Mahjong Fundamentals 5: Riichi

In this article we’ll look at the single most important yaku in riichi mahjong: riichi. Riichi is unique, and different to all of the other yaku in the game, because of the nature of making a riichi bet, and all of the other rules around it.

Let’s look at some of the upsides and downsides to riichi:

Benefits of Riichi

Positive 1: It is a yaku

The biggest and instant benefit of riichi is that you gain a yaku worth 1 han. For hands below 4 han this doubles your score, and allows you to win. For any hand, so long as you can get to tenpai while it is closed, then riichi will allow you to win. This is especially important for hands where we start with a number of dora tiles and no obvious yaku to push for. Riichi dora 3 is just as easy to go for as tanyao dora 3 if the tiles in the hand don’t suit tanyao.

Positive 2: Ippatsu and ura-dora

Riichi is worth 1 han, but actually we can usually consider it to be around 1.5 han on average, due to the chance for ippatsu and ura-dora. In particular, riichi can be an easy way to make an utterly rubbish hand suddenly worth a scary amount of points – hands like riichi tsumo ura-dora 2 go from 1000 points to a mangan with riichi. It’s not a reliable source of value, but it is often easier to get ura-dora from riichi to increase value than it is to try and incorporate another yaku and still reach tenpai quickly.

Positive 3: You slow down your opponents

In riichi mahjong, the first player to win each hand is the only one to get points. This means that the ability to slow our opponents down is very valuable to us, as it increases the likelihood that we’ll be the one to win. A key strength of riichi is that the riichi bet is scary to our opponents, and they are likely to fold or at least slow their hand to avoid dealing into you. This increases the amount of time we have to draw our winning tile, or at the very least means the game might end in a draw (with you in tenpai) rather than someone else winning.

Positive 4: You can push specific tiles out

This is a very minor benefit compared to the other three, but in situations where we have a tricky wait, such as waiting on an honour tile, or apparent suji (we’ll cover suji in the next article), it can encourage players to discard our winning tile and deal in. This is definitely not something to explicitly aim for very often, but it’s a nice side-benefit when it does happen, and it can sometimes tip the balance in favour of calling riichi on hands we might otherwise not choose to.

Downsides of Riichi

Negative 1: You must pay 1000 point deposit

When you riichi you need to pay 1000 points. Except in circumstances where points are very close, this is pretty negligible – 1000 points is the minimum you would pay in the case of a ron, and it’s less than what you can quite often end up paying through one of your opponents winning with tsumo.

You also gain the 1000 points back if you win, and given that being the first to riichi means you’re probably the first to tenpai, you can consider it reasonably likely that get your riichi deposit back, especially if you have a good wait.

Negative 2: You alert your opponents to the fact you are in tenpai

This one is a reasonably big negative, especially in higher skill games where your opponents are more likely to defend. For hands that are already big (i.e. approximately mangan or better), we should look not to riichi because we want to maximise our chance of winning. However for other hands, the additional value of riichi outweighs the lower chance that our opponents will deal our winning tiles.

Negative 3: You can no longer improve your hand

A lot of people focus on how you can’t change your hand if you draw tiles that might improve your wait or give you extra yaku, however in reality this usually not that big of a problem. Typically with a hand that is in tenpai there are very few tiles that will improve it or add value, and most of the time the chance of drawing an improvement tile is comparable to the chance that you’ll draw your winning tile.

Most of the time we are better off simply declaring riichi, because the backfire of drawing possible improvement after riichi isn’t anywhere near as bad as the backfire of drawing your winning tile if you didn’t riichi. Plus, declaring riichi is an immediate way to improve your hand!

Negative 4: You can no longer choose not to discard a tile

This is typically an important downside when there are very dangerous hands already on the table (e.g. a dealer riichi, or a dora yakuhai pon), or when you have a reasonable lead in 1st place and do not want to deal in by ron.

Most of the time, however, the situation on the table will not be that dangerous, and if you’re the first to riichi then go for it. There is a possibility that things may become dangerous later, but it is usually best not to worry about that possibility. Once you riichi, you become the dangerous situation for your opponents, and that is the better position to be in.

