The Beautiful Closed World of Shenmue III
Wednesday, 08 June 2022
It takes me quite a while to get around to playing games these days, which helps insulate me from jumping to kneejerk conclusions about what I've been playing. I like to have the time to engage in a game in the way that it requires, which isn't something you can do under the time pressure inherent to reviews. Thus, earlier this year, I completed my playthrough of Ys Net's 2019 title Shenmue III. It's a remarkable achievement on many fronts, not least of which is that it managed to pick up a franchise after an absence of a decade and a half and provide a sequel that is entirely in keeping with the aesthetic achievements of its predecessors. Yet many people have complained about the time they spent with Shenmue III, in one extreme instance lamenting that "Shenmue 3 is a Terrible Game and I’ve Wasted My Life".
Gladly will I concede that, as a commercial proposition, Shenmue III has serious problems... but those problems are the ones it inherits from Shenmue itself, and as a game that was funded by a Kickstarter pledging to provide a true sequel to 2001's Shenmue II, objecting that it is too much like the games that preceded it might rather miss the point. Frankly, if the flaws in an artwork are that some people do not like it, this really isn't as knock-down an argument as it may seem. Rather, to appreciate Shenmue III we have to understand why it is the way it is, how it fulfils the promises made by its creators, and why its beautiful closed world has more to teach about game narrative than most of its critics are prepared to allow.
Yu Suzuki's Shenmue series is a work of flawed genius, and it is not for the reasons its creator claims that they stand out in the history of videogame narrative. It's originator gave the original game the clunky genre title FREE, standing for 'Full Reactive Eyes Entertainment', supposedly to show the interactivity and freedom the player would have. Well, interactivity maybe - you can indeed spend hours meaninglessly opening each compartment in every chest of drawers, for instance. But 'freedom' is precisely the opposite of what a Shenmue game is about, and it is all the better as an artwork precisely because it is really quite uninterested in the player's freedom.
Suzuki-san and his colleagues often try to paint this series of games as a precursor to the open world genre. Indeed, in the Kickstarter for Shenmue III, the text expressly tries to make this claim:
Shenmue defied all convention and created the genre that later came to be known as "open world." An unparalleled level of freedom let you chose how you wanted to play.
But this is neither true nor fair. Not only does 1999's Shenmue not foreshadow or influence the open world games of the early 21st century, what it actually does achieve is artistically far more interesting than this claim would suggest. The open world genre crystallises in the Grand Theft Auto franchise, especially after its transition into 3D models from its sprite-based roots. As I have written about before, it is the playground worlds of 1985 (which were never exported to Japan) that influenced GTA, most especially Elite, although I still suspect Paradroid has a part of this tale to tell. Suzuki-san was neither influenced by these titles nor went on to influence the open world lineage that followed.
But so what? The open world genre may have become a commercial powerhouse, and certainly contains a great many games that enjoyed both huge sales figures and critical acclaim. But the open world is a narrative dead end in many respects... brilliant, and still evolving, but also highly limiting and increasingly stagnant creatively. To get stories into open worlds, designers have to seed the huge landscapes with signposts that push to set piece encounters, a technique I have compared to 'plot origami'. From a game design perspective, this is often brilliant - both Witcher 3 and Breath of the Wild have shown the immense appeal this can have when it is done well.
Yet the open world is a format that is fundamentally limited when it comes to characters. Precisely because the worlds are so large, characters either remain locked into a geographical slot they inhabit, or have to haunt the landscape as ghosts who can appear only when they need to miraculously reveal themselves to the player in order to advance the storyline. As a result, the open world's fevered desire to satisfy the player's desire to 'do anything' always risks becoming narratively flat... whatever interesting aspects of Breath of the Wild story I might have picked out in Zelda Facets, it cannot change the fact that the moment-to-moment activities the player pursues are basically a kind of bizarre subsistence hunter-gathering, collecting durians, horses, and weapons to fund the player's role as an itinerant trouble-maker.
