Consistency

Theory, theory… I can’t really write any. It’s all so very theoretical. Instead of that I can give you right here right now, today, a piece of advice that is sure to help most beginner and some intermediate players, and it is about consistency.

What I mean when I talk about consistency in your discards is… be faithful to your logic/style/beliefs/school of thought/theories/experience/that sixth sense for the occult that always helps you win/whatever. I’m not saying you should stick to what you know and never change it, you probably should improve your game every time more, but what I mean is… if you’re going to suddenly change your whole game, don’t do it in the middle of a game. Or even worse, in the middle of a hand. If you’re digital, don’t suddenly go occult because you had a sudden hunch; or if you’re occult, don’t go digital just because you’ve begun doubting yourself due to that mangan you dealt into the hand before.

It’s painful to see a guy switching between aiming for ittsuu or chanta three discards in a row, just to get himself into furiten later on because he couldn’t decide on one when he had to. You should remember that every one of your discards should have a meaning and a reason to it. If you go one way or another “just because” then your results will be substandard.

It doesn’t matter your belief, your system or your method, all discards can (and will) backfire at one point. If you panic at that point and attempt to play it off as if nothing happened, switching courses yet again, in a half-assed way, then your results will be half-assed too. Instead, you should remember the reason why you chose to go one way, and whether straying from that course will be beneficial or not in reality.

If at one point you considered a path to be inferior to another and you destroy that shape, when that shape comes back (if it comes back) it will be a lot weaker than at the time you destroyed it. So you must really think twice whether you wanna switch roads again. Is it really worth of it?

Also, if you happen to lose consistency because you are starting to doubt, don’t! Make sure your discards have a sound reason and intent for them (it doesn’t matter if they’re right or wrong, superstitious or logical, they just need to have a purpose to carry out) and if they manage to carry out their purpose, then you’re winning in a way!

When you learn better strategies, you’ll send your tiles to accomplish better purposes, but don’t just dispatch a lot of your units into quests just to cancel them halfway through.

If anything, think before you discard. Most of the time it’s as simple as that.

31 thoughts on “Consistency

  1. Also, reading Sasaki Hisato’s book, I now realized this article covers like two of his lessons from the first chapter!
    (“Go through with the plan you chose” and “You can’t hit two birds with one stone”)

  2. :)

    OT: I noticed something interesting yesterday when looking at gamelog stats from Houhou tables in the Tenhou analyzer… good players have a bigger winning edge in Hanchan, as compared to Tonpuusen.

    Perhaps this is obvious, but I was surprised at how big the difference tends to be. I’m inferring from this that the effect of skill is bigger in Hanchan.

    What I really want to look at is stats from Kyougi rules, specifically games without Ippatsu/Ura. Would these stats show that the effect of skill is even higher?

    Looking at some of the new translations from Puyo’s blog about when to Riichi, it’s clear that very often it’s a no-brainer to Riichi because of the Ippatsu/Ura. Surely it’s a much more nuanced decision to Riichi in Kyougi rules.

    What about Noten Bappu Ari/Nashi…. the questions are endless…

  3. Last time I checked, Kajimoto was doing a lot of broadcasts playing kyougi rules online.

    For renmei rules, since they also use a sort of shizumi uma, it’s crucial to be over the starting points. If you’re the only one under the starting points in the end, you get a -12 which is pretty much like dealing into a dealer mangan or non-dealer haneman.

    Every point counts. Hands like pinfu tend to be not worth reaching without dora. Seven pairs and hon itsu are common currency.

    Meh, I’ve seen the 10dansen broadcast and part of the hououisen tournament broadcast in niconico, and well, they’re different rules, but it’s not that much fun or “of skill.” I think those players would do just as good in normal red dora rules.

    Lots of broadcasts on competition rules in niconico.

  4. Is it a consensus that those rules are not ‘of skill’?

    I totally buy the argument that they are not as exciting as Ippatsu/Ura mahjong, but I would be surprised if they don’t increase skill at all.

    Purely the fact that “to riichi or not to riichi” is a more complex question, or that “every point counts” suggests that skill would have a slightly larger effect than usual.

    The fact that 90% of riichis in normal rules are mangan or higher seems excessive in some respects.

