The Elder Scrolls: quantum mechanics and game stories.

The Elder Scrolls: quantum mechanics and game stories.

[Editor's note: expect spoilers for the Elder Scrolls series generally, as well as Metal Gear Solid and Knights of the Old Republic.]


In The Elder Scrolls II: Daggerfall, timelines are strange. The initial conditions of the game are complicated, but fairly typical, when it comes down to it: there are a few different kingdoms in the Iliac Bay, all vying for power, and a kind of WOMD - Numidium - powered by a magical artifact, the Mantella. Using the Numidium, each of the individual kings is hoping to win ascendancy over his rivals in the Bay; the Emperor is hoping to unify the Bay and restore Imperial rule over the area; the Orcs are hoping to establish their own independent state; and some supernatural beings are hoping to achieve immense personal power through the Mantella. The player-character, although ostensibly backed by the Empire's spy network, can of course choose to pursue any of these goals. He or she will come into possession of the Mantella, and have to decide who deserves what end. Great! All those things we seem to love in games: multiple options, different consequences for our actions, and a good fight one way or the other.

Of course, the problem arises whenever you introduce multiple endings to a game - how do you negotiate the state of the world at the beginning of a sequel, if you want to produce one? The multiple outcomes of Daggerfall are quite robustly different: they involve the creation of a new kingdom, or of new gods, or the subjugation of people under an Imperial banner. The lives of millions are at stake. By the time TESIII:Morrowind rolls around, how do we expect people to discuss the Iliac Bay incident? The ES games have always relied heavily on in-lore writing: there's an acceptable degree of fallibility in characters' knowledge. People in Morrowind (the province, and the game) are sufficiently removed from those events, in time and in space, that they don't necessarily provide an accurate picture of what happened at the end of the last game. This kind of argument from ignorance is not an implausible way of explaining this phenomenon. Other games have used a similar system. In the original Metal Gear Solid, there are two possible endings. In one, Solid Snake's love interest, Meryl, is tortured to death, and Snake escapes with Otacon; the two remain fast friends in the rest of the series. In the other, Meryl survives, she escapes with Snake, and Otacon manages to extract himself of his own gumption. For a while, speculation remained as to whether Meryl had lived or died, and which ending was more appropriate to the series, and to Snake's character, but ultimately, Meryl showed up again in MGS4. The retcon was less a change of facts and more a change of interpretation; it was suggested that if the player had escaped with Otacon, Snake was simply mistaken - Meryl hadn't died, but was severely wounded, and was later found by government agents inspecting the site of the game's events.

A somewhat more nuanced (but perhaps less useful) approach was employed by the Knights of the Old Republic series, in which the player could (in true Star Wars fashion) either save the galaxy or set up shop as the new Sith Lord to destroy it. By the time the next game takes place, the main character was directly asked by an NPC if he or she remembered the events of the original. The player has clear dialogue options which will ultimately decide the truth of the matter: whether the player-character in the first game was male or female, and whether they sided with the Jedi or the Sith. In each successive playthrough, you could re-determine the issue, and explore the consequences of different actions.

To return to Daggerfall, however, what makes that game unique is that it does not take either approach. Instead of trying to negotiate five or six endings into one by virtue of information selection, or letting the player decide what happened on a meta-narrative plane, the solution is totally in-world, while simultaneously a tongue-in-cheek acknowledgement of the problem. When the magical Mantella is activated, it causes a Dragon Break, a magical event with the outcome that all the endings simultaneously and truly occur. The civil war ends in a relative stalemate, because every king wins and loses the war; the Orcs do create their own state, but are also defeated by the other kings; and the Empire re-establishes control over the area. There's a weird in-lore technical discussion of it over at the Imperial Library website which suggests what happens - namely, that when the timeline is skewed, the Dragon Break retroactively changes the past to fall in line with the new present conditions. Yes, it's confusing, and quite a few characters and books in the game devote time to speculating on what exactly happened, and how.

In a strange way, however, the game may have been looking forward to more recent developments in science - namely, quantum mechanics. There's a quality of particles known as quantum superposition which suggests how a particle can exist partially in multiple states of potentiality, due to the way probabilities overlap with each other. The double-slit experiment is the most common example of this process taking place. Scientists basically set up a reactive wall behind a shield with a single opening, and fired one electron through it at a time. So far so good. However, when they added a second opening to the shield, instead of seeing the electrons go through one opening or the other, they appeared to go through both openings simultaneously. More specifically, rather than going through both openings as particles, they seemed to act like waves, which (to put it bluntly) are quite different from particles. If you try to observe this happening, however, the waves coalesce back into particles. What's actually happening is relatively simple wave interference, like ripples on water where two stones have dropped, and the crest and troughs represents different possible outcomes. And this isn't only true of single particles like electrons - buckyballs do it too, and they're composed of sixty carbon atoms linked together in a soccer-ball-like crystal. If you're familiar with Schrödinger's Cat, that thought experiment is basically a way of asking the question, "What happens if we try to apply this quantum mechanical property to classical, normal-sized objects?" The results are weird. Dragon Break weird. In a sense, Daggerfall ends in a totally legitimate way - a huge scale-up state of superposition, in which every possible ending of the game does take place, no matter which one you personally observe as you finish playing the game.

