a ready hand is shown.
Factors to consider include whether or not there is a Pinfu (no-points hand), whether the one-shanten (one step away from a winning hand) is a perfect shape, a double-sided wait, a double-sided wait plus a two-sided wait, or a guaranteed awkward shape, and the potential score if you call a tile.
The conclusion will vary depending on the combination of these conditions, but there are some guidelines indicating when you should aim for a pon or chi.
I'm not trying to advertise, but since it's a highly likely criterion for encounters, those who are interested should buy and read it.
Reach rate
Looking at players in Mahjong Soul, the reach rate seems to fall somewhere between 16% and 22%.
A high reach rate leads to a higher winning rate and higher point value, while the rate of calling tiles decreases relatively.
また、ノーガードで打ち合う展開が増えるため、放銃率も高くなっていく傾向があります。
In Mahjong, the "Riichi" hand is quite powerful.
If you declare a preemptive reach, you will almost always end up with a positive balance for that round.
To put it in extreme terms, if you keep getting a reach before your opponent, your performance will automatically improve.
For players who primarily play closed hands, the goal should be to raise this to around 20%.
However, if you are leading from the second round onwards, it is important to refrain from making a risky, all-out attack by declaring a reach, and instead play defensively to get through the round.
You don't need to be too nervous during the East round, but in the later stages, the last-place player often doesn't fold and attacks, so you need to be especially careful about when to push or fold.
To increase your reach rate, just like to increase your winning rate, I think you can improve it by learning tile efficiency while playing with a closed hand and knowing when to push and when to fold.
rank distribution
This is the percentage distribution of the number of times each of the top four places occurred.
The ideal arrangement is an inverted pyramid, with 1st, 2nd, 3rd, and 4th place.
If this is distorted, you may need to improve your ability to push and pull or your situational awareness.
In terms of average ranking, it seems that many people are ranked around 2nd or 3rd to 2.5th place.
However, Mahjong Soul and Tenhou have ranking systems where coming in last place is a significant setback.
Therefore, it could be said that it's somewhat unavoidable, even if the proportions are a little unnatural in order to avoid coming in fourth place.
If your fourth-place rate is less than 20%, I think you can become a Mahjong Saint by playing more games.
To improve, you need to understand the point difference between yourself and other players, consider how many points you can afford to give to whom, and adjust your playing style according to the score situation.
Since it's not realistic to target a specific player directly, try to be in a situation where you can win by calling "Ron" on anyone, or where you can win by drawing the winning tile yourself—a situation with the easiest possible conditions.
However, you should be able to play normally until at least the first round of the South wind.
If your score is extremely low, you might be inclined to push forward, but if you still have your turn as the dealer, refrain from being reckless.
There are quite a few opportunities to avoid last place when other players move sideways.
If you're in a situation where you're about to lose and a promising hand comes along, you might be tempted to jump in, but if you still have your turn as the dealer, take a moment to calm down.
Furthermore, if you are in the lead from the second round of the South wind onward, you need to pay close attention to how you play.
You don't need to go along with the last-place player's riichi, and you can rest assured if the point difference is small enough that you won't be overtaken even if the second-place player draws a mangan.
If you are in first place, it is very effective to hold onto plenty of safe tiles, and if you have a winning hand, it is very effective to use a concealed hand (damaten).
Just because you have a two-sided wait doesn't mean you should declare a ready hand; doing so with only a ready hand can lead to a painful loss when your opponent declares a ready hand.
If there is a significant point difference, it is effective to extend your lead and apply pressure, but if the score is close, it is worth considering.
The situation where the leader joins the competition is the most welcome development for the other players.
Even if another player winning would jeopardize your chances of first place, there's no need to deliberately do something that would only benefit those in lower positions.
Be especially careful in the final round.
The conditions often change depending on whether or not you place a reach stick.
If you have few points remaining, it's not a good idea to win with a low-scoring hand in the late game, such as in the South 3rd round.
However, if you have a bad starting hand, slowly building your hand will only make things worse.
With the idea that it's enough to ease the conditions for the final round, it's also important to win with low-scoring hands.
(If gaining 1000 points can reduce a 4800-point deficit to a 3800-point deficit, then gaining 3900 points in the final round would be sufficient, which makes sense.)
こういった状況判断力は、和了率や放銃率といった数値には表れにくい力です。
Even if someone's individual performance doesn't seem particularly outstanding, if their average ranking is good, it can be inferred that they possess excellent situational judgment.
Added in May 2023:
A common mistake is,Choosing the option that maximizes your earnings per hand doesn't necessarily mean you'll achieve the highest average rank in a half-game.is.
A typical example would be in the last round, with you (the non-player) having 40,000 points and the second-place player (also a non-player) having 29,500 points, and the second-place player then declares a reach.
This is a situation where you can hold on to the top spot unless the second-place player draws a full house or you directly hit them with 7700 points or more.
If you have a hand that is Pinfu with one Dora, the best option to maximize your winnings in this round is to declare Riichi.