    I could imagine that 5200-point hands suddenly become much more important in Kyougi rules.

    If the overall hand values are lower though, it seems like Noten Bappu should also be removed. Where can I watch the nico-nico broadcasts and can I do it easily without being able to read Japanese?

  5. Few organizations use no ippatsu and no kan/ura dora. Only MU and JPML uses it, saikouisen used it but they now label it as “classic rules.” Some people believe JPML uses it so they can fix games without kan or ura dora interfering.

  6. Pingu, it’s not about understanding kyougi rules or not, but about understanding mahjong and its functioning or not. If you understand how mahjong works, you can make a winning strategy for any kind of ruling after a few games. The skilled player outplays the beginner in any ruleset. Also, pro leagues are played among pros alike, so it should already be a competition of skill as it is.
    Also, as a piece of information, Saikouisen Classic Rules (Saikouisen’s kyougi rules) don’t have noten bappu.

    No, not rigged as in putting the tiles you want in the wall.
    Let’s say there are 4 players in the table, A B C and D. A and B are old players who always alternate the champion title between one another, and C and D are younger players who made their final table for the first time. With some bad luck, B has descended in points to a point he is in a position where he cannot win anymore, and the game has turned into a battle between A and C.

    Exaggerated example 1: Last hanchan(s), B starts winning cheap hands from C and D, pushing to his riichis and pursuing him with a low scoring but fast hand every time he can. He is decreasing C’s chances to win greatly, as he barely even allows him to finish a hand, and even when he does win off of B, he doesn’t need to worry about his hand being high scoring (no ippatsu, or ura or aka to boost him either). If anything, he saved A from a direct hit. Also, if A is also playing to win, A’s chances to win go up. A wins, and B can try again next year; the challengers disappear into the abyss. A would have done the same for B if roles were reversed.

    Exaggerated example 2: A is far ahead and wants to skip this round. Just according to plan, B makes a cheap riichi so that A can deal in, letting him know he is in tenpai. A then proceeds to sashikomi. Do you imagine the danger or problem that could turn into if he had ura dora or ippatsu? Or how hard it would be for him to reduce his hand value if he had red dora? Or if someone made a kan, and he tsumo’d with quite a few kan dora and kan ura when A is dealer.

    If A was losing to C, then B (and A, too, of course) has a few ways to help bump A into first. But if C is losing, he has little chance to make a come back. Of course you can’t help it if he gets tenhou, or a kokushimusou in the third discard, but I’m sure you can crush his hopes in many ways.

    There have been incidents similar to my examples in a few games in the past.

  7. Also, the uma+oka in normal games is so high that it would be very inconvenient if someone who “wasn’t supposed to win” got a first place at the same time the “guy who was supposed to win” got last place. So, switching to a more moderate uma is far more convenient if you want to rig games.

    I’m not saying anyone is rigging games. In fact, now there are a lot of guidelines to follow to determine if someone has been cheating or not, and it is the job of the judges to determine that. I’m just saying that if you wanted to rig games that way, certain rules (like kyougi rules) make it easier than others, as there are less external factors that can screw your plans up.

  8. Other stuff in kyougi rules makes it easier, too. No tobi, for example (you can keep giving away points to your favorite). Or the fact that there is no miraculous kazoe yakuman (stuck at sanbaiman), all green doesn’t count without the hatsu, or even that you cannot get multiple yakuman.

    If you really, really wanted to cheat with someone else, it’s evident which one is easier to do.

    I’m not saying people don’t or have never cheated on regular rules either; people have been doing it and probably will keep doing it for a while, too. In most movies you see, there are a few guys ganging up on a poor fool to steal his salary, and they aren’t necessarily playing kyougi rules. People who want to cheat, will cheat no matter the ruling. And people who want to play mahjong intellectually, will play their best game for you in whatever ruleset you put them into.

  9. I see what you mean about game fixing!

    Besides that, though, I’m being misunderstood.

    I’m completely aware that a skilled player does well regardless of ruleset, etc. However, that doesn’t change the fact that certain rulesets give a BIGGER edge to skilled players, some give a smaller one.

    In the example I gave earlier in this thread, it seems obvious from Tenhou Analyzer that skilled players have a significantly bigger edge in Hanchan as opposed to Tonpuusen. This is nothing to do with greater variance, either. Purely how MUCH more they are able to place well on average if they are highly skilled.