Shenmue III is not an open world. Nor should it have been, because neither of its predecessors are either. This is most obviously apparent in the fact that each game consists of two different 'village' locations, as opposed to having 'dungeons'. This split into village and dungeon dates all the way back to the tabletop and Dungeons & Dragons, and was picked up by Ultima and Wizardry, and so spread into videogames. There are no dungeons in a Shenmue game, only villages - and each episode is about two such villages (although the second episode foreshadows the village that will feature at the start of part three). If Shenmue III were an open world game, you would be able to return to Bailu village from Niaowu. But you cannot. Once you leave Bailu village, the story has moved on and you have moved on with it. Shenmue III is a closed world.
Being a closed world affords enormous advantages for narrative in games. Precisely because an open world game features a vast landscape, the depth of conversation you can have with characters is necessarily curtailed. Open world games often do a great job hiding how shallow its cardboard cut-out characters are with lively dialogue and quips, but fundamentally every open world game saves its interesting characters for those narrative 'ghosts' who the player cannot access on demand, and populates the rest of the world with convenient stereotypes. Yes, you can run over anyone you like in Liberty City or San Andreas, and you can kill a bokoblin all over Hyrule. What you cannot do is learn anything about these characters, because they are not really characters in the literary sense at all. They are props to make the playground seem lived in, like the woman in the princess costume at Disneyland.
Precisely because it is a closed world, Shenmue III can be about place in a way that no open world game can manage. That's because human places - villages, towns, cities - are about the people that live there, whereas videogame places - inns and shops, dungeons and secret bases - are about the loot the player can steal and the advantages the player can eke out of them. To be fair, there's a little of that in Shenmue III as well - it is still a videogame after all! - but by building these games around pairs of 'villages', each Shenmue game evokes a sense of place that is rooted in the people who live in these places, and complemented by the player's character, Ryo, who is much more than just a blank slate like, say, the Master Chief. On the contrary, how much you enjoy playing Ryo will depend upon how much you are willing to become Ryo.
Here we come to the biggest complaint levelled against Shenmue III by its detractors: the kung fu fighting isn't good enough. But this is paradoxically not a game about kung fu fighting at all. Indeed, when you fight in the major battles, you will not use any of the skills you have practiced up until that point, because victory in these set pieces is determined by QTEs i.e. Quick Time Events, which is to say preset button sequences. Shenmue did, alas, invent the QTE, although the 1983 arcade game Dragon's Lair had already pioneered the form (and for detractors such as myself, shown how tedious it could be). If I have warmed to them over the last forty years, it is only thanks to the efforts of Shenmue... and I am still quite frosty about them.
In so much as it is about kung fu at all, Shenmue III is a game about practicing kung fu. This is not about the freedom to go and beat the living daylights out of ten thousand cardboard cutout characters creating the illusion of a city, this series is partly-yet-significantly about Ryo's restless youth gradually coming to terms with the commitment and patience required to master a martial arts. Thus the player is asked and expected, in every game in the series, to spend a significant part of their time performing repetitive actions to master the individual moves, or the elements of their form. Thus in the original Shenmue I spent a part of every day repeating sweeps and kicks, because as a game-player I wanted the advantage of levelling up those moves.
By Shenmue III, while that carrot still dangles, my engagement with the training exercises such as Horse Stance and One Inch Punch is practically meditative. Yes, as a player I am conditioned to want to complete my bars and gain higher levels. But as a visitor to this world, I am choosing to train at the dojo high in the mountain so I can watch the clouds drift lazily across the mountains as I do so, immersed in the beauty of this world which is categorically not reducible to polygon counts. So effective is the game at getting the willing collaborator into this state that it is possible, as happened to me, to be sad when the training exercises are completed, and no more points can be earned. The game has made me into Ryo, and given me the slightest taste of his impatience along with a parallel yet opposite flavour of what it means to practice a real martial art (or, for that matter, any other art): patience and commitment.
