Iteration in games: a mini-rant

 Posted by (Visited 17186 times)  Game talk
Jan 172006
 

I saw this comment tossed off in a thread on Evil Avatar responding to the Moore’s Wall post.

That is the problem with Raph Koster’s ideas. No other entertainment media is iterative. Before a movie starts shooting, they have the entire film written, storyboarded, cast, locations rented, sets made, etc. Books have a complete story outline and usually a sample chapter written before getting a publisher’s go ahead. None of this is left to chance. . . none of it is left to a simulation.

My first reaction is “he’s nuts. Of course other entertainment media are iterative.”

Nine out of ten Hollywood screenplays never get greenlit. They fail to make it out of the phase called “development hell.” Many of those scripts are “written on spec,” meaning they originated conceptually from the writer; many others are actually “packaged” by a producer or agent, often to a known concept or IP. Even after that draft is written, there are numerous further drafts; the frequency of bringing in more writers is so high that the Writers Guild of America has very specific guidelines about who gets credited on a movie — to the point where “and” versus “&” mean very different things in writing credits. And even then, the script is iterated dramatically as the director gets ahold of it, as actors change line readings, and so on. And the writer is just about never there to vet the changes; most scriptwriters don’t have on-set privileges for that.

Recently I read an interview with Gene Wolfe, author of The Book of the New Sun in which he stated that he often goes as high as five drafts of a story. It’s not uncommon for years to elapse between drafts.

In Hanging Out With the Dream King: Interviews with Neil Gaiman and His Collaborators we learn that until recently, scripts for Marvel Comics weren’t scripts — they were plot outlines. The cadence of panel by panel was left to the penciler. DC Comics, such as those that Gaiman wrote for his run on Sandman, did work to scripts, but the penciler can and often does adjust things, sometimes in negotation with the writer and sometimes not.

The process of doing a piece of pictorial art often involves doing multiple studies of the same scene, including pencil sketches, a “cartoon” (the source of the word) in the case of oils or frescoes, and often, multiple versions of the same painting, until you get it right.

In theater, a play is written, and then goes in to “workshopping,” which often consists of table readings, which are not real performances but just actors sitting around a table. The troupe will refine the play this way, often putting the playwright through many drafts, before moving on to staging the play in rehearsals. Early on, as the play is seen on a stage, the process of blocking (positioning actors on the stage), set design, and lighting may well dictate significant changes in the play. Once the play opens, it generally does so in smaller venues, for test audiences, and the play goes on tour before heading to the big markets and finally “opening.” Once it has opened and met with approval, of course, it might be performed many times by many troupes, with no control exerted whatsoever by the original artist behind the work.

The points here are several:

  • Iteration is a standard and highly fundamental part of of all creative crafts.
  • In collaborative media, it is quite common for collaborators to change things; in team-based creative endeavors, it’s quite common for the original creator to not even have significant say over the final product.
  • None of these are directly analogous to games, but the original quote is still wrong.

The reason why the analogy breaks down is because of the nature of the games medium. You can tell from the screenplay whether a movie has a potential of being good. In most creative fields, this is called a “shit filter.” You can tell whether something is irredeemable crap — you cannot, however, tell whether the final product will be good (since it can be screwed up in myriad ways from that point forward). But a screenplay is, itself, a creative endeavor in a medium. It’s legible as a final work of art. In fact, sometimes, it’s all we ever get.

Games do not have this. In all the other cases I listed, you can tell from that first draft whether or not there is real potential precisely because the initial document (be it a sketch or a script or a riff) is representative artwork in its own right.

Design documents are not representative artwork. They are specification documents. Design documents are analogous to creating a movie based on the director’s film commentary; they are comparable to finishing a painting bsed on a verbal description of what it’s like. Reading a design document and assessing its strengths and weaknesses is akin to judging a song seeing only its harmonic analysis.

Because of this, the typical way of judging a game is to build it and see. If it’s not fun, you tune it through iteration. Hopefully it’s clear that this latter stage is no different from what other media do. And in fact, in writing and in composing and other “basic media” (as opposed to what I’d call “composite media” that are multidisciplinary), that’s pretty much what you have to do anyway. You write the story, because the idea alone will not suffice to convey the quality of final execution.

Lately, the ways in which the industry has tried to avoid this problem are by following Mark Cerny’s Method or creating a “vertical slice.” In both cases, the idea is that you create a game, just one that only has a fraction of the content.

…before meeting with the developer, the publisher makes it known that it will only sign games with full demos. The developer has to have a ‘Vertical Slice’ (as coined by EA) to show to the publisher. (For Vertical Slice, think of a thick slice of black forest gateaux with cherries and thick cream oozing out the sides. One bite and you know how the rest of the gateaux will taste.)

The most widely heralded example of a Vertical Slice was the Medal of Honor Demo that showed a full D-Day landing. In that moment you could see, hear and understand the complete vision of the game.

– William Latham, in Develop Magazine, August 2004

This is still the creation of a game, has therefore has the concomitant problems.

Of course, the typical Vertical Slice costs the developer several hundred thousand dollars. The cost to the publisher is typically around zero dollars.

The Vertical Slice is directly analogous to the sample-chapters-and-outline from the book publishing business. Its objective is to minimize expenditure in the creation of a game. But let’s not kid ourselves that the rest of what remains to be done is the “game” — it’s usually actually the “dressing,” to use my term from A Theory of Fun. It’s content — statistical variation on systems — and it’s both the most expensive and the least risky portion of the the final interactive entertainment product.

In other words, all the risk lies in the initial creative step. Just as nine out of ten (or more!) scripts do not make it out of development hell, and just as most chapters-and-outline do not get out of the slush pile, we should expect that nine out of ten first takes on gameplay mechanics will fail. And the take in the design document doesn’t count because it’s not actually a game, it’s a description of one.

Development studios known for fun factor, such as Blizzard and Popcap (to pick two from very different parts of the biz) have publicly stated that they do indeed discard most of their prototypes:

“Our path of development is extremely prototype-heavy,” said Gwertzman. “We’ll make half a dozen prototypes, and pick just one of those to be a hit casual game. And once we develop that one, it’s a very iterative process. It’s a sandbox model. We try different things out, and find out what’s fun. Only when we find out that the core mechanic is fun do we worry about the art, content, and all the other little details.”

“We really obsess over the core game mechanics. In a game like Bejeweled, hardcore developers look at that and might think it’s kind of…it’s very easy to kind of dismiss it, but we literally spent weeks on just the right way for the gems to fall when you make a match. In a game like that, it’s little details like that. How does it feel? Getting those little details right is what we prioritize. So when we’re designing a new game, we’ll spend months and months prototyping core mechanics.”

In that same Evil Avatar thread, I got knocked repeatedly for being too interested in theory. Not to dive into defensiveness here, but I make a good case study in this particular instance.

Raph writes great stuff, and his theory is always fascinating. Which is why it’s a shame he’s never actually created a truly compeling game.

IMO Raph has ideas that just don’t matter in games. The whole eco system he developed in UO was pointless to the gamer who only used the deer to grind his skills on and could care less that there was a predator/prey relationship with the wolves, etc. Instead of focusing on the fun factor he seems to take field trips to dullsville and become enamored with the local inhabitants.

He is more concerned about showcasing his human psychology theories than creating compelling gameplay.

Leaving aside the numerous posts that claim that I am not interested in content, the underlying points being made in these are interesting precisely because of what they point out about the game development process. It’s directly analogous to what a (justly) famous and multimillion-selling game designer said about my Grammar of Gameplay talk, when he called it “intellectual masturbation.”

First off, the question needs to be asked, “how much iteration typically gets to go into MMO games?” The answer is “not much by launch,” because the systems tend to be heavily interdependent, and there’s an enormous amount of technology that needs to be in place to do even the most basic sorts of things in an MMO. The result is that a vertical slice costs a ton, and you don’t get to throw it away once you have committed that much (even though you ought to, the “good money after bad” paradigm is alive and well when the “bad” money column has a very large figure in it; it’s a commitment fallacy).

Secondly, the question needs to be asked, “how do we get to that core game faster?” Because the faster we get to it, the faster we can decide whether or not to throw away the game (since we ought to be paying attention to Sturgeon’s Law).

The only real answer to these dilemmas is rapid prototyping. And this rapid prototyping can take a couple of forms:

  • It can be ways of getting us more quickly to the stage of tuning, such as
    • Technology platforms
    • Libraries of already-implemented stuff
    • Procedural ways of implementing stuff
  • or it can be ways of improving our visualization of the core game in advance, such as
    • Allegedly useless psychological theories about why players do what they do, which may help in prediction of how they will react to gameplay
    • Systems of notation that better capture the core mechanics of gameplay, such as the intellectual masturbation I apparently wasted that designer’s time with at GDC
    • Prototypes that do not use computers, but are still playable and thus can be assessed

If you detect a note of resentment here, you’re right to. I tend to take the preceding chain of logic for granted, at this point, in large part because I’ve just about never gotten to iterate a system of any sort prior to launch. The few cases where I have been able to (say, chat bubbles, moods, and emotes in SWG), it has had a measurable and noticeable impact on the quality of the feature. And what gets the iteration time? The thing that is smallest, often least important, because it’s what happened to get done first and has spare cycles for iteration when you’re engaging in massively parallel development. No, this does not make me happy.