However, reaching this point likely lowers your chances of getting first place for the following reasons:
- By placing a reach stick, the point difference becomes less than 10,000, and the second-place player can draw a mangan and overtake you.
- The risk of directly impacting the second-place finisher increases.
- Declaring "reach" increases the level of caution from players other than the second-place player, lowering the winning rate.
If we consider this hand in isolation, it's true that declaring "reach" would significantly increase the score and maximize the overall profit for that hand.
However, you can intuitively understand that this reduces the probability of winning first place in this half-game.
(It's intuitively obvious → It means, "We don't have any data, but you just kind of feel it, right?")
Therefore, there is nothing inherently wrong with using the profit/loss per hand as the basis for judgment, but the important thing is how to adjust the judgment criteria when looking at the entire half-game.
As a general trend, the further into the game it gets and the more points you have, the more likely you are to play defensively or keep your hand concealed. Conversely, the fewer points you have, the more likely you are to play aggressively or declare a reach.
Average total (runs)
This shows the average score when winning a hand in mahjong.
When I watch Mahjong Soul, I get the impression that there are a lot of players with around 5600 to 6500 points.
A score exceeding 6000 points is considered high, characterized by a high rate of declaring riichi and a slightly lower rate of winning and calling tiles.
At around 5600-5800 points, the rate of declaring riichi is relatively low, but the rate of winning hands and calling melds is high, giving the impression of a fast pace.
However, if this type of player has a low rate of open calls, it could indicate a defensive playstyle that frequently uses concealed hands.
To increase your average score, it's important to focus on declaring riichi and aiming for specific hand combinations,An easier approach is to focus on the drum.is.
If you start with a bad hand, focusing on accepting dora tiles or red tiles can sometimes turn it into a hand that you can actually win with.
パッとしない配牌だからといって安易に染め手に向かっていては、放銃リスクだけが増えていきます。
Even with tiles on the edges of the hand, it's important to always be mindful of the possibility of drawing a dora tile. For example, you might hold onto a tile longer if there's a chance of it being paired with a dora tile, or prioritize accepting a 5 if a red tile exists.
When comparing tiles, whether or not they are close to the dora tile is a significant factor.
Being mindful of when you might draw a Dragon will greatly help improve your scoring potential.
While it's important to aim for high scores by focusing on hand combinations, it seems that pursuing a hand combination to the point of returning to a state of being one step away from completion is not very common, except for flush hands (and even then, I feel it's limited).
I think the ideal scenario is to keep tiles that have the potential to form a winning hand when discarding tiles, and ultimately, to be able to form a winning hand.
For example, if you need to discard either "34" or "78", I think it's better to keep "34" as it guarantees a Tanyao hand and leaves room for a red 5.
However, if you have three dora tiles and enough points, keeping "78" which includes a 9 that is likely to appear as a waiting tile may be advantageous (although there is the disadvantage that it becomes more difficult to transition to an open tanyao).
If you feel your average score per hit is low, you might want to reconsider your approach to accepting the "Dora" card.
If you're a player who prioritizes speed over dora, then you should focus on increasing your calling rate and winning rate to specialize in a speed-oriented playstyle.
You can learn more about the finer points of accepting Dora tiles in this area from G. Uzaku's "301 Selections of Standard Discard Strategies."
It contains many useful problems that help you incorporate perspectives you might not have considered yourself, so I think it will be quite helpful.
Number of winning rounds
This is the average number of turns it takes to win a hand.
Looking at the averages, it seems that most players are around the 12th round of the draft.
Those with a score below 12.00 tend to be quite fast players, with a high rate of calling tiles and a tendency towards lower scoring.
They are likely gaining the upper hand by actively utilizing open melds.
Players with a score of around 12.10 to 12.30 often have an average score exceeding 6000 points, and tend to be closer to a closed hand style with a low rate of calling tiles and a high rate of declaring riichi.
Conversely, I think it's quite impressive when someone has an average score of over 6000 points but an average winning turn of less than 12.00.
If your average score isn't particularly high, but your average number of turns to win is high (12.50 or more), you probably need to learn the basics of hand building and other fundamentals.
平均打点や和了率・放銃率とセットで見ることにより、その人のスタイルや実力が浮き彫りになると言えます。
To improve the number of turns it takes to win, you need to focus on basic hand building while being mindful of the number of steps required to complete a hand.
ただし、早ければ良いというわけではないので、和了率・放銃率・副露率とのバランスを見ながら考えていくのが大事だと思います。
Once you understand your own strengths and preferred style, it's a good idea to approach studying to improve your winning percentage in the same way you would study to increase your winning percentage.
A collection of articles on improving your mahjong skills, techniques, mental game, Tenhou (online mahjong platform), etc.
How to Use and Interpret the Mahjong AI "Mortal" | Bad Move Rates and Rating Guidelines, the Difference Between 4.1 and 4.0, and How to Choose Between It and the Official MAKA in "Jantam" (Jantam and Tenhou)