In this regard, it is a design flaw of the combat system in Shenmue III that you can assign moves to controller shortcuts and therefore execute them effortlessly. Having spent hours mastering the control inputs during sparring, we really ought to be tasked with executing those moves ourselves, not handing them off to an automated surrogate. I imagine this was a design decision intended to support weaker players, but since all the moves power up, even a button-masher who persists can get by without being able to execute the stronger moves. Letting them be triggered by an all-too-convenient shortcut control undercuts the focus upon training that suffuses the player's experience. I am almost ashamed that I made use of this feature at all, and if I play the game again, I will certainly try to avoid doing so.
So much is Ryo's kung fu training foregrounded in the game, that one of my favourite characters turned out to be someone who has really quite a peripheral role in the story: Su Zixiong. This jocular, overweight tai chi (taijiquan) instructor is found at the village square on the edge of Bailu village every day, training the young children in their moves. When Ryo meets him, he declares himself preposterously to be the Lu Bu of Bailu village, referencing the unbeatable warrior of the legendary Three Kingdoms era of China. Pressed on the topic, he admits to being more of the Zhang Fei of Bailu village - another legendary figure from the Three Kingdoms, and one that is a slightly more honest fit to this character, for all that it maintains a self-aggrandisement that is both harmless and charming in its excess.
Every day, I went to spar with Su Zixiong (although I think his name ought to be Zixiong Su...), even though he is a deeply minor character in the plot. My primary motivation for acquiring new move scrolls was not to learn new moves (I didn't need a wide range of moves in any Shenmue game), but to have something new to practice while sparring with my friend in town, since he declines to spar if you have nothing to learn. That I was able to form a surprisingly robust relationship with such a peripheral character could be dismissed as merely a quirk of my personality. But I do not think so. It speaks to precisely why the closed world of Shenmue III works: it is full of interesting peripheral characters, so many that whoever the player bonds with really is up to them. No open world can claim this, because the major characters are 'ghosts' who appear only when the plot requires them. Shenmue III offers a greater sense of place than almost any other game you might choose to mention.
I adored my time in Bailu village, which is among the most beautiful villages in any videogame to date, and easily my favourite settlement of the six to feature across the Shenmue series. I was literally saddened to leave, and as the plot inexorably pushed me to the point of 'have you done everything you want to do here...?' I found myself wanting to tarry longer in Bailu, to spend more time fishing her streams and lakes, to continue my training in her mountains - even though I could not, because I had acquired all the available move scrolls, and maxed out the bars for my training exercises. The game conspires through its design to force you to co-operate with its plot voluntarily, an extremely clever trick that perhaps is difficult for most gamers to appreciate.
There was another reason I did not want to leave Bailu: Shenhua's house. It is your home in Bailu for the entirety of your stay, and I found myself becoming absurdly attached to it. Not as a mere building, as I had done with the house Link purchases in Breath of the Wild, but as a genuine home. For to my surprise, I wanted to talk to Shenhua when I came home every night, even though the conversations were pointless if judged in terms of game benefit or plot advancement. As someone who has to write videogame dialogue as part of my job, I have very little patience for the so-so conversations most games have to offer. Many a time do I find myself clicking impatiently through weak dialogue. This never happened with Shenhua. I wanted to talk to her - I wanted to learn not only about her, but about Ryo through her.
One remarkable moment in my nightly discussions with my host came when I was given another strictly meaningless choice about how to respond to something Shenhua had said about her quite literally unbelievable ability to speak with animals. I made my comment, and clearly caused offense with my (and Ryo's) down-to-earth scepticism. I wanted to take it back, but I did not, not least of all because reloading the game was pointless (this conversation did not have plot consequences) and would not in fact undo the choice I had made anyway. The very fact that I even considered reloading to undo a 'meaningless' choice reveals the depth of meaning these conversations had to me.