On the meta-narrative level, this also happens in KOTOR. At the beginning of KOTOR2, your character from the last game might be any of four archetypes - male/female and good/evil - and they probabilistically overlap. The possibilities only coalesce into a real person when you decide, at the beginning of KOTOR2, that Revan was a man who fell to the Dark Side and regained his throne as a Sith Lord. But there's a meta-narrative conceit that takes place: we understand that although those three alternatives exist in the dialogue box, they exist because the game designer is allowing us to take command of the storyline in this particularly instance. He's winking at us and saying, "Go ahead, pick whatever option you went with from KOTOR." In TES, this isn't the case. Even the eponymous Elder Scrolls themselves are blind to the events of the Dragon Break. And as in the double-slit experiment, without an observer, the wave form will never coalesce. All possible states are superposed.

Can we find a way to make this happen in games? A way to model worlds in which unobserved events are superposed until an external factor forces them to coalesce into a single state? There's a problematic way we currently deal with side-quests: they're static, or they're static until we create conditions which make them impossible. (In The Witcher 2, for instance, side-quests are open for a whole episode, until you make one step past the point of no return and the game will suddenly alert you that you've failed five quests all at once!) What if, instead, the smaller conflicts and cycles of the game world waiting for you to arrive, they actually developed upon your arrival? Instead of finding the same set of side-quests waiting for you each time, you might that one side in a conflict had gained an advantage over the other - that the problem had been solved without you, or had become a different one entirely - or you might find that nothing was the matter in the first place. Maybe on its own, it doesn't seem like much, but the mentality of possibility could go a long way towards the next step in building rich game worlds which actually seem alive.

Story where there is none.

Story where there is none: how Demon's Souls and Dark Souls tell perfect video game stories.

 [Editor's note: Expect spoilers, such as they are.]

Gamasutra recently featured a great little article by Robert Boyd about the surprising relationship between the perceived and actual difficulty of Dark Souls - a game which flaunts its difficulty, but uses that very fact to draw in (and keep!) a large audience. It's a neat piece which succinctly considers the way the game generally depends on the player to determine difficulty. Boyd notes, for instance, that the flexibility of the leveling system gives the player freedom to play any way they like, to tackle any challenge using any strategy they conceive, and the option is always available to summon another player to help.

Overall, it reminded me that I did a lot of thinking about the storytelling in Dark Souls, and its predecessor Demon's Souls, when I wrote my undergraduate thesis. So, as a companion to Boyd's meditations on how gameplay is fruitfully directed at the player's experience, it might interest some to consider one specific design decision which makes the story as flexible and rewarding as the gameplay itself: the banishment of exposition to item descriptions! It may seem absurd, but that one decision may be the only thing that makes the story in the Souls series successful at all.

Of course, some people will express some degree of incredulity at the notion of the series having "successful" stories, since more than a few players generally came away with the sense that there was no story being told at all. This is somewhat true, but to be more accurate, it's merely a recognition of how flexible the games are with their storytelling. Players can't really be faulted for thinking there's no story, when the medieval fantasy RPG as a genre often trains people to look for very specific narrative markers. When people think of MFRPGs, they think of large-scale games like the Final Fantasy series (which may be a dated example, but remains an example I love for its cultural weight); games like that often sink a lot of narrative material into cutscenes, very direct instances of narrative intrusion into the gameplay experience. There are other ways to achieve similar effects - scripted sequences, for example, which take place within gameplay, but effect situations which cannot be interrupted by the player - but the DSs totally avoid this structure. Other games, meanwhile, have taught us to look in different places for story clues. The Elder Scrolls games have taught us to mind our bookshelves, as the multitudinous in-game books are informative, fun, and can boost our stats. It has long been the case in RPGs that non-player characters populating the world can provide useful information, with the extra tint of unreliable character biases. But the DSs go in a different direction entirely.

What do you accomplish, as a designer, by doing a whole bunch of work creating a vast, rich, multi-faceted world, only to reduce much of it to verbal descriptions of one or two sentences, buried in an inventory screen? Two things.

First, you put control in the player's hands. I think that's important for nearly any aspect of a game, and I think a fair number of designers would agree. A game is meant to be played, and the more pieces you give the player to play with, the more rewarding the game. It highlights and explains, moreover, the diverse responses to the story in each game. Many players would say there was little or no storyline at all; others would say there was a fairly complex one. Both are fairly accurate descriptions. I'd hazard a guess and say that the different impressions of the situation are highly motivated by the biases of the players themselves - the narrative system creates a kind of index of players. The players who are more interested in the gameplay system will largely ignore the story, because they're not profoundly interested in it in the first place; those players who are explicitly interested in the story will find it when they look for it. The story is optional, to rather bluntly oversimplify the point. Ultimately, the decision to nestle narrative information in a secondary screen strikes an effective balance: it's readily available, but never gets in the way of playing the game. Unlike lengthy cutscenes, which deliberately monopolize a player's time, this system only gives the player precisely as much information as they want at any one time.

However, we'll also find that secondly, and more importantly, this system somewhat ingeniously finds a way to bridge the gap between these two stereotyped groups of players (we'll say play-types and story-types, broadly). The specific linguistic form of the "item description" is one particular to video games, and as such, it's wise and effective to make use of it when telling a video game story. It only exists in games, and more specifically, only exists in games which place an emphasis on having a variety of interesting items to collect, match, improve, and utilize in combat. In most RPGs, and certainly in Demon's Souls and Dark Souls, much of the essential activity of the game is in exploring the world, collecting items, and using those items to do more exploring and fighting. By placing narrative exposition in this tiny niche, the designers intimately tied together these two impulses - collecting items, and discovering more of the story. A player whose major effort is in creating a sweet build for a character will invariably pick up some juicy storyline details; a player who wants to learn more of the story will do so by picking up some new pieces of armor or unique superweapons. Neither detracts from the other. Each bolsters up the other. The result is a game experience whose balance is determined by the player, and which will ultimately be rewarding no matter how that balance is achieved (heavy on story, heavy on action).