There are three more ways of getting better at creating the first draft of something.

One of them is to mostly do the same thing you already did. This is a very common path.

One of them is to have so much practice at making drafts that you learn the tricks of the trade. This is the path followed by improvisers in all fields. (As it happens, I’m pretty good at this in several media; I’ve even been known to invent playable games on short notice sometimes, and I can improvise passable short stories live in front of an audience).

The last one is to be a shining genius.

Since you don’t get to improvise MMOs, and I make no claims to being a shining genius, and I am not all that interested in doing the same thing I already did, I’m going to keep going with my masturbatory theories about player psychology and game mechanics, even though they do seem to consistently lead to people concluding that I’m not interested in fun (to be frank, that was a major motive in writing the book: proving to people that I was!). If you’re reading here, presumably you feel the same way.

Frankly, it gets irritating seeing people out there making statements like those, because there’s generally a consistent misread on my motives. I spend time on this stuff because it’s what I can affect; what I really want to do is make content and entertain people, believe it or not. Otherwise, I wouldn’t spend all that time putting songs and stories and poems up on this blog.

Anyway, you’re already here, so I won’t take it out on you. 😉

  63 Responses to “Iteration in games: a mini-rant”

  1. Commentary by Raph Koster on the interative processes in entertainment mediums . If it’s good, it’s been polished.

  2. Iteration in game design. The general point being that Game Design is not the only creative field that depends upon iteration, but it is one of the only fields that often does not have iteration built into its development budget.

  3. *chuckles*

    I am, by nature, a theoretician. My favorite thing to do is to come up with ideas on how and why things are as they are and do as they do. My second favorite thing to do is a conflation: design, which is imagination + engineering. Thus, my third and fourth things to do is to imagine — to make stuff up — and to engineer — to put stuff together.

    My friends generally find me too whimsical and head-in-the-clouds to take seriously, most of the time, but I still seem to have captured their respect nonetheless. Being a dreaming idealist, the abstract and the theoretical are my constant companions, because nothing else exists in the land of the ideal and the horizon of the future.

    Sure, it’s intellectual masturbation. People say that like it’s a bad thing, to spend intellectual activity on something that gives you pleasure. But then they turn around and go, “Follow your bliss,” in the now-ageless words of Campbell. So what gives? It’s like there’s something wrong about serious thought that’s also pleasurable; I don’t buy. The establishment that Einstein spent the last years of his life to was founded on the premise that “useless” disciplines, like mathematical logic, should be pursued.

    I think that, without a doubt, games are important. There’s something about them that goes above and beyond the simple and dismissable. The puzzle of fun, the question of interactive aesthetics, the curiosities of social context: they’re as useless as String Theory. Or should that be, as important? Ludology is to social sciences what physics is to the natural sciences, or I admit, perhaps a very close approximation.

  4. Interesting. Evil Avatar thinks no other industry iterates, Raph thinks every other industry iterates much, much more.

    I agree with Raph. In the field of commercial software (and gods know, i’ve done too many years of that not to notice) iteration is regarded as about the only way of Getting it Right. Open Source software often iterates far more rapidly than Commercial software, with the end result that it is more stable and often provides more closely what users actually want.

    Google for Agile or Extreme Programming to find how iteration-based is currently being used to reduce the number of point releases and thus bring software more rapidly to optimal.

    And again, Raph’s point about parallel systems and costs is extremely well-made. There is a solution to this that leaps immediately to mind, but it wouldn’t suit many MMOs as they currently work, with levelgrinding and molebashing currently reigning supreme.

    That answer is the modular MMO, with each activity (game) existing within, not a vacuum but an amniotic fluid that we call the “world”.

    Then you can iterate and iterate on the “games” while keeping the “world” intact.

    Anyway, that’s my method. Pat pending, as usual *grin*

  5. […] https://www.raphkoster.com/?p=267I saw this comment tossed off in a thread on Evil Avatar responding to the Moore’s Wall post. That is the problem with Raph Koster’s ideas. No other entertainment media is iterative. Before a movie starts shooting, they have the entire film written, storyboarded, cast, locations rented, sets made, etc. Books have a complete story outline and usually a sample chapter written before getting a publisher’s go ahead. None of this is left to chance. . . none of it is left to a simulation. […]

  6. * Prototypes that do not use computers, but are still playable and thus can be assessed

    We used this technique a lot on our current project. We printed out lots of grid paper, drew big game maps, taped them all together, and playtested our combat mechanics before we even had our prototype up. I was in on some of the early tests, until it became rapidly apparent that I have such impossibly bad dice rolling luck that I was throwing off our statistics.

    “Maybe it’s your dice,” someone helpfully suggested.

    “We’re all using the same dice!”

    I think that if you have a team of people who focus primarily on rapid prototyping, they’ll get to a point where they have the tools and the savoir faire to bang out prototypes pretty damn quickly. It’s a lot harder when you take a team that has just spent three years plodding through the entire development cycle on a full blown commercial product, and then expect them to suddenly turn around and bang out a prototype for something else.

    I think that one of the reasons that theory is important is because while you’re presenting your Brilliant Design, you can point to features of the design, and instead of just saying, “This is a good design,” you can say WHY it is a good design.

  7. The points you make are relevent to other kinds of software development. As Cael says, iteration is a universal method for developing software, but I think for problems that are “hard” or “big,” that method breaks down. Raph, you say that this occurs in MMOs, because there is a huge infrastructure investment required just to experiment. In my own development group, we see it in the planning and scheduling components of our software. We have to build a lot of the system just to begin iterating on those components, and we can (and have) made bad decisions because we couldn’t iterate until late. We’ve always struggled with this, and having done so, I appreciate your motivation for developing tools to do better planning and experimenting earlier in the development process.

    As practical advice for presenting your work, perhaps you could acknowledge that iteration is a great way to develop many kinds of games and game subsystems. Then you can say that there are some problems where iteration is awfully expensive, and it’s *those* problems and game mechanics that you are trying to develop tools to tackle.

  8. Well if content is what you want to do, you’ve certainly ridden your career path as far away from it as possible. Maybe it’s time to tell big daddy Sony that Raph wants to get down and make a game he’ll be remembered for rather than another round of conference speeches.

    Your closing was odd because it assumed we all agree with you and already understand you. I read this site not because I’m a fanboy but firstly because you’re a friend. Beyond that I do believe your ideas are important to all of us. Frankly I’m glad that *you* (and all the intelligence and creativity that word brings with it) are spending the time to worry yourself out and I can then come and read the conclusions. We can all take your theories or throw them away and it’s just a passing moment in our day. However I agree with those guys up there, not in a malicious way, but in a tragic way… I want to play fun Raph games.

    I can feel your fire to express yourself and you put up weekly poems and write a song just because your daughter tried to ride a bike. You don’t write a three page explanation of how one would write a song about a daughter learning to ride a bike, you don’t do a world tour of speeches explaining the process, you write the song. So why allow games to be the one outlet of creativity you forgo in favor of discussion and an insidious fame?

    I appreciated your rant here because it explained your thought process on how you got to where you are now. That doesn’t get any games made. Your reputation for the intellectual and deep understanding of the *why* has led you down a road where there is no more *do*. The most wonderful thing I could see in a local game store would be a brand new “Raph Koster presents”… *anything*.

    You have the name and respect to get anything done, if Sony won’t let you do it, go to another publisher until someone does, but for the love of all that is Holy don’t retire as a CCO or Sr Executive Vice President of Who Cares. MAKE GAMES DAMMIT! You don’t have to stop telling everyone why you’re making the games how you are or how they should make theirs =) I don’t mean to belittle your job position, you have certainly earned it, but I don’t think it’s what you want. I don’t know if you think there’s no going back from being an Executive but you have already proven you can accomplish whatever you please. So maybe stop for a moment and see if your “intellectualism” is really getting you what you please. I feel a lot of frustration in your post but maybe it’s not all about people not understanding you.

    We get ONE ride on this planet my brother, take the trip you and your family would *enjoy* best =)

  9. My 2p is this… people always knock the theorists. Then 20-30 years later someone points out that they can apply *theory x* to their problem.

    Of course, all this means is that for 20-30 years people have been implementing without understanding. Which might be nice for them and even fun but ultimately is just process. Do you really want to be a robot in a car factory, or the guy who fundamentally understands how robots work and then specifies what they do?

    Or to put it another way – being able to systematically provide entertainment by copying and repeating, or by devising what is fun, why it is fun, then applying?

    As a former engineer I vote the latter. New science is more fun – and when it works it changes things. Then the engineers get replicating.

  10. Heh. I’ve been in a creative field myself (radio), and I can definitely concur about iterative creativity in that field. Ad-copy writing is especially notorious for its tortured creative process, with scripts and “spec spots” and storyboards upon which their creators lavished hours of creative sweat, tossed aside like used toilet paper. And there is enormous ferment behind many actual radio shows, with program directors and producers frantically scrambling on a daily basis to generate fresh content and fresh laughs.