    The same is true in Poker. Some rulesets are known to favor skill more than others. Indeed, a good player does well regardless of rules, but that does not invalidate the point at all. They would simply do even better with more skill-favoring rules, that much should be somewhat obvious.

    On this basis, I have yet to hear a convincing reason as to why Kyougi rules have no effect on the skill-edge. Surely someone must have complied stats on this, comparing the results of a given player in Kyougi vs Normal games?

  10. …and in case my Analyzer example is not clear, try downloading gamelogs from top-rated Houhou players and compare their detailed stats on Tonpuusen vs Hanchan.

  11. You are aware that some people are just better at hanchan than tonpuusen, right?
    Most people play only hanchan, which mean they are not as skilled when it comes to push-pull decisions in tonpuusen (because they lack the same experience they have in hanchan with all the number of games they played there). Parting from that point, surely most people won’t do as well in tonpuusen if they only casually play there. But if somehow tonpuusen was the main current, and people played hanchan only casually, I am convinced the results would be reversed. This kind of reminds of meta-gaming in card games, kinda.

    That, if you refer to average placings. If you refer to point totals, it’s obvious that, as hanchan are longer, the final scores (and the resulting point difference) will naturally be bigger. Also, expertise isn’t simply measured by “point numbers” in a game where you’re competing -for placing-, such as tenhou.

    Also, what do you think the result would be in a match between a person who was played kyougi rules for years, and a far better player who has never played kyougi rules before, but he is the top ace in aka ari games? Clearly, the kyougi player has an edge. But once the better player has learned the differences in kyougi rules, before long, he will outplay the other one. And that being the case, what is the point in having them both play kyougi rules? Wouldn’t having them both play a string of sessions in normal (regardless of the existence or absence of aka dora) mahjong rules display the difference in their skill as well with an appropriate margin?

    It’s like saying professional soccer should be played with smaller goals.

  12. I am referring to average placement, and the phenomenon seems to be sufficiently widespread that it’s not just an issue of one player’s preference versus another.

    I’m not sure why you keep sidestepping what I’m getting at. Sure, your points are absolutely valid…

    …but they don’t go against what I’m saying. Both lines of reasoning are true. Some rulesets, from a mathematical point of view, allow a bigger skilled edge- regardless of experience level and player preference.

    Of course, there’s a limit on what is fun and sensible (i.e. football with tiny goals) but that doesn’t invalidate the question. Moreover, Kyougi rules do exist.

    This is precisely why I’m interested to know how big of a difference they make. Could be big, could be small. Despite the fact that you can have more experience with one ruleset versus another, it cannot be denied that certain things are inherent in the game parameters- and not the player.

  13. People who are really good at tonpuusen have a better average placement in it than people who… well, aren’t good at tonpuusen. It’s pretty much self evident. Even if they are both houou level players. So, when you indiscriminately take a number of games from both types of game, and compare them against one another, obviously the sample with the highest amount of skilled players in said mode (in this case, it would be hanchan) will score relatively higher than the other.

    (Do take into account that while Tonpuusen may have its negative factors (you coincidentally deal once into a dealer mangan and you’re pretty much over), hanchan has luck-increasing factors of its own as well (the guy in dead last has a lot of opportunities to build a bullshit comeback hand to overtake you)).

    Also, did you know the top player in Jansou mode tables in tenhou can barely stay over 7d R2000 in normal tenhou, even though he plays against all sort of houou players in there? Does it mean the normal tenhou modality, including houou tables, are more luck-influenced than jansou mode? It doesn’t. He’s just better at one, than the other. And the other players, are just better at the other one. It’s more like looking at -where- you are assigning your stat/skill points than looking at how much HP you drain from a certain monster when you hit it.

    Also, I don’t think I’m the only one who finds kyougi rules to be as fun as football with tiny goals. Sure, running around the field is always good fun, and it’s nice to get to kick some balls, but in the end, you’re just making a normally fun sport more painful to play/watch, without increasing the “skill” level in a matter that makes it worth of it.

    Players who win at kyougi rules, are good at kyougi rules. Just as expected.