Understanding the folk process in Irish music.

The man, the music, and the box: why going back to the source is important.

[Editor's note: I don't own John Kelly's music, but I've given you a chance to have a listen to it. Also, the concertina I've used for an example photo is one of Jeff Thomas'. He's awesome, makes wonderful concertinas, and his website is http://www.thomasconcertinas.com.]

Both of my previous entries were about games, but I did let on that this blog would be about Irish music as well. As such, I’d like to take a similar tack to those entries, and use a specific example to draw out a realization which has practical benefits. It probably helps to know a bit about concertinas, and music theory, to totally grasp what’s going on here, but it might be fun and interesting if you’re not in the know in either niche.

A while ago, my friend Armand and I were playing a very West Clare West Clare reel: the Bunch of Green Rushes. It’s a spectacular three-part tune in a D mode, with Fs and Cs very fluid in their tuning. We launched into it…to find that in the very first part, he and I quite spectacularly played the F totally differently: he played it sharp, I played it flat. The result sounded like getting stabbed in the eye with a cheese grater. He told me it was sharp, and I just shook my head, because I’m not talented enough to play the fiddle and speak at the same time. When we had finished the set, he told me he always heard the tune played with the F-sharp in the first part. I insisted that John Kelly played it flat. John was a noted fiddler and concertina player, so Armand suggested he might have played that note sharp on the concertina.

This is where I tried to pull my concertina-player knowledge on him; it might require a little bit of backstory if you’ve never held a concertina. See, you’ll take a look at your standard concertina in Irish music and see three rows: the inside row, the middle row, and the outside (or accidental) row. The most common layout, by far, is a G major scale on the inside, a C major scale (one fifth below) in the middle, and relative accidentals (notes which aren’t in either of those scales) on the outside. Although, for a music that’s primarily in D major, that may seem silly, the layout actually provides an incredibly useful and flexible system for playing in all the major keys of Irish music, if you’re comfortable going between all three rows of the concertina to reach all the buttons you need. The thing is, older players often tended not to cross rows too much – they certainly did it, but usually only when necessary. Furthermore, many older players learned, or continued to play, on two-row German concertinas, which didn’t have any accidental row at all! If you play the Bunch of Green Rushes on the concertina, the first part very nicely fits into the C row, with a big juicy F-natural right under your middle finger (and the F-sharp cast off to the G row, under the relatively uncomfortably left-hand little finger). So I figured, if anything, John probably played that note flat on the concertina, since it’s a lot easier to reach.

Here’s where things got weird. We went back and listened to the track from John Kelly’s solo album, which includes him playing the Bunch of Green Rushes on both fiddle and concertina. He plays it on the fiddle with glorious F-natural notes, and big meaty C-natural as well. But on the concertina…he plays the note sharp! And in fact, he plays both the F and the C sharp – which made very little sense to me, since the C-sharp note is on the third, accidental, row on the concertina, and John’s style very rarely reached into that row. Now I was thoroughly confused – especially since, in the second part, the C was fluid, flat the first time and sharp the second. This was totally at odds with the accepted understanding of how John played the concertina; while he would normally be expected to play more or less in a single row at a time, this sounded like he was switching rows quite a bit, and in some rather unnatural shapes. What was going on?

Have a listen yourself.

When I finally checked the pitch, and thought about it a bit, I figured out how he was playing the tune, and why. He was playing the tune in G, rather than D, and this had totally changed the availability of notes in the tune, and how he could ornament them. G is a remarkably flexible key on the C/G concertina; G major and C major overlap considerably in notes, and the difference – F – means that you can play with the tuning of the seventh in the scale, which happens quite often in Irish music (in the key of D major, that would mean the change of C-natural and C-sharp). That means you have the whole G major scale, in the inside row, and can quickly go to the C row just for F-natural; John had chosen this layout for the tune, simpler and more flexible than playing it in D. However, this also made different options implausible. Whereas playing the tune in D makes F-natural accessible in the first part, playing it in G means the corresponding note would be B-flat – neither in C major or G major, banished to the outside row of the concertina which John rarely used.

So what’s the big deal? Nothing really. But if you play Irish music, you know that there are countless different settings of tunes, many of them with regional connections, or linked to particular musicians. Usually we chalk up these differences to the “folk process” in general – it’s true that this music was primarily orally-transmitted, and therefore allowed for a lot of variation based on mishearing and misremembering. Differences in regional and personal styles also prompted people to outright change tunes to suit themselves better, or to suit the needs of dancers. But here’s a very particular instance of how a setting of a tune changed due to the necessities of a particular instrument: the Anglo-German concertina, immensely popular in Ireland during the latter half of the nineteenth century. Especially because John Kelly plays the tune totally differently on his two instruments, it’s also a neat example of how musicians negotiated these sorts of problems in the past – rather than contorting his setting on one or the other to fit both instruments, he did what sounded the most awesome on either! On the fiddle, he slides dramatically into big F-natural notes; on the concertina, he hits big meaty chords on the B-natural, and plays in octaves during the second part of the tune. Especially for concertina players these days, who have heard a generation or two of concertina players whose ornamentation, phrasing, and style was very much based on fiddle and pipes playing, it’s refreshing to hear, and understand, how someone really plays the concertina with an ear for the specificities of the instrument.

To John Kelly!