    That said, I think the debate over iterative games is a useful one, especially now, because the iterative process has crossed over from games in development, to games that have been pushed to Live Retail.

    At what point do different versions of the same game cross over from being “the same game” into what should be described as “a different game?” What about new iterations of old movies? Things like so-called “director’s cuts” of popular movies? Or George Lucas going into his Star Wars movies and replacing actors in certain scenes with shots of the actors who played those parts in later movies? (Two examples that come to my mind here are the scene in The Empire Strikes Back in which Lucas replaced the original actor who portrayed a holographic Emperor discussing Luke Skywalker with Darth Vader to Ian McDiarmid, and replacing the original Anakin Skywalker we see in ‘ascended’ form at the end of Return of the Jedi with Hayden Christensen.) Are they the same movies? Only the hardest core of movie purists would say no.

    But now, if Lucas had replaced Darth Vader with, say, Pinky and the Brain from Animaniacs, THEN people who loved the original movies would have ample reason to gripe.

    Enter Star Wars: Galaxies, stage left.

    SWG underwent two massive revisions in 2005. The first one, the so-called “Combat Upgrade,” changed the way the core combat system was managed, but still retained the quasi-turn-based flavor of the original game. The second, one, however (the “New Gaming Environment,” or NGE), fundamentally altered the core combat mechanic itself, changing it to a quasi-first-person-shooter model And most of the PC game-news industry characterizes the NGE as a new game entirely in all but the name of the game.

    Here, SWG opens up a couple of Kosteresque lines of discussion: What makes a game ‘fun’? And how do game developers get to that ideal ‘fun’ model — and can they experiment with an existing game?

    As such, SWG presents an interesting case study.

    If we accept the premise that choice and consequence is the core dynamic that gives a game its “fun factor,” then SWG’s new click-and-shoot combat model can be seen as unfun, because the person controlling an avatar in combat is essentially bound to a single static action: Repeatedly clicking a mouse button. Sure, the player is given a row of “special attack” hot buttons, but they just don’t seem to have all that dramatic an effect. Basically, all they do is trade off varying amounts of extra Action Points for varying amounts of extra damage. And their re-use timers vary somewhat. On paper, it sound rather dull, but in practice it’s even worse. By contrast, the original combat system had several VERY dramatic effects: You could target any one of three different character stats, or all three with a single massive attack; or you could inflict a ‘poison’ attack which would repeatedly tick off points from the target’s affected statistic; you could use an attack which reduced the target’s maximum Health (especially notorious here was the Bounty Hunter’s Torso Shot, which would set the victim on fire), limiting the ability of the target to regain his health on-the-fly; or you could knock him down or hamper him with Blindness, Dizziness, or a Stunned or Intimidated state. Or, you could call upon a pet to fight alongside you.

    These options are either gone or severely limited in the New Game Environment. It certainly doesn’t feel nearly as multidimensional. And so, the combat model breaks down. Yes, it’s more interactive — but who wants greater interaction with a more tedious system? Not me.

    So, what can be done to deal with such a situation? With the first revision, Sony Online spent months preparing the in-game community for the Combat Upgrade. By contrast, the New Game Environment was developed in complete secrecy, perhaps even while the Combat Upgrade itself was in development! This apparent duplicity on SOE’s part (or was LucasArts the culprit here?) complicates the situation immensely, and is worthy of discussion in it’s own thread. But for now, it’s a digression.

    SOE had several resources at it’s disposal, all of which went unused prior to the imposition of the New Game Environment upon it’s customer base. The first was it’s Test Center servers. Rather than gather feedback from the SWG player community as to the quality or desirability of the new system, the developers rushed the thing, bugs and all, through a ridiculously inadequate two-week run on Test Center. SOE’s second resource was the gamer Forums. SWG is rather notorious for having a very loud, vocal community. But in this case, that community was kept completely in the dark. Amazingly, the veil of secrecy was maintained right up to the day Julio Torres dropped his big red-letter bombshell on us announcing the NGE. And when the swirling hurricane of player protests reached unmanageable proportions, SOE gutted the Forums themselves. They even ended the volunteer Player Correspondent program, which appeared to be invaluable at one time for locating in-game bugs and issues affecting particular professions.

    So now, the problem for SWG appears to be infinitely worse. It’s earned an unenviable rep as a horribly bugged, poorly-realized attempt at a game, with scathing reviews such as GameSpy.com’s Green Banana Award for 2005 littering the landscape.

    So I’m back to my earlier question: What to do about it? Clearly, both Sony Online and LucasArts have dug this game into a very deep hole. The only real solution to me would be for them to backtrack all the way back to the point of origin for this mistake, and then re-think this whole situation from that point. And in my opinion, the point from which SOE/LucasArts diverged from the path of creating a fun game is found in the theoretical question of what makes a game fun. Their error is the apparent premise that a first-person shooter combat model is inherently more fun than a turn-based combat model. This ain’t necessarily so, as Perptual Online discovered in the course of their pre-development research for Star Trek Online. Their findings directly contradict SOE’s assertion that a FPS-style “fast-combat” model is the way to go. For their report on their research, check this link.

    And so, we return to my earlier observation: that the NGE breaks down because it binds the player in combat to a static action with limited tactical choice.

    Somewhere, the game-creation process broke down while the NGE was in development, probably because the devs were under a severe time deadline imposed to satisfy the marketing priority to have a new retail product ready to roll on the same day the Star Wars: Episode III DVD was released for sale. Whatever the reason, someone pulled the trigger to release a product that represented a significant downgrade to the quality of the game’s fun factor.

    And that doesn’t even begin to address peripheral contributing factors, like the effect the level-based combat damage multipliers (the ones introduced with the Combat Upgrade last spring) have on players’ interactions with the worlds in these Galaxies, or the reliance on generic Creatures Of Mutable Level from mission terminals (I mean, come on, Ralph… level-82 kaadus??), or the complete elimination of once-popular professions like Creature Handler. For my part, the imposition of level-based combat-damage multipliers has effectively ruined most of the worlds in SWG. What’s the point?? Once, all those worlds were vast playgrounds. Now, the only places worth exploring are those where I can find creatures that fit within my characters’ narrow challenge parameters. Too low in level, and they’re a waste of time. A few levels too high, and they’re purple-Conn instant death. I want my game back!

    For me, the bottom line is starkly simple. I don’t see SOE or LucasArts showing any hint of willingness to admit they made a huge, monumental series of mistakes in 2005, so I have chosen, with great reluctance, to allow my subscription to a game that I once loved, to expire.

  11. The thoughts on iteration here are very interesting, powerful, and I tend to agree with your sketch of the totality of the culture industry. I want to note that at the start so I don’t appear to just be going to the one thing where I have a more pointed response.

    I’m definitely one of the folks who have from time to time suggested that there is an issue with your concern for systems and your modellings of player psychology, which dominate even A Theory of Fun, inasmuch as you largely understand “fun” as the epiphenomenal consequence of a cognitive prediposition to puzzle solving and pattern formation. I think that’s a good case of where theory can lead to a problem in practice that stands in serious opposition to the other impulse that shines out so strongly in your writings here (and in the pre-blog pages) about creativity and expression, about “content”.

    The problem with the systems and cognitive theories is that they lead you to a place rather similar to where some of the structuralist ludologists end up, where play and fun are largely secured by making a match between fixed structures and predicates in the minds of players or by designing a perfect system or mechanic. When you’re driven by that concern, it doesn’t matter how much you care about content in your other persona or how much you believe you’re driven by content and story, content is going to come last because you don’t believe, at that moment, that it is the dependent variable that makes a game work or succeed. Much as Espen Aarseth said that it didn’t matter to Tomb Raider that Lara Croft was a busty woman.

    I think there’s a genuine tension, even contradiction, in what you say you want to do with games, and even more, what you say about them, even in A Theory of Fun. Even in a Theory of Fun, the way you talk about content is in a cart that very much follows on systems-thinking and models of cognition, again, as a difference which matters in deciding whether a game is aesthetically successful, artistic, communicative, expressive but not in how it produces the experience of play (and fun).

    I know I disagree with that myself: that the systems of SWG, as interesting and sometimes compelling as they were, could never achieve their “fun” potential without content, and content that went way beyond mechanics like chatting or emoting. Without a sense of how to create the experience of being within a gameplay fiction that bore a mimetic resemblance to its source fiction, that re-enacted the consumption of its root fantasy in the body and mind of the player. I think that’s also what explains some of the success of WoW, a factor that its critics (and those envious of it) tend to dismiss or minimize–that the visuals, the narratives, the representational scheme of its world, aren’t just a bit of frosting on top of a game-mechanical cake, but the keystone on which the satisfactions of play in WoW often rest.