The game is rather less successful with its second area, Niaowu, which although enjoyable suffers from not being able to include quite enough passers by to really capture the sense of a busy market town in rural china. Don't get me wrong, there is a great deal of character to Niaowu, with its vast array of shops - and Shenmue's signature capsule machines ("I love those!") - giving the player many corners to poke and prod. But I enjoyed learning about the herbs and plants of Bailu far more, and although you can still find these in Niaowu they have lost a certain part of their resonance in the market town, where the shops and their shopkeepers are far more crucial to the spirit of the settlement.
There is so much more I could say about the closed world - or perhaps, closed worlds, Bailu and Niaowu being quite distinct - of Shenmue III but let me close this critique with a reflection of the role that time plays in the player's experience, and the tension it reveals with the expectations of gamers. A great many players complained that you could not skip ahead, just as I know many players bitched about being forced to work the forklifts in the original game. This could not be further from the spirit of Shenmue. Just as Ryo is asked to learn patience, so is the player. While as a commercial decision, skipping time would be justified, the entire issue is parallel to the way that the quick travel of Breath of the Wild guts the brilliant horse riding of its joy and wonder. It is precisely because you will frequently have time to spare that you will find yourself playing pachinko, or dropping coins into capsule machines, fishing, or working the docks. Not out of the love of money or the other game currency, XP, but for what that XP is named after: the experience.
Likewise, those who complain that this should have been used to tie up the storyline and give players some closure are missing the point. Ys Net did not promise to conclude the story with this Kickstarter, they promised to make another Shenmue game. If this game had ended the story, they would have failed to have done so. This desire from players was in tension with the spirit of the Shenmue franchise, which is neither as swift nor as impatient as most players of games have become. On the contrary, in not only bringing back Shenmue II's delightful rogue Ren but in establishing both him and Shenhua as companions on Ryo's adventure, the game escapes the very real risk of being just a super-polite, gently paced version of Herge's intrepid boy adventurer Tintin. On the contrary, Ryo is so much more than just 'Japanese Tintin', precisely because he is Ryo.
Shenmue III invites us, once again, to become Ryo. How much you enjoy that experience depends to a great extent on how much you are willing to be Ryo, to endlessly train for fights you won't actually have, to talk and to listen to strangers who may yet become friends, and to obsessively drop coins into a capsule machine in the hope of that one rare drop. This is not, despite Suzuki-san's protestations, a game about the player's freedom, for you are not free, just as the world is not open. On the contrary, you are constrained throughout by the invitation to become Ryo, and if you accept that invitation you will find yourself in a beautiful closed world of surprisingly vivid characters who inhabit a place that transcends the usual limitation of videogame villages: the world is there for you to discover, and is not merely a pile of resources for you to loot and pillage. Ryo's adventure is not over, and even if there never will be another Shenmue game, it is inescapably appropriate that Shenmue III ends once again on 'tsuzuku' (to be continued) not 'owari' (the end). It is far more wonderful that the narrative tree that is Shenmue was given another chance to blossom.
Very interesting view on the "meta" aspect of playing a game--that the game constrains you to experience the same constraints that taking on Kung Fu in real life would involve, and thus allowing you to enter into Ryo's character.
I highly recommend that your readers also view the mentioned video "Shenmue 3 is a Terrible Game and I’ve Wasted My Life", which, in spite of its title, is a rather deep discussion on what games need to be from a narrative and experiential perspective. The narrator of that video also shares your desire to play Shenmue for intrinsic rather than extrinsic rewards.
Keep on playing slowly, we all benefit!
Thomas O'Dell
Game Design Lecturer
Singapore
Posted by: D | Tuesday, 28 June 2022 at 10:23
Thomas,
Many thanks for your thoughtful comment! It is gratefully appreciated. And also for correcting my misapprehension about that video, which I confess I judged solely by its clickbait title. I am not one for watching videos, but I knew this video by reputation and so included the link anyway. Clearly, I misjudged it!
All the best,
Chris.
Posted by: Chris | Wednesday, 06 July 2022 at 11:53