The first two hours, part 2; or how I learned to stop worrying and just not play games with too many features.

Even small worlds should start big

[Expect some degree of spoilers for Lost Odyssey, The Witcher 2, and Skyrim.]

So, as per the last post here, we are playing the first two hours of Lost Odyssey and the first two hours of The Witcher 2: Assassins of Kings; and then here we are, wondering why we’ve only ever played the first two hours of the former, while playing multiple times through the latter. Both games are epic fantasy RPGs, but the player experience is designed from two nearly antithetical standpoints. I started off by taking a look at the way the world and story are introduced in each case, since epic fantasy RPGs generally make the exposition of a grandiose narrative central to the game; this time, however, I’d like to get a feel for how the introduction of mechanics in a complex system affects the player.

The thing is, big worlds tend to have big, complex systems that players use to interact with them. If that’s to be the case, then the challenge of the first part of the game is essentially doubled: the designer has to hook us not only the idea of exploring the world, but on the way we’ll actually be undertaking that task. There are some games with great stories that never get played, because the system and controls are horrible; there are others with shit stories which people still play, because the system is effective and gameplay is fun. (There are notably more games of the second type than the first, because most games persist in having generally shit stories.) Or, on the other hand, there are plenty of mechanically fun games that people will put down because they’re just boring.

This issue is central to both Lost Odyssey and The Witcher 2. LO takes place in a huge world mishmashed from magical and mechanical elements, with apparent social and political structures in place maintaining and overseeing that complicated system; meanwhile, it has a conventional turn-based RPG interface system, attacking, defending, using special abilities, and using items. TW2 is a more conventional world, drawn very deliberately from late medieval and early modern source material in eastern Europe; to balance that out, though, we have a fairly unique action RPG interface, which demands some complex balancing of potions, items, spells, and swordplay, both in and out of combat. In a certain sense, TW2 has a more complicated mechanical system than LO – at least, it has one with which most gamers will be less familiar (if only for the single craziest reason that potions cannot be used during combat). And yet, because of some dubious design decisions early on, my roommate and I found TW2’s system easier to navigate, even though both of us are used to LO’s Final Fantasy-style mechanics.

Two points demand consideration: one, the streamlining of tutorials, and two, the availability of actions early on.

Both games have player tutorials, although it’s notable that TW2’s tutorial was retroactively added only after players had complained. We were playing the Xbox port, the Enhanced Edition, which included this little extra bit to ease us into the game. Without passing judgment on the issue of hardcore gamers, hardcore games, and how balls-crushingly difficult a game should be right from the get-go, I do want to give CD Projekt Red some credit for making one design decision (no tutorial, dropping a player into combat early on) due to some principles and beliefs about their players, and then amending that decision when given the chance (added a tutorial as DLC, or including it automatically in the Enhanced Edition available on the Xbox). That’s ballsy in and of itself: getting feedback from players, especially about a touchy subject like difficulty of what’s billed as a kind of “game for gamers,” and then incorporating that feedback, goes a long way towards making a good game experience better. So points for you! But we should also note that they didn’t give ground willy-nilly, like Rosecrans at the battle of Chickamauga; they added a tutorial, and it’s an optional one, which takes the form of a discrete episode chronologically separate from the main quest.

It seems stupid to make such a big deal of it – the game asks, would you like to play the tutorial? Yes or no – but since this post is about observing the differences between two games who take different approaches to these things, it really struck me how vast of an experiential difference arose from the option to play the tutorial, or to skip it, and to have the tutorial as a single wholesale unit. My basic take-away from the fact: tutorials should be discrete episodes, and they should be optional.

Who likes tutorials? It’s very rare that a player will remark, “Oh yeah, man, that tutorial rocked my socks.” More often than not, it’s, “Man, that tutorial was just annoying and awkward and still confusing and fuck that tutorial.” Does TW2 avoid that? Does it make a tutorial that’s enjoyable and informative?

Kind of, not really. It does achieve something pretty upfront and basic which seems necessary insofar as having an attitude about tutorials, as a designer. Let’s all agree on something. Tutorials break the fourth wall. Period. They have to contend with the problem of interface, not only in navigating menus but in many cases literally teaching the player which buttons to push, where, when, and how. That’s never going to be a unique, beautiful, streamlined system. It’s always going to be kind of shitty. Since that’s the case, why don’t we more often go for broke the way that TW2 does? During the game, sure, I want to be really into the game experience, but right at the get-go, fine. Ask me outright if I want a tutorial, and make that tutorial as blatant as possible. TW2 clothes its tutorial in the pleasant skirt of a totally meaningless story: your boat has sunk in a little marsh, and you need to walk to the arena to compete for some prize money. No biggie. It doesn’t have any pretensions about being a training sequence, as though Geralt and I are learning the same skills at the same time; it doesn’t have any pretensions about being well-written or incredibly fun. But amazingly enough, I damn well appreciate that. What the tutorial from TW2 tells me is, quite literally, Geralt knows what the fuck is going on, and the game is going to be more fun if you don’t make him look like a floundering douchecanoe. That’s what I love about it: it has a clear goal, gets to that goal as efficiently as possible, and then takes its hands off your shoulders to let you ride (and maybe tip over) on your own.