    What this suggests for iterative work in MMOGs is that maybe we need to stop talking about game-mechanical innovation for a while, at least until there are real technical changes that make meaningfully emergent AI possible, and start talking about content-creation models, about how to “re-skin” some of the same old mechanics to make compelling new content-worlds, new stories. And about which kinds of content should be simply avoided until the technical capacities or design features appear to support them–this is one reason why I think a MMO based on Lord of the Rings is such a terrible idea, because the content almost demands game-mechanical changes that I don’t think Turbine (or anyone) has a clue about how to bring about. But this is not so for many stories, many representations, much creative or imaginative content: the kingdom is not always to be found in the structural, the mechanical, the systemic, the cognitive.

  12. David that was awesome… WHOAH. I was worried I was going off topic by discussing the last few paragraphs compared to Raph’s suggested intent of “iteration”… but… wow. Hahaha.

  13. I don’t know what kind of time you have on your hands, but making a few tiny games using your theories would probably convince a lot of people that (A) you’re not all mouth and (B) you’re not all academia.

    Even if you don’t make them, you might think about getting a few college students to do so. They come cheap, and I know dozens that would do it just for the chance to say “I worked for Raph Koster”.

    In short, the AAA titles you’ve been working on have only proved that current game development systems are wrong, not that your game development systems are worthwhile. The first is obvious, the second needs to be addressed.

  14. Raph, I can see where Q is coming from. You and I had a conversation at GDC two years ago while waiting for a session to start. I breeched the idea that you should use your name to develop the game that you want to develop and the feeling I got from you was that you weren’t ready. Mind you this was two years ago.

    The only way you’re going to prove that all this theory and such is worth pursuing is if you can apply that theory to a game that is successful and its success is largely attributed to your design theories. SWG and UO allow the excuse of an existing IP that is the cause for the success of those games. The naysayers will shut up if your name is staked to a game raking in the cash. Frankly, I believe this line of reasoning will create a fun and immersive game beyond anything that we have out there today. Alas, I do not have the name as you do which is why I and others would love to see you take that advantage and turn it into The Next Big Thing(tm). Accept the fact that your name has become synonymous with MMO development in many circles and use that power to influence the direction of the industry as do many of your peers in other genres.(Wright, Meier…) BTW, tell me where I can sign up and let me know if you need help. With your name and the team that will follow you to The Next Big Thing(tm), I would think it would be easy to find the venture capital to found the company.

    Something to chew on but then again, I thought it was something to chew on 2 years ago…

  15. For those who like a little more background on things:

    The form of comic book writing which Raph refers to in Gaiman’s book is commonly called “Marvel Style”, in which the writer provides a page-by-page (sometimes panel-by-panel) story description, then the penciller draws the book, then the writer goes back and dialogues the book. This was how Stan Lee and Jack Kirby reportedly worked in the 1960s, and it was quite different from how comics had been created before. Famously, in one issue of Fantastic Four Lee got the pencils back from Kirby and it included scenes with a whole new character who hadn’t been in Lee’s script: the Silver Surfer. (This story might be apocryphal, however.)

    The major alternative method to Marvel Style is Full Script, which is about what you would imagine. Alan Moore’s intensely detailed style of scripting is widely known.

    There is, probably, less iteration in comics than in many other media, not as a function of the medium, but because of the deadline pressures in producing a monthly book. One could argue that Marvel Style production was invented in part to encourage more iteration while still meeting deadlines (though I don’t know that this is actually true).

    For a hilarious – but, I suspect, dead-on accurate for many people – fictional take on writing a novel, complete with loads of iteration, take a look at Edward Gorey’s comic strip “The Unstrung Harp (or, Mr. Earbrass Writes a Novel)”. Revising a novel, is, of course, iteration in a pure form.

    And in any event, writers differ widely regarding whether they write from an outline, how much they revise, etc. (Rumor has it that George R. R. Martin threw away hundreds of pages of text while writing A Feast For Crows and backed up and tried again.)

    All of which is just more support for the notion that the poster Raph is responding to doesn’t know what he’s talking about. But I have all this trivia in my head and it has to go somewhere… 🙂

    (I am, BTW, a new reader. Hi!)

  16. Raph’s notes about Popcap are particularly interesting given the relatively similar, derivative nature of most casual games. I’d played a variety of other games that had an essentially identical mechanic to Popcap’s Chuzzles, for instance, but Chuzzles is a vastly more fun, satisfying game than its competitors due to its polish, particularly a control scheme that feels ‘more right’ than in other similar games. There’s a great deal of purity to that, since the only thing a player needs to do is to drag a row of objects up and down, or left and right. One wouldn’t immediately think that one way of implementing that is any better than any other, but that turns out not to be true.

    I think that the quality of the UI makes a huge difference to an MMO, for instance, yet I don’t believe that these UIs have gotten the same kind of attention that user experience receives in other industries. It’s one of the places where polish is hugely valuable.

    UI, notably, is one of those places where it should be cheap to do lots of prototyping. (Or, I suppose, you could make it scriptable and let your players do the work for you, as WoW has done.)

  17. The funny thing about WoW is that the UI is arguably too open. Anyone that has logged on one tuesday after a major patch (9 – 14 of these happened last year depending on your definition of major) can attest to the problems of the UI. I recently took a week hiatus from WoW because I didn’t feel like going through the hassle of updating my UI after the 1.9 patch. That and I needed the player-developers to release the updates in the first place. Good thing I did or in successive weeks I would have had to update to 1.9.1 and then 1.9.2. Perhaps it’s their scriptable design that’s flawed but I tend to believe much more of the UI needs to be taken care of by the company instead of the player base. In anycase, UI does make a huge difference and recent games have proven that its difficult to do right for everyone.

  18. Rapid proto-typing for me (2 released commercial games this year, one getting close to release) is all about having a small, “agile” team which can try out lots of ideas quickly. I think it is necessary to first lower one’s standards about controlling for player psychology and certain “grammars”. Get a decent engine designed and then just try stuff.

    Of course MMOG’s face the difficulty of having to “try” things with thousands of players before having a realistic idea of how it will play out in release not to mention many other details.

    Which is why I think more and more that truly massive MOG’s just aren’t the place for innovation. MUD’s were much better suited for rapid prototyping. I could revamp my mud’s combat system, add a new class, add a new zone, etc., in a day or two, then get immediate feedback from a small playerbase whose particular psychology matched the design ideas I had to offer.

    To echo what others are saying, I don’t think you’re going to prove any of your ideas to anyone without having them work in a given game system. And I don’t think a large publisher like SOE is the way to create these game systems. Innovation is going to happen on the periphery of the market. Businesses like Sony will just pick up on whatever emerges from the periphery and is big enough to market to the mainstream.

  19. Well, now I am all depressed because nobody instantly answered with “but you ARE a shining genius!” Ah well. 😉

    More seriously:

    As practical advice for presenting your work, perhaps you could acknowledge that iteration is a great way to develop many kinds of games and game subsystems. Then you can say that there are some problems where iteration is awfully expensive, and it’s *those* problems and game mechanics that you are trying to develop tools to tackle.

    Yeah, I guess I feel like I have said that before, but it can always stand to be reiterated.

    I just replied to StGabe in another thread stating,

    I don’t think anyone creates anything in any medium based initially on the grammar.

    In other words, knowing there’s a grammar is helpful, and it often guides us into new and fresh areas, or it helps us learn our chops (I saw in your newly launched blog that you emphasize how important it is to just learn chops). It’s like learning scales in music. You don’t start writing music based on a scale. But internalizing the scales is hugely helpful.

    Really, a goal with a grammar is to lay open the field of possibility. The very boundaries of the grammar suggest new ways to develop ideas — and often ways that lie outside the grammar itself.

    So, in short, I’d never try designing a game directly with a grammar. But it can help us anyway, both in understanding what we are doing, and in broadening our palette. IMHO, of course.

    I think that, perhaps, is a key thing that I haven’t expressed before. In all of the various training that I have in various art forms, there’s always a sort of formalist phase you go through, where you end up developing your skills by slavishly following forms and theories and stuff. But eventually, you break out of it. You’re much the better for it, though.

    I don’t design a game by starting from theories. Like everyone else, I start with “wouldn’t it be cool if…” So that’s a significant misapprehension of what I am like that colors pretty much everyone’s opinions of what I say.

    Well if content is what you want to do, you’ve certainly ridden your career path as far away from it as possible.

    Keep in mind that I don’t mean content in the sense of “levels and stories and mob stats,” but in the sense of “gameplay,” which to me rolls up all of that stuff. Part of why I took this job was because I saw it as an opportunity to direct more resources towards finding more ways to improve gameplay.

    Your closing was odd because it assumed we all agree with you and already understand you.

    Ack, no. What I was getting at is that if you’re here you’re not dismissing this stuff out of hand. If you were, like some of those Evil Avatar folks (and folks elsewhere) you wouldn’t even bother reading it. So by your presence, you’re showing at least neutrality, if not interest. 🙂 I think the discussion threads are evidence that there isn’t at all unanimous agreement with what I say. Nor would I want there to be.

    I think one thing that people also often don’t get is that I like to think aloud, and get reactions from people. I’ll even argue a more radical position than I actually hold just in order to get a response. I do that a lot, and many of the more notable quotes that people bandy about were in fact phrased that way in order to get a rise out of people, and stir up the discussion pot.