  14. Also, they all require experience and skill, obviously. But they require experience and skill for its own mode. When you play with people with your exact same skill level at that ruleset, all things will just be the same again (at a 25% chance each). They end up balancing out in proportion.

  15. Instead of players with the same level, if you would rather like to compare the skill of a beginner with a skilled player, then sure, that margin would indeed be bigger in kyougi-like rules, but only because there are -more chances to make mistakes-. Against player with a similar skill level, you’re back to very marginal results, easily affected by variance.

  16. It doesn’t seem like you are drawing sound comparisons.

    The jansou mode has so few games logged that I don’t think you can draw any firm conclusions about who the best player is.

    The reason I am asking this is because if indeed Kyougi rules do to some extent reduce variance and increase skill, then they are a good choice for tournaments.

    I want to know if this is the case. Certainly, the amount of variance in normal rules is too high for most tournaments played over one day to be a reliable indicator of skill.

    I think it’s normal for a player of any game to want to know what the best indicators of skill are. I’m also very interested to know the nuanced ways in which the game is different across rulesets. Some changes are obvious, some less so.

    I can’t say I understand why it’s a bad thing to be interested in this. Also, if Kyougi rules are as awful as you say they are I don’t see why people would be making popular broadcasts about them.

  17. Also, this logic seems faulty to me:

    “…they all require experience and skill, obviously. But they require experience and skill for its own mode. When you play with people with your exact same skill level at that ruleset, all things will just be the same again (at a 25% chance each). They end up balancing out in proportion.”

    Or rather, it sidesteps the issue. If one ruleset provides a greater differentiation of skill, then to play against players of your level is a more elite distinction of good play.

    Your assumption is only accurate if the skill level required to be a top player at Kyougi is the same as for Normal rules.

    If you look at the top rankings at Tenhou, they are very volatile. Would Kyougi rankings be less volatile, and would there be a deeper spread of player ratings? Would the best if the best stand out higher, statistically speaking?

    That’s what I want to know, and it’s a valid question in terms of assessing one ruleset against another.

    Whether or not one likes them personally is another matter.

  18. If that is the issue, I can answer in quite a simple way:
    I don’t think all of the people higher up in renmei’s standings are the best of the best, as I don’t think some of the people in lower leagues are worse than the people higher up.

    Also, you say there are not enough games in jansou mode, but there aren’t many kyougi games to analyze either. As for the reason why people watch the broadcasts, they are official broadcasts by renmei (hence, their rules), and mahjong fans, players who think they can beat the pros playing, players who want to learn from the pros, and people with diverse reasons sit to watch. As to why I watch them, they make for good conversation topics afterwards. You get to read Fukuchi’s opinion on how some of the people playing there are belittling other players while doing some very bad moves themselves, you can see some drama, and you can also know what the people in 2ch are talking about when they accuse someone of ganging up on a player. Reasons are many. All in all, it’s fun. But it would just be as fun and successful no matter if they were playing kyougi, aka ari, or wareme de pon rules. You can see that’s the case, with Saikyousen being just as (if not more) successful in attracting viewers.

    The downside on making tournaments kyougi-rule based, is that while you may think that if they actually are rules that allow for less variance then they’d be the most appropriate choice for competitions, the only thing they do is force players to practice kyougi rules in order to keep up with the other players. If the point was putting all players in the same level, we should all just play zung jung, with their “more balanced” point system and whatnot. Or invent a new mode every year. Doable, but just a nuisance. There’s a reason why only renmei (and now janryuumon, after some sort of deal with them) use kyougi rules. I’d say it’s for custom. Some would say it’s for game rigging, and even Yu Takehana has had problems with that while he was in the league, but that’s no reason to discredit (or credit) that set of rules.

    I personally don’t think there’s enough reason to think kyougi rules imply more skill. They imply -different- skills.

  19. You’d have to explain how some people who cannot move up from D or C League for the life of them, suddenly get to final tables in Hououi, Juudansen, or Oui tournaments, despite all of them being the same ruleset (and viceversa).

  20. Well, except for C and D leaguers in Hououi. That’s the exception because they can’t make the final table without being in the top of A1. w

    But the other possibilities, I mean.

  21. This is mixing up two issues, I think.

    The whole issue with Renmei being corrupt has nothing to do with what can be inferred about the rules themselves.