When we should riichi

At a glance, you might think that you have four pros and four cons, and riichi is a fairly balanced bet. In reality though, the strength of the four positives to riichi massively outweighs the four negatives to riichi, and riichi is almost always the right option to take when you get the opportunity.

The answer to the question of when riichi is a good idea, is pretty much “almost always” if we are the first one to riichi. Especially in lower skill games where opponents are less likely to defend, the number of times where riichi is a bad idea is tiny in comparison to the number of times where it is a good idea.

If one or more of the following is true, you should riichi if you would be the first player to do so:

  • You have a good wait
  • You are dealer
  • You have at least one han other than riichi

A “good wait” usually refers to a two-sided ryanmen wait or better. Waits on a single tile (such as penchan or kanchan) are bad waits. Waits with two pairs (shanpon) are also usually considered bad waits, though if one or both of the pairs that you’re waiting to complete is an honour tile or a terminal then it can be seen as an OK wait for riichi.

It is common for players to overthink the riichi bets, and to try and push for more value, or a better wait, but this is usually a negative choice overall. It is also not a good idea to ‘wait and see’ with riichi – if the above rule is met and we can be the first to riichi, we should take the opportunity and riichi immediately. Waiting to see if the hand might improve before riichi is typically a bad move, and insta-riichi is almost always a better choice, as it makes it harder for our opponents to reach tenpai as well.

When we shouldn’t riichi

Taking the rule on when to riichi above, it follows that we shouldn’t riichi if:

  • We are not dealer
  • and we have a bad wait
  • and riichi would be our only han

In general situations, the only time we shouldn’t riichi is when all of these points are true. Being non-dealer with a riichi only hand and a bad wait is one of the few common situations where the positives of riichi and the expected hand value aren’t high enough to justify the negatives of riichi. In this case we should usually not sit in tenpai without riichi (known as damaten, or dama for short) and hope we can tsumo, but instead look for other ways to add value or a better wait to the hand and break tenpai if we need to.

There are a few other times where we might want to not riichi, such as:

Somebody else has already called riichi

If we are not the first to riichi, then things can become quite tricky. Whether it is a good idea to riichi or not is down to a large number of factors, however generally:

  • Do not chase against a dealer riichi unless your hand is at least 3 guaranteed han after riichi (4 han if one is from pinfu) and a good wait
  • Do not chase against a non-dealer riichi unless your hand is at least 2 guaranteed han after riichi (3 han if one is pinfu) and a good wait

Other factors can come into it, for example it is often better not to riichi yourself if one of your winning tiles appears safe to the existing riichi. The point distribution at the table, and the current round number, should also factor into whether you choose to push or try and fold. If in doubt, if someone has called riichi before you, it is probably best to play safe and fold.

The hand is already expensive

There are two mechanisms at play here. One is that riichi adds proportionally less value to an expensive hand – in particular going from 4 han to 5 han, or from 6 han to 7 han will not make any difference to the value. This lessens the impact that riichi has for increasing our hand value.

The other effect is that our opponents are more likely to defend and our chance of winning will decrease. If our hand is already good, then it is better to maximise our chance of winning, than to be greedy and try and increase the value further.

As a general rule, do not riichi if your hand is already mangan or better

We have a large lead

Typically if we are ahead by a long margin (e.g. more than 15000 points with only 2 or 3 rounds left), it is better for us to play safe and to avoid riichi. Calling riichi leaves us exposed to dealing into large hands, and also pays an extra 1000 points to the winner if we don’t win. If we already have a comfortable lead, then avoiding dealing in is more important than chasing extra value, so we should instead be cautious and not riichi.

We might also not want to riichi if there are potential tiles we can draw that might greatly improve the wait or value, however it is fairly rare in these cases that waiting is actually better than just being aggressive and calling riichi first. In general, if you riichi whenever you can be the first to riichi, then you will be correct far, far more often than you are wrong. So if in doubt, insta-riichi!

For more on riichi judgement, please read Chapter 7 in Riichi Book 1, or Puyo’s series on Riichi Theory.

In the next article in this series we’ll move on to considering defence, and what we should do when our opponents get the chance to riichi first!

Leave a Reply

Your email address will not be published. Required fields are marked *