    MAKE GAMES DAMMIT!

    I don’t know what kind of time you have on your hands, but making a few tiny games using your theories would probably convince a lot of people that (A) you’re not all mouth and (B) you’re not all academia.

    Since I took this job and moved to San Diego, I have in fact:

    • designed four tabletop games, two of which I am pretty darn happy with and one of which I think is really, REALLY good despite being incredibly simple
    • designed around 40 small computer games, mostly arcade and puzzle (sometimes both at the same time)
    • prototyped around 14 of them (meaning, functional code to the point where you can play the core gameplay and see if it’s fun)
    • more fully implemented around 9 of them (to the point where you can actually play a full game, advance through the content, save a high score, etc)
    • “finished” 5 of them, meaning they are complete enough, including sounds, music, and graphics, that I have been willing to share them with friends and pass them around as playable games. They could all stand a couple of months of polish, but they are fun and could probably be sold commercially with some more effort
    • designed to the 30-50 pager level two separate MMOs, and one more to the 10-pager level

    So believe me, I am making games. 🙂 It’s just that nobody is getting to see them, which is a whole different issue from what you are describing…

    Tim, I need to owe you a MUCH more detailed response in a separate comment.

    (I am, BTW, a new reader. Hi!)

    Hooray! Where did you come from? I am always curious about the spread of the blog’s readership (which is, btw, growing by leaps and bounds every month).

    Also, add yourself to the reader map. I want to have a special post when we break 100. 😉

  20. OK so wow. That’s pretty impressive.

    However you can’t be surprised when the public perceives you as a pure intellectual if you don’t share the games with the world. Since you call it an “issue” yourself then I’m sure you fully understand. =)

  21. No, a shining genius would be shunned for the first fifty years of his life, viewed with suspicion for the remainder of his life, and lauded as a shining genius fifty years after his death. I don’t know if that’s worth genius. =P

    Raph, you might consider taking some of your better mini-games and adapting them for the MMORPGs you’re on. Matt Mihaly’s Achaea has a wonderful little chess game that I’d make a point to play every now and then. It doesn’t make the MMORPG itself any better (well, maybe it does), but cult(ure)s can develop around such things… Oh, and you could let slip that it’s your design. =P

  22. BTW, I’m a new reader too, from Rocket City, USA… home of Marshall Space Flight Center. And you’re right, Ralph. I chose to post in this thread because I value thought-provoking discussion, and it seems this site is a haven for it.

    I wasn’t quite sure whether to direct my earlier comments about SWG into this discussion on iterative creation, but I hope it sheds an interesting lighting angle on the discussion: Basically, at what point do we (or should we) cease monkeying with our creative output and let it be what it is?

    We can approach the issue from yet another direction, and contemplate one of the “holy grails” of MMORPGs: Dynamic, changing content. Sony Online’s made a few attempts in this direction, with multipart storylines and such. If such things were the rule rather than the exception in a MMORPG, such updates would also qualify as “new iterations” of the game. And this seems to be something game developers are striving for. Turbine recently revealed a few of it’s tools and techniques for creating quests and stories for it’s forthcoming Lord of the Rings Online. It’s intriguing! Basically, their system revolves around a quest-scripting toolbox that appears similar to Neverwinter Nights’ level-creation toolset. The author of this article implies that previous developers for other games had to craft quests individually. This toolset appears to allow developers to string together ALL the elements of a quest — dialogue, NPCs, target enemies, scripted events and event triggers, loot, etc. — together in a cohesive tool set. Shoot, I’m beginning to think these just might be the sort of tools a great raconteur could learn to use to craft compelling stories within this burgeoning computer-game genre. Heck, I may even get brave and take a crack at it, myself.

  23. Quest creation tools aren’t really anything new. The big strength of EQ, for example, was its template-driven quest architecture. That said, the tool they’ve got sure looks slick. 🙂

    The real challenge with all template-driven quest systems is that they reduce down to the same basic templates. The real challenge is finding ways to come up with quests that offer new narrative potential. That’s what generally requires custom scripting. The Beowulf quest is an example of something custom-scripted that has behaviors not typically possible in a quest engine. That said, once you develop the behavior once, you can add it to the engine if the engine is flexible enough.

    Oh — and David, SWG is a good illustrative case for something that I mentioned in the post. If it helps, you can think of SWG as being a script that is being performed by a different troupe than that which originated it. I generally make it a policy not to comment on how people (whom I like and respect and work with) are handling something which I was once involved with but no longer are. It’s just too easy for that to backfire in ways I do not intend. So you’ll always see me skip over discussions of SWG’s changes here.

  24. Raph, SWG is always going to be the millstone around your neck and to be honest your response No. 21 is going to be viewed as a dodge of the highest order. Repeated referencing of the Beowulf quest is irrelevant re. SWG

    SWG changes (re. NGE) are pretty much irrelevant, the important issue is the initial implementation. Who designed the original HAM system? Who thought mission terminals were *content*? Complete lack of overall narrative content? Skill systems completely FUBAR or worse (Bio-engineer, Smuggler, Droid Engineer etc. etc.) Who is responsible? Koster? Vogel? Blackman? Raise your hand please…

    Anyone with Google and half a brain can find reference to the tensions between Austin & San Diego, the overall *relaxed* development atmosphere, Koster quotes from the pre-release boards. The sooner you acknowledge the disaster of SWG the sooner the threads on Evil Avatar, F13, CorpNews, the official SWG boards ad infinitum will stop bashing you round the head with them… The sooner you might fix your reputation in the game development field.

  25. Haha Vogel. Yeah he made a skill system. HAHA

  26. In a nutshell, AJ, I just can’t talk about it for numerous reasons. Maybe some day. 🙂 After all, I did eventually write a UO postmortem of sorts.

  27. I imagine, AJ, that what Raph is saying obliquely in response 21 is that it would be unprofessional for him to talk about what has happened with SWG. I imagine that there are corporate nondisclosure agreements involved, as well. A goodly number of companies these days also have nondisparagement policies — i.e., employees are not allowed to be critical of the company in public.

  28. Well, this might sound disingenuous, but I hope my post (#8 above) doesn’t come across as a flame. (grin…) An honest, sincere rant, yes… but not a *flame.*

    Anyway, I am a firm believer in the importance of a firm theoretical foundation as the base for creative work. I find that when I don’t prepare myself on a theoretical level going into a project, my productivity suffers in myriad ways. I become haphazard in my idea development; I become tentative; I fail to focus my energies on the actual problem; I wind up solving the wrong problem; I get writer’s block; etc.

    Sounds like a game developer’s worst nightmare, doesn’t it?

    Actually, it’s ANY creative person’s worst nightmare.

    I think the best question any creative person can ask him/herself at the outset of any project is, “What are we really wanting to accomplish here? No… REALLY… down at the very core level: What are we REALLY wanting to accomplish here?”

    The question often has several possible answers. Which answer you choose depends on how deep you want to go in uncovering your motivations. But when you uncover that magical REAL goal, amazing things happen. It energizes you. It frees you. It eliminates doubts and distractions. And I find it makes me ten times as productive, at least.

    So, Raph (or is it Ralph? I’m confused….), theorize on, bro!

    But now, back to SWG….

    This is what puzzles me. I can’t figure out exactly what it was you were trying to accomplish with the game.

    I mean, yeah, you and the original dev team crafted a truly amazing universe. Ten huge worlds, detailed out to the last hillock and blade of grass. Fascinating indigenous life forms, dozens of competing factions, cities, loot, combat systems, crafting metagames, sabacc tables, dungeons, caves, loot… on and on it goes.

    But I would really like to know: What were you trying to DO with all this “stuff??” So much of it was just sorta…um, there. The Narglatch Cave. The Nightsister Stronghold. Lord Nyax’ compound. The tribespeople on Dantooine. The Blue Leaf Temple. And on and on.

    I found my mind wandering down endless branching pathways: “Just what ARE the stories behind these things?”

    Infuriatingly, the game gave me precious little opportunity to find out.

    Whenever I traversed the worlds of SWG, I couldn’t help marveling at the untapped storytelling potential of it all. The Wild Force Wielders prowling the steppes of Dantooine… what were they looking for? What if I could establish a relationship with a Dantari Shaman… would he be able to give me some insight as to the nature of the Force as it manifested itself on Dantooine? What if I could become palsy-walsy with the Nightsisters on Dathomir? What was Exar Kun up to? Did he have a hand in the evolution of the half-humanoid, half-crustacean Mereks of Endor? Are the mereks sentient or nonsentient? Or are they sentient in ways we are unable to comprehend, communicating among themselves in a language comprised of ultrasonic clicks and chirps, punctuated by pheromonic markers? What if we could do a database search or randomply browse in the Theed Palace Library? Could we uncover a dangling loose end or two that leads us to a fabulous quest? What if non-combatant Entertainers could stumble across the start of a text-based adventure that leads them to fantastic locations: Dancing before Jabba the Hutt, with a bad guess as to his whims leading to a swift death in the Rancor Pit, or infiltrating the Neo-Cobral hierarchy?