The tutorial works by giving you a series of small tasks and text prompts in rapid succession. As a general rule, I’d say avoiding text is preferable, but the decision to give you written on-screen directions and reminders is actually fine here, because again, the tutorial is set apart from the game experience itself. If, eight hours into a game, I had to read a bunch of written directions overlaying my screen, I’d feel pretty unhappy about it, because I was really getting into the world and enjoying it. Even relatively early into a game – if I start it, get the basic controls down, and then immediately find myself deluged with new directions – I’ll be quite resistant to text direction. But if your tutorial is totally up-front with itself, and tells you, “Hey; this isn’t the game, this is just a way for us to teach you how to play the game,” then I’m totally fine with it breaking a few “game design principles” for the sake of efficiency. By leading in with the option of the tutorial, the game weeds out people who don’t want a tutorial, and as a result, the people who do want one get a pretty good one. It seems so simple, but I don’t know that it ever occurred to me so directly before this. If your players have chosen to play the tutorial, rather than having it foisted upon them will or nil, then they won’t complain about it if it’s not a majestic wonderland of edutainment.

It’s not often that intensive, story-driven fantasy RPGs can take cues from fighting games, but I will say this much: the best tutorial I have ever played was from the Xbox Arcade game Skullgirls, a cartoony fighting game with a lot of style. Although fighting games as a genre do tend to have tutorial modes, the tutorial system for Skullgirls involves an intensive, multi-chapter mode for learning everything from basic movement skills to advanced combat techniques. It’s so totally set apart from the game that it feels like a mini-game, or series of micro-games, unto itself. It’s so blatant about its tutoriality that you don’t mind it. And amazingly, it very effectively teaches you the astoundingly complex combat system of Skullgirls. The experience served as a devastating litmus test for what I once thought was appropriate for games. If people can learn to be good at Skullgirls, then whatever infinite crystalline snowflake of a game you’ve created can definitely be taught just as well. It may not be seem great, but giving your players a whole bunch of information all at the once could be the best way for them to actually retain that information.

Maybe it has something to do with the brain goes into crisis mode, but TW2 basically just told me, “Get out of your boat, collect some ingredients, and make that dude a potion,” and I DID. Granted, I figured out more of the system as I played through the game, but the basics of every part of the game – potions, combat, menus, gear – I had right from the beginning.

Lost Odyssey, on the other hand, takes the opposite route – one which is much more frequently employed, in my experience. It was clear, from the scope of the menus and buttons, that the game had a lot to offer with its fighting system. At the beginning though, we could only attack or defend. We were dropped into a battle, we hacked our way through, and then, when we found a ring, we discovered – calloo, callay! – we could equip rings for special abilities, the first of these being a timed-attack reflex system. The menu told us so by interrupting our game to point out to us precisely how to equip a piece of loot, select its abilities, and employ those abilities in combat. It was, frankly, annoying, and worse was seeing the other menu buttons which were grayed out, unselectable, or had no entries yet, because I knew I had more moo-interrupting-cow tutorial moments ahead each time the game decided I needed to know something new.

I told you in my last post that I had trouble getting into LO’s whole world, characters, etc. What doesn’t help is the game stopping that experience constantly, to tell me it’s a game, before leaving me to go back to playing. There’s a bit of a representational problem in this design. You’re breaking the fourth wall a lot in most turn-based combat RPGs, since seeing time bars and menus of actions during combat is obviously not “realistic;” but that system still adequately suggests how a person might actually fight, by taking time to prepare oneself, to wait briefly, to choose one’s next move carefully but quickly, and then to act. There is no representational analogue to the magical omniscient narrator fairy who pops in to tell you how to put a ring on your finger and know how to do a thing. If it had better results, I’d let it slide, but the fact really bugs me that an inefficient, and un-fun, and ineffective system persists in being employed by plenty of designers out there.

This line of discussion is leading steeply into my second realization: tutorials are streamlined most efficiently by giving the player all (or most) of their relevant abilities in the beginning of the game.

“But wait!” we might cry. “This flies in the face of every experience most of us have ever had! How many of us have turned off a game because we were overwhelmed by information and ability, and how many because we found the game was too slow in giving us more information and ability?” And in a certain sense, that’s true. You won’t necessarily lose gamers on account of making a system that’s very deliberately simple, and only gives them one new piece of data at a time; they may find it tedious, but they will rarely find it unusable. However, that’s hardly an argument for such a system being the best, or even particularly useful. Especially in the context of a story-driven RPG, a tutorial should contribute to your experience of gameplay and story at the same time. When it comes down to it, we’ve had too many games that somehow convince us that a veteran of military service still hasn’t figured out how to parry and counterattack until just now, or that a graduate of a magical academy knows only a few rudimentary spells, just for the sake of forcing upward progression through the game.

It is, admittedly, important for some sort of upward progression to exist; it’s a huge part of why gamers play games. Particularly in the case of MMORPGs, where you want people to play for a long time and keep paying subscriptions, ideally, the deep psychological satisfaction of leveling up or acquiring a new ability can be a deadly weapon. But in certain types of games, and especially in action RPGs, you have fundamental design concerns which conflict with this model: in some games, the acquisition of actual skill as a player, and in some, the vaunted importance of the player’s ability to choose and develop a character and playstyle. As was the point of this comparative project, this is an observation I would only make in directly comparing two games, but I found it much easier to deal with TW2 for the simple reason of knowing what I could do. At the beginning of the LO, you can do nothing but attack and defend. You’re some kind of insane warrior with the ability to leap dozens of feet into the air, destroy huge electro-magical tanks, and mow down trooper after trooper wearing full plate armor, but your martial skill in gameplay is limited to swinging your sword or blocking with it. You can’t do anything else until you’ve learned how to do it, at some arbitrary point defined by the game (oh, immediately after this battle ends, you find a magical combat ring? HOW INTERESTING).