    Also, how good any of the Renmei players are, whether A1 or D1, is meaningless. There are not that many people who play in those leagues (compared to a site like Tenhou) and there is no reason to believe that the best players join in the first place. The barrier to entry for ‘pros’ is not high.

    Also, the issue of ‘different’ skills for different rules is also besides the point. What I want to know is which ruleset, within itself, allows the biggest and most consistent differentiation of player skill. I still haven’t heard a reason to believe that Kyougi rules don’t increase the edge between skilled and unskilled players.

    Does Janryumon keep stats on Kyougi rules now?

  22. I think you’re missing the point when you believe playing kyougi rules has anything to do at all with playing regular mahjong.
    It’s like asking if playing basketball can measure the skill level among handball players (with its more complex rules, strategy and whatnot). If it can increase the skill required. Well, sure, it might, but even if it did, it’s a different game.

    No, Janryuumon does not keep stats of anything. In fact, they don’t even save your replays unless you pay for an in-game item to record it.

  23. And it just makes your question not any more valid than questions such as “What is the best sport?” “What is the most intellectual table game?” “Which requires more skill, Chess, Shougi or Go?”

    Different rules, different games, different way to do things. Irrelevant of skill.

  24. I understand where you are coming from, but to me that’s a pretty perverse way of looking at things.

    You are implying that different rulesets of Riichi mahjong are as different from one another as Chess and Basketball.

    Personally, I think different Poker variants can all be considered Poker. They are all very different and require different winning strategies, but that doesn’t make them entirely different. Poker is still poker, whether you are playing Texas Hold ‘Em or Omaha.

    I think the same is true for Mahjong. In this case, the difference between Kyougi rules and Free rules is not even as large as the difference between Hong Kong rules and Zung Jung, say. Nevertheless, I think it’s completely appropriate to call all of these games Mahjong.

    Moreover, some allow skill to have a greater statistical effect, some less. If Kyougi rules can do this, while still being closer to Free rules than something as different as Zung Jung, then I think they are a worthy adjunct to the Riichi rules we all love to play.

    It’s not clear to me why this is such a sensitive subject for you.

  25. It is not perverse, but natural. They -are- that different. Just in the same way you suddenly stop aiming for sequence-like hands in Zung Jung (because they pay less), you’ll find yourself doing the same in kyougi. It’s not a one or two rules change, but quite a few; you can’t compare them anymore in that way. If the riichi mahjong variant adopted was kyougi rules, why not accept more universal riichi mahjong rules? Instead of a set of rules rarely ever used by more than one organization, and whose ” ‘diminished’ luck influence” are still doubtful and arguable? There are a lot of incredibly great mahjong players out there; but making kyougi “the rule” and placing them in kyougi rules against other people when they go to tournaments? For what purpose exactly? The luck factor that is said to diminish the breech between strong and weak players doesn’t really affect the table any less than its counterpart.

    If people start with too good of a haipai, with all dora tiles and whatnot, there will still be hands over which you have no control in either version, while the guys with a bunch of dora or speedy starting hands keep taking the points. That doesn’t change much. It’s just the volume of the hands being won being smaller, but not the amount of luck affecting them. What you call “skill” is mere “adjustment;” adjusting yourself to the fact that now nouten bappu and riichi deposits are relatively more important, adapting to the fact that 2 dora in your hand is almost certainly a shoubu hand, as well as adapting to playing a different type of mahjong. And people who are adapted to it, may win, or not.

    Take the currently ongoing Tenhou Meijinsen for example. They are tenhou rules, with the only modification being 30k in, 30k out; Uma of 1-3, and no tobi. I wouldn’t say that if they played in kyougi rules instead, the results would have been less luck-influenced. And that was my personal impression every time I played kyougi rules as well, as well as I do when I’m playing others.

    I suggest you try them yourself and tell me whether you really feel like they diminish luck influence (and also, what you call “luck” and “skill” must be well defined, first. If you can manage to make profits in a highly luck-based game, you should be pretty damn skilled as well!). And if you feel like they’re an improvement, adopt them as your preferred way to play. But I don’t think they deserve to be made into the default for anything. There’s no merit in that.

Leave a Reply

Your email address will not be published.