Winning in riichi mahjong can imply a lot of hard work, and sometimes it's hard to tell whether you're actually a winning or a losing player. What type of mahjong do you play? Winning mahjong? Losing mahjong?

If you're reading this blog, I assume that you want to start playing your best winning mahjong. I also assume that you've been playing riichi for long enough to know that you cannot (and will not) win 100% of the time.

At the same time, no matter how bad you are you won't lose 100% of the time either.

Skilled pro players only take 1st place about 30% of their games, which in turn means they're "losing" 70% of the time.

In fact, most strong players unavoidably end in 4th place around 20% of the time. That's once every five games.

Strong players win 30% of the time, but they still lose 20% of the time

RISK vs REWARD: The difference between a Winning Mahjong and a Losing Mahjong

Suppose for a second that you have access to three choices that will always get you the following results every game with the following probability:

  • Choice A
  • 1st place 30% of the time
  • 2nd place 20% of the time
  • 3rd place 20% of the time
  • 4th place 30% of the time
  • Choice B
  • 1st place 20%
  • 2nd place 40%
  • 3rd place 20%
  • 4th place 20%
  • Choice C
  • 1st place 35%
  • 2nd place 0%
  • 3rd place 0%
  • 4th place 65%

The doors line up before. You knock at one... which one is it? What choice do you make? I will give you a minute to consider your answer.


If you answered A, you’re wrong. I'm sorry...

If you answered B, I have to tell you... you’re wrong as well.

However, if you answered C… boy, you’re so wrong, too!

So... all answers are wrong... what gives?

The only one right answer to this problem is: WHAT ARE THE RULES? "What is my win condition?"

Depending on your winning condition, i.e. on the rules you’re playing, each of the above options can become the ideal choice.

Yes, this was a trick question. But an important one.

Examples of Different Mahjong Rules

For instance, at the final table of a tournament where 1st place wins 200 point$ (and the fancy title + bragging rights), and then 2nd, 3rd and 4th place all get 30 point$ each, then surprisingly Option C becomes the best choice.

I saw myself in this exact situation during the Shinjinousen Finals.

Especially for once-in-a-lifetime events, since you only get to play so few final tables in your life, it’s especially important that you choose the option with the highest probability for 1st place, regardless of all the other placements. 30-70 is still a good gamble.

Depending on your winning condition, the ideal choice will be different!

However, if you chose option C all the time in NPM League Games, where you play 80 games a year, and the point distribution is +50 +10 -10 -30, you wouldn’t be very successful. Unless, of course, it was the final game and you needed a 1st place to win.

Under NPM League rules, choice A would be optimal and preferred even over option B, because the bonus for 1st is much bigger compared to the penalty for 4th.

It's all in the rules. Let's run a thought experiment using the NPM League rules as an example.

Winnings Simulation

Suppose 100 games are played:

  • A: 50*30 + 10*20 – 10*20 – 30*30 = +600
  • B: 50*20 + 10*40 – 10*20 – 30*20 = +600
  • C: 50*35 – 30*65 = -200

That would be the EV (expected value) for each playstyle, as detailed above.

“Wait! A and B are both +600! You said choice A would be better than choice B. So what gives? Aren’t they just equally good? They have the same EV so they must be just as good!”

Yes, of course... if you played an infinite number of games. For league games, however, professional players only play 80 games a year. And even fewer for the lower leagues.

Since just the people on the top of the ladder move on, you want to prioritize the chance for first place, all other things being equal.

When the minimum score for progressing to the next league eventually comes down to “you need to have +200 points to pass the other player ahead of you and promote to the next league”, there will be no difference between holding 199 points and having 0 points.

On the other hand, when you already have a good lead in the scoreboard, switching to B and reducing the risk of 4th is an excellent choice as well.

However, keep in mind a few tournaments don’t add that huge bonus to 1st place, and might adopt a +20 +10 -10 -20 point spread alone instead. In those cases, B suddenly becomes the strongest option.


  • A: 20*30 + 10*20 – 10*20 – 20*30 = 0
  • B: 20*20 + 10*40 – 10*20 – 20*20 = +200
  • C: 20*35 – 20*65 = -600

Other Mahjong Winning Conditions (e.g. Online)

How about Tenhou (the internet mahjong server) for example? Or MahjongSoul?

Your best decision might change according to the point spread of the game you play.

Mahjong parlors, championships, etc., are all different. You have to find your winning condition and play for it.

1-day tournaments, where you only play 5 games at most, are the perfect place to go all-in most of the time since there’s no risk of losing anything (5th place and 48th place likely all hold the same value) and you can just keep signing up to one-day tournaments over and over again over the year. There is no shame in getting last place at the end either. It's not about how many times or how hard you fail, but about how many times you succeed.

For example, Tenhou championships are made in a way that you need to win 5 games in a row to have a shot at winning the entire thing, and you can play any amount of games while the championship is going on. That’s the perfect time to use something like strategy C and increase your chance for 5 wins in a row.

Winning in Mahjong Parlors

In mahjong parlors, you can either be more conservative or more aggressive, depending on the rules.

One extreme and odd example, the ruleset in Killer Tune (Umeda) right now gives 2nd place enormous value (+70 +30 -40 -80, rake:20 per game). There, falling from 1st to 2nd is only -40, while dropping from 2nd to 3rd is a whopping -70 (almost double).

At the same time, climbing from 3rd to 2nd is a good +70 while falling to last is only -40.

This means players are less inclined to chase 1st place if it will risk them falling down to 3rd, but a lot more likely to fight for 2nd place.

If you think winning at mahjong is about "getting a lot of 1st places", you need to reconsider and think again.

To win in mahjong...

You cannot and will not win every game. So you have to understand what your win condition is and play according to it; adapt to it. Don’t play the same way under every setting, and don't follow statistics without context.

Especially, don’t be afraid to end up 48th in a 48 people tournament just to prioritize a play style that will only make it more likely to end 10th most of the time. Play for the top, even if it means falling to the bottom - and don't fall victim to Tenhou Syndrome. (Unless, of course, the rewards for 10th place are well worth it!)

Mahjong isn’t a game of discards, it’s a game of rules. The optimal play will vary greatly depending on what you’re chasing after. So find your win condition, measure risk vs. reward, and make your choices accordingly. “Winning mahjong” is all about making the right choice under the proper rules.

(PS: If you would like to discuss your strategy before entering a tournament, click the "Lessons" button and schedule a meeting together with me and I'll gladly guide you)

Good luck, and choose your strategy wisely from now on!