The problem? When you only give your player limited abilities, what you’re literally doing is limiting their playstyle, and forcing them to play the game in the way you want them to play it. You’re reducing player choice, and shunting them down a small number of paths. Players naturally rebel against this impulse. Big arrow point to the right? They want to go left. Character supposed to die? They’ll try to sequence-break. Or, if you give them no options, they’ll a) begrudgingly follow along or b) turn off the game because I’m not a part of your system, man.

What I happily discovered about TW2 was that, even though there was clearly a level-up and advancement system, all the basic assets of that system were already in place. You already know all the spells that Geralt will know right at the beginning, and they all already do what they’re supposed to do; they simply get stronger and more effective over time, as would naturally occur with regular use. This had a particularly interesting effect: it meant that my skill as a player and Geralt’s skills as a character actually advanced along similar arcs. From the start of the game I understood that I could use one spell -  which, for lack of recollection, I will call the Force push – to stun an enemy and then one-shot them. My combat style gravitated towards this system, using a spell to halt enemies and then my sword to quickly dispatch them. As I went through the game, my Force push got stronger, to the point that I could stun four or five enemies at once and one-shot all of them before they recovered (which was awesome). That’s only a smallish example. The availability of playstyle options from the beginning also means that I’d have had no problem changing gameplay paradigms if that system stopped working. Again, it might not be something we think about too much, until it becomes evident from a game that does it right, but frankly, it’s much better to have a playstyle that develops in reaction to the world, and to the player’s abilities, than one that develops in reaction to arbitrary new information and new abilities. Or maybe it’s not, but damn it, it certainly seems like it is to me. Again, we can appreciate the value of unlocking something new, and how it positively affects the player, without building a kind of deus ex machina ability system that drops new stuff for us on a semi-regular basis.

In fact, one of the best examples of balance here, in my mind, is the discovery of shouts in The Elder Scrolls V: Skyrim. Some people went through whole chunks of the game using only regular combat skills, stealth and magic to get through their missions, deliberately putting off the main quest. Then they discovered that holy crap Dragonborn I can shoot fire from my mouth, and all of a sudden the balance of the game changes pretty meaningfully (no-mana-cost spells, just a recharge time), as does the player’s combat style in relation to that change. It’s intense, it’s awesome, it’s tied into the story quite intimately – and it happens once. Not a million tiny times, without any storyline justification (you found a magic ring that you can use in combat! Now I’ll teach you how to wear rings because you apparently understand how to weave the fabric of the universe, but not fingers). Once. That’s it; one big change to the game mechanics, everything else you have to get up front. At the very least, we can appreciate that any further big changes need to be…er, small, for one thing, and for another, storyline-derived. More or less any Final Fantasy game gets the picture. Because party-based RPGs introduce new abilities by introducing characters, you always have a certain leeway in giving people more stuff. The problem arises when your cast of characters is waaaay too big, for both story and gameplay purposes, and it just feels like being nitpicked by meaningless people with useless abilities. In a way, this provides a good size test for a project. Could you introduce most, if not all, of your characters and abilities quite near the beginning, and still reasonably expect the player to succeed with them? If not, maybe cut some shit down.

The first two hours; part 1.

A few weeks ago I had the pleasure of sitting in with a friend as he began two RPGs in one night – one eastern, one western. For whatever reason, my roommate decided that it was high time he began playing Lost Odyssey, which he’d owned for some months without ever opening, just at the instant I’d showed up with The Witcher 2: Assassins of Kings in hand. We cranked up the AC, poured out some good beer, and gave each game around two hours of intro playtime to get a decent feel for each. Both games have gotten good reviews – both games were noted for having great stories, in particular.

So why have we played through The Witcher four or five times, without touching Lost Odyssey since?

There are plenty of discussions about the differences between western RPGs and JRPGs, but I really wanted to sit down and think about why such a disparity in play experience took us to such dramatically different ends. My friend was playing, and I was watching, but I still found it difficult to navigate the intro to Lost Odyssey, whereas The Witcher felt pretty solid, right from the get-go. For that matter, it might be noteworthy that plenty of gamers have found The Witcher to be turgid and over-difficult (Zero Punctuation’s review was fairly accurate in its assessments, even though I disagree with the conclusion). So why did my friend and I have that negative response to Lost Odyssey, but not The Witcher? Let's start with that question.

And maybe, ideally, we could get some nifty game design worthwhiles out of the experience.

[Editor's note: I can certainly appreciate and anticipate the comment that these games are totally different, trying to accomplish different things, etc., and that same comment is true of the varied other games I reference in the article. The main decision to compare Lost Odyssey with The Witcher 2 was the literal fact of having played both of them in the same night, and the rest of the examples simply come from my experience. That experience is limited, so example and counter-examples are expected and welcome.]

Expect spoilers for the first hour of Lost Odyssey, The Witcher 2, Final Fantasy 7, Kingdoms of Reckoning: Amalur, and Knights of the Old Republic.

Big worlds start small

Evidently, both LO and TW2 take place in vast, wonderful, lovingly-crafted worlds full of heroes, heroines, assholes, and common folk, all wending their way through a complicated superstructure of magic, mundanity, mechanics, politics, and general mystical hoo-ha. There’s a lot to deal with in both instance: tons of characters, all with particularities to their relationships, belonging to different organizations which themselves have particularities to their relationships, and a world with some basic rules and attitudes that are totally different from our own. In any medium, that’s a serious issue to overcome, and in a game, even moreso, since most people come to a game expecting to experience “fun” more or less consistently.