    You and the SWG dev teams deserve kudos in many cases. Y’all can write some damn good stories when you put your minds to it. I just wish I knew why the original worlds were so underutilized.

  29. Well, the easiest answer would be to elevate the role of design documents to art.

    arrangements of four blocks/falling like leaves in autumn/farewell, complete row!

    It seems probable that there is some loss of utility by this method.

    Being but a mere student right now, I can’t comment on how much iteration happens behind the closed doors of various development factories, but it doesn’t seem anybody is really trying to iterate beyond approaching some mysterious holy grail of Game Balance. Due to the time span and multiple instantiations of a massive, it should be possible to perform iterative design on the game as time goes by. There can be some test server where alterations on the basic gameplay, and on narrative structures and all that delicious contenty stuff are tested in such a way that they can be rolled back and tried again. Instead, we get to watch as numbers are shuffled into Working As Intended. I can’t comment on the NGE, as I’ve never really followed Star Wars Galaxies, but it does seem to have been handled in a Bad Way. Which is also the case with the rest of Star Wars Galaxies, so it’s par for course? And just to be cruel, I don’t see what really separates the Beowulf quest from any number of “deliver this I.O.U.” quests, besides involving more interesting challenges than foozlestomping. And stylistic writing, of course.

    P.S.: I’m a first time poster from the Terra Nova mess, and your frappr map has busted 100 people.

  30. David: we actually talked very openly about what we wanted to accomplish. I think that the “What My Job Is” post you can find in the Gaming/Essays section of this site conveys some of it. And AJ was there (under a different name), so he can talk about it too. 🙂

    WF:

    What, you want MORE than just more interesting challenges and an engaging narrative? What, exactly? (I ask only partly in jest).

    Terra Nova is a mess?

  31. Not to suck up but I’d vote for Raph as a genious in several fields of communication and vision. Much like a skilled scientist who has the ability to inspire listeners into taking on a difficult problem with their minds and possibly time. ^^

  32. I will give Raph the answer he wants. I think he is a game genius.

    SWG was my first MMO, so I don’t know the previous games. But after SWG, I have played wow, lineage2, eq2, daoc, dnl, guild wars, so I have a good idea about mmo games.

    I feel sad when people point at SWG and say it’s a failure and / or take it as an example and comment ‘sandbox games will not work’ and praise wow all the time. In the Evil Avatar, the same thing happens. Raph is criticized for SWG, its sandbox content, crafting system, etc. But the problem were never those.

    I know the problems exactly, because I have watched my friends leave the game one by one, in the end me joining them.

    Sandbox content : Factional bases. Perfect idea and content. But it had flaws. What flaws?

    1) Taking a base or defending it did not grant a reward that we can feel, show or touch in the e-world. The one and only thing it granted was bragging rights. We asked repeatedly in forums for rewards. We were ignored.

    2) Ignoring problems raised by playerbase: We have repeatedly reported that turrets were shooting at incredible speed allowing a single defender to kill an entire army. Same for combat medics. We were ignored in both cases. We were told things were working as intended.

    Then, we started not to raid bases anymore. We let one of the biggest sandbox contents ingame sit there unattended, we deserted it. It forced us to leave the game one by one. SOE watched very loyal players leave the game because they refused to accept and fix a problem.

    Those base raids were prepared in secrecy, spies were used, meetings were held. We used to setup rally points, and gather silently. It was a long preparation. Just think of losing all your hours of work in seconds repeatedly. How many times can you keep doing it ?

    I will jump to wow for a similar example. I don’t remember the name, but one of the bosses in an instance was reported to be too strong for the 10 men raid group. 2 weeks later, that boss was fixed to match that level.

    BH-Jedi conflict is unique among the games I have played. I am sitting in a cantina, not a pvp zone, talking to my friends, suddenly I see gunshots and a saber igniting. In all other games, you see the fight coming. BH-Jedi conflict is something (almost) both sides enjoyed.

    This goes for player cities, too. Was a very good sandbox content. There were parties, tournaments organized in player cities. Players prepared dances for events. But when we asked for some powers to remove condemned houses, powers for mayors, other stuff, we were ignored. This led to to concrete graveyards. Another sandbox content ruined.

    In summary, I think Raph’s SWG theory and implementation was quite unique and successful. But the follow-ups were awful. And the follow-ups were probably technical stuff, not Raph’s fault.

    One thing is certain: Raph created the strongest social community in SWG among all games, and at least 1/3 of current SWG subscribers are still playing SWG despite their dislike of the game because of the community bonds formed in early times. It’s like a pub for a good number of people where you go and talk to your friends after work. And these bonds were formed while raiding bases, helping each other’s Jedi to survive, etc. all sandbox content.

    I, and a lot of people are waiting for Raph to make another game as deep as SWG with a better team who will not ignore follow-ups.

  33. I use “mess” in the most loving of ways.

    As for what I’d want? Off the cuff answer would be a quest structure that exploits the fact that massives take place in shared, persistent worlds. This could be implemented in that quest results have impact beyond an alteration to the player’s vocabulary. So, stopping the foozle hordes actually means they won’t invade Capital City, Humansylvania. The game would have to be structured so this is a possibility, yes. It should also be noted that you don’t lose the possibility for interesting challenges if smartly implemented. I will grant that the Beowulf quest had some really neat challenges, made possible by LegendMUDs expanded vocabulary. They generally still boil down into fetch quests, although some are maze challenges (finding your way into Rabin’s house).

    I generally don’t feel engaged by the narratives of the quests I go on. I recognize that even if I defeat Fadhnir or free Modhthryn (sp to both!), they’ll return and the world will be As It Was for the next person to try the quest. Maybe this is just me, but I’ve always been of the opinion that Fadhnir could only be slain once, that the Holy Grail could only be found once, that Sleeping Beauty could only be awaken once. There are ways to write and to generate quests that account for this. Or maybe I should just get the willing suspension on my disbelief checked?

    I’d also like to see another massive out of Raph that shows Lessons Learned from Star Wars Galaxies.

  34. Now I’m torn between magnanimously saying “more genius suck-ups, go right ahead!” and going into hiding. 😛

    Oropher, you’re effectively bringing us back on topic regarding iteration! 🙂

    I will grant that the Beowulf quest had some really neat challenges, made possible by LegendMUDs expanded vocabulary. They generally still boil down into fetch quests, although some are maze challenges (finding your way into Rabin’s house).

    Oh, yes, of course. The reason to point to Beowulf isn’t to say “look at this radically different technology,” it’s to say “our tools can get pushed a lot further than they do now.” I’m not making the claim that Beowulf is the greatest quest ever written; instead, I offer it up as an example of better writing in quests, and as an example of some slightly varied mechanics in service of the story, and as evidence that I do in fact make content and not just sandboxes. 😉

  35. Raph, this point is what intrigues me about the ferment going on in computer games:

    I’m not making the claim that Beowulf is the greatest quest ever written; instead, I offer it up as an example of better writing in quests, and as an example of some slightly varied mechanics in service of the story….

    One of the best elements of the iteration of SWG known as the New Game Environment was the addition of quest-based experience-point gains for the Mustafar quests, and subsequently, for Kashyyyk. To this day, I remain convinced that if this improvement had been made within either of the two previous game environments (the original game or the Combat Upgrade), it would have done FAR more to strengthen the “StarWarsiness” of the game than yet another change to the combat system. I consider the combat mechanics to be a means to an end, the end being the chance to pretend to fight stuff. And it’s the stories — the quests — that give the battles meaning. And I think this really focuses players in a positive way on those stories.

    It sounds like Turbine is preparing to take this quest-based XP model to it’s logical extreme, with the publication of Dungeons & Dragons Online. To wit, NO direct xp for killing creatures. At most, you get bonuses for clearing a dungeon. I’m not sure that extreme is such a good idea, but it’s definitely noteworthy. I wonder how the powergamers (those who like to level their toons as rapidly as possible and zero in on the best loot available) will react to this next-generation MMORPG. (Or is it even truly a Massively Multiplayer RPG with it’s all-instanced-content approach?)

  36. Isn’t Vertical Slicing a virtual world rather difficult? MMOs must entertain for hundreds of hours. Single player games only need to provide entertainment for 40 – 50 hours on average. I can see a Vertical Slice created for Mario Kart or for Resident Evil because like you suggest the elements of gameplay and content will not vary from that Vertical Slice by a huge margin. In this case the slice is representative of how the final work might come out.

    MMOs on the other hand are a beast that Vertical Slicing doesn’t seem fit for. Many people asked me to vertically slice Ages of Athiria but I’m not capable of boiling the design down to a single thread of an idea. Wouldn’t that imply that all other subsystems are subordinate and therefore not as important? If the slice is supposed to define your game, then the rest of the game can only be seen as dressing on the slice or you lose your vision as portrayed by the slice you chose. Unfortunately, if I had to choose a slice it would need to show how all the subsystems of the game interract to create a living virtual world but I might as well just create the game given the funding needs of creating such a slice.