Because I’m an idiot, I thought that a comparison to a different medium might actually shed some light on this question. For those of who you know that The Lord of the Rings is a book, you might also recall from having read it that it is fucking gargantuan. There really is an entire world crammed between the two covers of the red leatherbound edition which I expect any sci-fi/fantasy fan to own, padded deliciously with a whole slew of appendices on kings’ dynasties and how each chieftain of the Rangers died (wolves, evidently). But do you also remember how LOTR begins? It’s really tiny: a domestic drama. Two birthday parties, and one guy complaining about how his in-laws are trying to get his house. Everything that happens subsequently, from meeting elves to fighting off millennia-old servants of an exiled demigod to running barefoot up the side of a mountain, begins with the story of two tiny dudes eating and drinking with their friends and family. There’s a number of reasons why this works, one of them being that it allows for appropriate expansions of scale through the book (you’ve got more room to up the ante with each successive challenge), but more important to the question of how a new visitor experiences a huge world, it’s small enough to digest altogether. First off, you only have to deal with three really crucial characters at once – Bilbo, Frodo, and Gandalf when he shows up – and second, their relationships and troubles are things we already understand. Instead of giving you a whole huge chunk of world at once, we get a single sliver.

How do our games hold up to the same principle?

LO begins with a big epic battle, chock full of all the awesome things that should push all our awesome buttons. Dudes in big clanky fantasy armor are tossing around huge swords while towering robot tanks crush guys, while reinforcements are begin driven up in what look conspicuously like steampunk coffins. The player-character shows up and starts kicking ass and taking names, except all the enemies are faceless evil robot-men with no identities, so he basically just kicks ass until the sky turns to lava and everyone dies. And yet, interestingly enough, between the two vast armies and their tanks and artillery pieces, the numbers of actants in the opening battle sequence could not begin to approximate the number of fucks that I could not give about what was happening.

I understood the basic cues I was given – guys in black armor with red eyes, guys in gold armor with…faces, one dude who’s a third the body weight of everyone else and has his midriff exposed like a teenage girl but still ruins everyone else – but they were nothing other than internally-consistent cues which presented the basic shape of a totally generic epic sci-fi/fantasy world. It feels especially ruinous saying something like that, since LO received some stellar praise from reviewers regarding the uniqueness of its world and the emotional beauty of its story and characters, but frankly, I want to talk about the effects of design decisions, and the effect of the early design decisions in LO was that I never saw any of that beauty in the first two hours of the damn game. LOTR begins with a sequence that invites you into the home of a main character, and rather than confusing you, only makes you care about them on purely human terms: they have family relationships; they celebrate birthdays; Bilbo is happy to see an old friend again. These are totally mundane facts, but they give you the most rudimentary basis for caring about the larger events which will later happen. LO doesn’t give you that at all. Instead, it plays precisely into the human problem of large-scale mechanized warfare, a spiritual problem which we have understood for centuries – if you see row after row of identical-looking, uniform-clad soldiers, it’s harder to think of them as unique, living people and easier to think of them as bullshit sword-fodder. So even though we understand the basic visual cues of good guys and bad guys, we have no real basis for assigning good or bad to either side, and they’re really interchangeable for all intents and purposes. The addition of the player-character to that equation did nothing to change that balance, because I still had no understanding of who he was or why he did what he did. The bigger the introduction got, the smaller my interest became.

What’s particularly intriguing about this fact is that TW2 also begins with a huge epic battle, but it treats the subject in a totally different way. (Firstly, I’d like to point out that I played the Xbox version of the game, which includes an optional tutorial to be discussed a bit further on.) The intro to TW2 definitely isn’t perfect – it starts with text, for the love of God – but it very quickly strikes our key points by taking the reverse approach to the same big-battle set-piece. Instead of introducing us to very broad-strokes groups, good guy army and bad guy army, without giving us any real human referents for their attitudes and actions, TW2 gives us a handful of primary and secondary characters clustered together. We begin with our player-character, Geralt, in prison, with two rowdy guards keeping watch. We’re subsequently introduced to Vernon Roche, our captor and interrogator, who eases us into remembering the big battle at LaVallette Castle. That’s it – two major characters (Geralt and Roche), and two minor characters (the guards) who teach us more about the major characters by gossiping or remarking on them. It’s tight, informative, and not overwhelming. Even though we don’t understand what Geralt’s doing in prison, or who Roche is, there’s just enough mystery to not let us know, but still let us care.

We then get to our big battle scene, in flashback, but the treatment is totally different than in LO. We begin with characters, and only gradually work our way into actual combat. It’s certainly difficult to keep track of everyone, but it’s made much easier by two simple decisions which the developers wisely made. One, each character (or character cluster, like the mercenaries you meet near your tent) has a few unique relational details which let us situate them. We learn in the opening conversation that Geralt and Triss are lovers, and servants of Foltest. Geralt saved Foltest from an assassin. Geralt is a monster-slayer by profession. Foltest is about to fight a battle to recapture his children from their mother. That’s all the salient information, and we get it all literally in one conversation, which primes us for the whole prologue. Two, each important point is reiterated multiple times before we jump into combat. We’ll hear from Geralt, Triss, and Foltest that Geralt is a monster slayer, that he saved Foltest from an assassin, that he’s carrying on with Triss, and that this battle is about Foltest sleeping with someone else’s wife. It strikes a useful intermediary position between the mundanity of LOTR (in-laws, birthday parties) and the totally alien experience of LO (huge battle with robots and tall hats). Someone saved someone’s life; someone’s sleeping around; possession of children is in question. Those are the same basic situations we see in our own world, and they let us stabilize ourselves in a way LO doesn’t. Like I said, we can understand basic visual cues from LO’s opening (gold armor, good guys, black armor, bad guys), but we can’t really understand why anyone is good or bad, much less our player character. We can immediately begin digesting and responding to the world of TW2 because we have enough material to form opinions about it. Foltest is a king who slept with a nobleman’s wife and now has to deal with the consequences; maybe we think he’s an okay guy or not, but at least we can begin deciding that for ourselves given the basic facts. We know Geralt doesn’t like the situation, but that teaches us about Geralt as well.

We might also rename this principle, or at least frame it differently, by saying that LO tries to build a world from the beginning, but TW2 builds characters instead. It’s a huge difference that gives us a reason (or doesn’t) to actually keep playing and see more of that wonderful world. Maybe it’s not a piece of advice particularly useful to just anyone making any game, but for any narrative game – especially epic sci-fi/fantasy RPGs, for which the narrative is a focal point of the game experience and a major draw for your players – it can be a crucial stumbling block when handled poorly.

Think about Final Fantasy 7. It had a large cast of well-drawn characters, each of them with a rich backstory that could be explored ad infinitum later into the game; however, it began mostly with Cloud and Barrett running around fighting Shinra soldiers together, to destroy a mako reactor. There are a few conversations between them about the reactors, Shinra, and mako energy. Barrett makes it clear he doesn’t trust Cloud and Cloud seems blasé about the state of the world, in it only for pay. Sure, they come off as archetypal at first, but at least that serves as the basis of a relationship between the two characters, the world, and us as players. If the developers had instead dropped Cloud into the whole massive world with a looming extraterrestrial threat on the verge of ruining the planet, we would have been lost. Sephiroth and Aeris, two of the more memorable (and ultimately, more important) characters, are quite alien, and don’t make for relatable introductions to this place. Instead, terrorism, corruption, and ecological damage offer us points of comparison to our own world, and the broad-strokes portraits we first get of our characters again allow us to start forming opinions about them right away, since their opinions about events and ideas allow us to form opinions about them as people.

By contrast, we might consider a more recent attempt at a fantasy (action) RPG, Kingdoms of Amalur: Reckoning. The game has a really juicy opening, well-conceived, well-written, and well-executed. Yet in spite of all that, it never seems to have the proper effect of drawing the player into what’s happening. We see in an opening cutscene that the evil Fae king, Gadflow, has waged war on humanity for his god, only to find that our character is a soldier killed in battle with his forces. The character wakes up in a pile of bodies beneath a gnomish laboratory, only to discover that Gadflow’s warriors are assailing the laboratory. We quickly find that this place is the Well of Souls, run by gnomish scientist Fomorous Hugues. All of that is well and good, but like LO, we don’t much care, unfortunately. We see a couple of people we don’t know, being killed by other people we don’t know. Cloud professed not to have any emotional investment in his mission to blow up Shinra’s mako reactor, but at least we knew enough of him to know about his attitudes and opinions, and how they contrasted with Barrett’s. The only real characters in the beginning of KOA:R are the player character, who just returned from death and has something in the way of nil for a personality, and Fomorous Hugues, who appears briefly and then jets off. There are no character interactions which ground us in the world – everything is strange, and nothing wonderful enough to draw us in on sheer essential value. The most compelling people in the prologue of the game are the two gnomes who banter with each other while wheeling your corpse down a hallway!

Even the intro sequence to the first Knights of the Old Republic game managed a decent balance here. Although the prologue was spent fighting off a Sith attack on your Republic ship, you’re immediately awoken by a fellow soldier who tells you to find Bastila Shan, the Jedi knight on board, who’s crucial to the Republic war effort; as soon as you open your bedroom door, Carth Onasi, captain of the ship, contacts you and tells you to head to the bridge to aid the soldiers there. Bam. The developers wasted no time in introducing you to both of the major characters who would form the center of your entire experience on the first world of the game. Even though you won’t see Bastila’s face for hours, you hear her name over and over from the second sentence of the game onward, and since much of the game’s story hinges on your relationship with her, that’s a smart writing decision. It’s especially noteworthy because the game comes from such an established IP, and the devs could easily have gotten away with a fairly generic Sith vs. Republic and Jedi battle. Instead, they gave you a Sith vs Republic and Jedi battle, but framed it around the three main characters – you, Bastila, and Carth.

I called this first part “big worlds start small,” and by that I meant not to introduce too much of a world in the very beginning of the game. I might tweak that notion by instead suggesting a more nuanced approach: do your best to balance introduction of ideas with the introduction of characters. Lost Odyssey introduces you to a whole war, two huge armies, strange technology, magic and mages, a natural or magical sky-rending cataclysm, combat moves, item mechanics, and navigation before you know a single person’s name. In The Witcher 2, you know Geralt, Roche, Triss, and Foltest literally before you can take a single step (although you do have a few dialogue options before that). Those are totally different ways to begin an epic story in an epic world, and one immediately turned me off, while the other kept me going on another three plays-through. Especially when a big demographic of players only has so long to decide whether or not to play a game - two hours to decide whether or not you like a game can be fairly luxurious for some of us - making a deep connection right away can make or break the sixty hours of RPGing that follows. Lost Odyssey may indeed have a wonderful story in a beautiful world filled with heartbreakingly real characters, but I'll never know, because it lost me, barely ninety minutes in.

Next: "small worlds start big" - at least when it comes to mechanics and tutorials.