    Don’t we have problems fitting a single playstyle into the a large playerbase already? Vertical Slicing doesn’t seem like a fit for MMOs and I’m not sure why anyone would try it other than to secure funding.(different problem all together) How does one build a world where politics, merchants, military, hunting, combat, sailing and thievery are all important parts of the world? There’s no one way to slice this such that a publisher will come away with the feeling that it would be fun. Hence the dillema. If you Vertical Slice your game then you predispose your game to a very narrow design(the one you received funding for after pitching your slice) which likely will limit your subscribers. The method in this case seems more dangerous to me than worth it. There are projects in life that require large amounts of money and huge commitments of vision. Would Walt Disney have launched Disney World by pitching a single ferris wheel to investors? Not a chance. Will the revolutionary MMO come from a team that pitches a combat system or a trade system or a political system… Nope. The genre is still too much in its infancy for iteration to be the end. Iteration of design from game to game seems to me like a sign that I should get out of the industry because margins are going to start shrinking as games become more and more iteratively different.

    I don’t think the problem of creating these huge games can be broken down any further to get to a slice or core gameplay. For me, this is why the understanding of social dynamics, game design grammar and other forms of theory are so important. Definitions, categorizations and testable theories all lend themselves towards helping a designer wrap their hands around a much bigger project. Without them, you have no choice but to iterate from game to game or patch to patch meaning your doomed to be just like what came before you only slightly different in the same way that the majority of MMOs share core game mechanics.

  37. Raph writes:

    Hooray! Where did you come from? I am always curious about the spread of the blog’s readership (which is, btw, growing by leaps and bounds every month).

    I came from Ceejbot. She is, after all, the one who got me started with on-line journalling (as we called it back before the term “blogging” or even “weblogs” had been coined…).

  38. Raph you wrote:

    More seriously:

    As practical advice for presenting your work, perhaps you could acknowledge that iteration is a great way to develop many kinds of games and game subsystems. Then you can say that there are some problems where iteration is awfully expensive, and it’s *those* problems and game mechanics that you are trying to develop tools to tackle.

    Yeah, I guess I feel like I have said that before, but it can always stand to be reiterated.

    I’ll take your GDC2005 slides as an example of where it would have been nice to observe iteration as a fine design paradigm in some cases but not all. I just skimmed the slides again and didn’t see it in the first 10 or 15. So if it’s there, it’s too subtle. 🙂

    At any rate, I think this one post and the comment you made about games you’ve recently designed shed whole new light and context on most of this site’s content. Hyperlinks have a nasty way of landing readers right in the thick of things, but it might help to educate people if you had an “overview” page or “research statement” page that made all of this clear to readers.

    Maybe this sort of thing is redundant to people who know you personally or have worked with you, but I think it would be terribly useful for people like me who only know you through your writing, UO, and SWG.

  39. Oof, I am not sure what it would say. Hell, even my Wikipedia entry avoids any material like that (compare to say Lum’s, or Brad’s!).

  40. From Raph’s essay, What My Job Is….

    …Star Wars is a universe beloved by many. And I think many of you are like me. You want to be there. You want to feel what it is like. Even before we think about skill trees and about Jedi advancement, before we consider the stats on a weapon or the distance to Mos Eisley and where you have to go to pick up power converters-you want to just be there. Inhale the sharp air off the desert. Watch a few Jawas haggle over a droid. Feel the sun beat down on a body that isn’t your own, in a world that is strange to you. You don’t want to know about the stagecraft in those first few moments. You want to feel like you are offered a passport to a universe of limitless possibility.

    Call it an MMORPG, call it virtual world design, call it a graphical mud, I don’t care. My job is to try to capture that magic for you, so you have that experience. That’s my goal, the goal of this team, and we will do it to the best of our ability. I hope we’re well on the road to doing so….

    Excellent essay, Raph. Thanks for the heads-up, and for your patience with a nOOb who’s only just beginning to wander your website.

    Reading this over (and I had a hard time getting to it, cuz I kept slipping off into side issues, like the fascinating essays on player killers and what-not), I feel more comfortable with the mindset you appear to have with SWG. If your Big Goal really is to capture a sense of Wonder with this thing, then I’d say it’s only a matter of time before it all comes together. Granted, it may be a LOT of time, but still….

    I guess I’m just by nature frustrated with the inevitability of iterative processes in game development. I want a finished product, and I want it NOW, iterations be damned. Or, maybe I should cut myself some slack and just remember that the creative process itself is by it’s very nature a spawning bed for contention and frustration.

    So, what captures that sense of “Ooooh” for people? In a world where people pursue all sorts of divergent goals, how CAN a developer elicit a favorable reaction in everyone — the powergamer, the Killer, the Explorer, the Gentle Soul? And how in the world do you pull it all together in a single environment that doesn’t tear itself to shreds? Especially in a genre still in it’s infancy, relatively speaking….

    Heck, maybe you SHOULDN’T try to keep it neatly tied together, or “balanced,” or whatever. For my part, I played a Master Bounty Hunter for the better part of a year in SWG, prior to the Combat Upgrade. Yeah, I was a laughingstock among the hard-core PvP’ers, and I didn’t bother at all with hunting Jedi. But I had my fun. And I learned, and adapted, and eventually learned to excel within my limitations. I learned that as long as I didn’t stick my toon’s unprotected butt out onto the forefront of a battle, he was a killer. Even with all the gimping BH’s underwent over the course of that first year of the game, he was still a killer. But then, when the Combat Upgrade made Bounty Hunters one of the toughest characters in the game, all of a sudden EVERYBODY was a Bounty Hunter. And for me, that spoiled the fun of being a Master in a difficult profession.

    But then, I’ve always been a bit of an oddball. And I suspect games don’t sell well by marketing themselves to oddballs.

  41. So, what captures that sense of “Ooooh” for people? In a world where people pursue all sorts of divergent goals, how CAN a developer elicit a favorable reaction in everyone — the powergamer, the Killer, the Explorer, the Gentle Soul? And how in the world do you pull it all together in a single environment that doesn’t tear itself to shreds?

    By the way, I have over the course of my involvement with SWG tossed out a few ideas in the SWG Forums in an attempt to share some possible partial answers to the question I posed here. One thread I started in the old “Quests And Missions” forum seemed to generate some lively discussion. Look it up when the Forums’ Search function is working again:

    AUTHOR: Iakimo
    TITLE: I Want To Write Some In-Game Content!

    A professional developer from another game (a SWG fan, herself) jumped in and we had quite a bit of interesting back-and-forth on one of the ideas It’s a pre-Combat Upgrade thread, so some of the battle narrative is out of date, but they could probably be fleshed out nicely with appropriate tweaking.

  42. Raph: -I- never faulted you for making sandboxes. I like sandboxes. You should totally make more. Literal sandboxes. It would be So Cool to build a sand castle on Tatooine. But they are difficult to get started in.

    David? Finished products don’t exist in any medium. The initiator at some point stops working on it, whether for reasons of cost, time, or seeing no further actions to take. Then somebody else starts working on it, and gets far enough away from the source material to dodge claims of outright plagiarism.

    There’s a name for a world where everyone gets what they want. It’s called a utopia. Oddly enough, we seem to have trouble implementing them.

    -Wuh Fuh

  43. To be fair, you were not critiqued for “only making sandboxes” so much as not having enough directed content to properly frame the sandbox content. There were also critiques that the sandbox content in SWG was not fun, and that is probably the core problem. I think that SWG COULD have been great, and the sandbox elements could have shined (the crafting system in particular is quite good), but the creation of world environments with no stories to tell left a lot of players quickly bored.

    This is interesting, and I know you can’t talk about SWG yet, but I think that there are a lot of lessons about sandbox content that can be learned from the game.

    These elements had meta-challenges, but ones that were too easy to complete and did not build on one another. Giving players the ability to build on the world allowed them to build houses, or construct towns with guilds. They could even, if they chose to do so, build Rebel or Imperial bases and choose to defend them or attack oppossing bases. The problem comes out when you see that architects could enable a large group of players to acheive this goal without the work that the architect put into it. . . they simply paid the credits necessary to keep the architect able to mine ore (and thus advance his crafting ability), and then placed their home. Or they went on enough Rebel or Imperial randomly generated missions to purchase a base, and plopped it down and forgot about it (because the GCW materials were so bugged forever and because there was not a good method for knowing when to attack, and because PvP was so hopelessly imbalanced). Unless you were a crafter, the sandbox elements of the game were accomplished in under 15 minutes. They could be extended if you enjoyed home decoration, but with everyone having a home, most people worried only about decorating guild halls and used private homes as warehouses.

    If the game designers did a map of the player’s experience while in the game world, shouldn’t they have seen that the sandbox elements would have been exhausted in moments by every type of gamer but roleplayers?

    There are other examples. . . but mostly they appeal to roleplayers (entertaining/socializing in cantinas, raising your Imperial or Rebel “rank”).

    That’s why I give the critique that there was not enough directed content in the game to properly frame all of the sandbox content. In the end, the sandbox content really worked for crafters and roleplayers. That is fine and good, but that leaves a significant portion of players without anything “worth doing”.

  44. Vertical slicing an MMO is certainly possible. I’d suggest that no matter how complex an MMO, there will always be core gameplay elements that clearly say ‘this is what this game is about’. It’s going to be much more expensive to produce an MMO’s vertical slice than a single-player game’s vertical slice, of course. It’s also complicated by the fact that an MMO developer should probably also have to prove that their technology scales, relatively early in the project, given that such a high percentage of MMO developers fail to be able to do so within reasonable time and operational budget. (At least, if I were a publisher, I’d be highly dubious of a team’s ability to scale their technology unless they’d either successfully built a scalable and stable MMO before, or could prove that it scales.)

  45. To Ralph,

    I have been checking out this blog for a few months now. I have to give you credit, you expand my vision of game design at every turn. While I don’t think you are perfect, you certainly have my respect. This conversation about iteration, for instance, drove me to start my own blog.

    Its’s rough for sure, but perhaps you want to stop by and check it out. The article about re-iterating an MMO was one that I wrote about a month ago. Your gumption, if that’s a word, validated my points in the article so much that I decided to post it. I won’t give the address here, because this is, after all, your blog.

    I don’t want to sound like a fanboy, I’m far from it and have some very severe critisism of your design principles. However, like any true artist, I enjoy hearing what you have to say concerning game design.

    Don’t worry about the SWG detractors, you made a great game. Keep up the good work. However, some of those smaller games you’ve made…maybe pass them around a bit more so that we can see what you are thinking?

  46. What’s the URL to the blog? I can’t stop by unless you tell me where to go. 🙂

    And thanks for the kind comments…

  47. Amberyl says:

    Vertical slicing an MMO is certainly possible. I’d suggest that no matter how complex an MMO, there will always be core gameplay elements that clearly say ‘this is what this game is about’.

    My response to that is “maybe”. Perhaps a better term would be “mission statement”?

  48. some of those smaller games you’ve made…maybe pass them around a bit more so that we can see what you are thinking?

    Well, it’s not like they’re going to provide any major revelations. They’re mostly just puzzle games of various sorts. No theoretical advances, nothing even all that experimental.

    I can’t release them without permission from Sony, though — even though Sony is unlikely to ever do anything with them.

  49. In response to you not being able to release the games, due to sony…That’s too bad. I’m sure they are beautiful gems that should be recogonized by the public. My wife is a painter, and she has been bound by restrictions of these types many times. I will post my blog site soon, since you have given me permission to do so.

  50. Pff, they’re not beautiful gems — don’t overestimate me. A few of them are fun, but that’s as far as I will go. 🙂

    Also, it’s not that Sony is holding them back or anything — I just haven’t asked them. By contract, Sony owns all my game output, is all. If they decide they don’t want it, I can do whatever I want with it, but I have to get permission in writing and I just haven’t bothered.

    Also, you don’t need permission to post a link — just go for it. Trust me, if there’s a link in a comment that I consider spam, I zap it — I do that a dozen times a day.

  51. okies, it’s at: http://dsutton.blogster.com/thoughts_mmo_paradigm.html

    I’m not sure if is a good idea or a pile o’ stink…but its interesting. 🙂

  52. Well, the SWG Forums’ search tool is working again, so I thought I’d toss in the link to the “Idea Thread” I alluded to earlier here, FWIW.

    When we’re talking about iterations in a MMORPG, it’s this kind of thing that I think most people would normally expect. Not a major revamp of the core combat model… or, egad, TWO such revamps….

  53. DSutton, what you’re suggesting has been done a lot, actually; a lot of the Kesmai games were that sort of session-based game, resolving after a set period and then restarting. I believe Air Warrior ran week-long campaigns for example. And even the original take on MUD1 was a “groundhog day” sort of environment, I believe.

  54. Does game iteration really matter in the first place? It does in the sense that we want to get things right, but in the overall scheme, games, especially MMO’s, are always in a state of beta. We think that games are art, right? This translates to other mediums as well.

    I thought about this while watching my wife paint tonight.

    As a painter, she feels that her paintings are never finished. It’s simply part of her philosophy. To my frustration, she will continually work on a painting for the duration that it is in her possession. When she does more to a painting, however, it becomes unfinished once again. When she sells a painting, she can’t work on it anymore, so the piece is essentially finished at that point. However, she was never truly satisfied with the end product and thought that she could do more to “finish” it before it was released.

    Games, I think, are the same way. We’ve been adding rules to chess for thousands of years. Does the introduction of en passant change the nature of chess, or merely add to it’s value as a game? Is a game ever finished? I dunno? We finish one part of the design doc and there is another section added. Game’s are always unfinished, and that’s not a bad thing, always.

    Do us youngsters play Monopolpy the way that are grandparents did?

  55. DSutton, what you’re suggesting has been done a lot, actually; a lot of the Kesmai games were that sort of session-based game, resolving after a set period and then restarting. I believe Air Warrior ran week-long campaigns for example. And even the original take on MUD1 was a “groundhog day” sort of environment, I believe.

    Thanks for the insight! I appriciate you coming by to take a look…any thoughts on how those projects turned out?

  56. In a game that’s essentially a psuedo-persistent strategy game, with frequent tournament ladder resets, your end result isn’t much different from a lobby-based competitive multiplayer game.

    In a world-like environment, frequent restarts favor the hardcore, in general; more casual players tend not to get to see the extent of the available content, and usually carry over less in the way of rewards. Usually, in something like this, you’ll see a persistent community-oriented area, coupled with a resetting strategic environment of some sort. (Carryover bonuses tend to affect the persistent world in some way, more so than affecting the resetting environment, ensuring that the latter remains a relatively level playing field for newcomers.)

  57. The main problem with iteration in games is the trick of keeping the player ‘willingly’ participate in them.

    This is why wow is a huge success. Wow is a game of endless iteration. If you want to reach Rank 14 in your faction, you need to do 3 Battlegrounds repeatedly. But wow sets the targets and awards so good, players ‘blindly’ do everything again and again. You reach rank 14, but it’s not enough, you need to win more battlegrounds to be Exalted with one clan to unlock more weapons, receipts. You ‘obey’ your instincts and keep doing them all to unlock all.

    Same for instances, all players do the two end-game instances; Black Wing Lair and Molten Core repeatedly for months. Because wow has perfectly set you up. Every player has 9 slots for armor and 2 for weapons, and these instances are 40 player instances. 440 slots to fill. Players don’t even realize they are not having fun anymore in the game, because after a certain time, experienced groups work like machines and drop bosses one by one.

    And just when players complete their ‘epic’ sets, wow releases another instance.

    I don’t understand why other companies can’t see the simple pattern here. Players will blindly be hypnotized by iteration if appropriate awards are given.

  58. I think that’s a different sort of iteration, Oropher. That’s the players iterating over the content. I do still have a post called “Do treadmills suck?” half-written that I need to finish. 🙂

  59. Raph wrote:

    I think that’s a different sort of iteration, Oropher. That’s the players iterating over the content.

    Heh… chasing the old carrot on a stick? As in striving to be uber, but when they reach “uber” status, they find the target’s been moved?

    Idunno… I haven’t played WoW yet, so I can’t say definitively how well they’re keeping their player base fixated on the carrot, but it would seem they’re doing their iterative work pretty darn well.

  60. *dryly* It’s Blizzard. They take “iteration” to these incredible heights that most people still don’t copy. =P

  61. I’d be interested in seeing whether treadmills suck according to a shining genius, so please do finish that. 🙂

    WoW and Diablo both hit the one-arm-bandit mechanic very, very well in this regard – just enough incentive to keep feeding those coins in.

    While iterating gameplay mechanic and in turn providing the player with repetitive play will get you a game, it won’t typically get you an innovative game. Of course, in this marketplace innovative isn’t nearly as nice as market-successful – which Blizzard have very aptly proven.

    I think the vertical slice model falls on its face in the development of non-MMOs more times than not precisely because of the amount of work that can be required to prove the mechanic and the team. While the results are often similar to the scriptwriter rejecting, the cost of doing so is markedly different. For this reason I’m always a fan of – where possible – prototyping using other media or demonstrating the minimum amount required for the gameplay to work. The trick is proving this to the publisher…

    In MMOs the big hurdle here is your game mechanic needs a large number of players to be proven, and by the time you start inviting your large numbers of real people you’re a lot further into your development than prototyping…

    I’m realising I’m parroting a lot of what Raph said in his original blog entry here. As he inferred, the issue is that people don’t appreciate that right now the only realistic way to prototype a MMO mechanic is theoretically. Which is probably easier if you are a shining genius, but is still prone to failure.

    I would hate to see iteration of mechanic become the driving force in successful games because of lack of risk-taking. The theory is valid and useful – but the bad press is pretty much inevitable. We attach so much more to action than to words, and too often theory is seen as just words.

    Academics/geeks hold an interesting social position as both admired and pitied probably for this very reason.

    Okay, stopping rambling and returning to lurk-mode (and yes, I’m a newcomer here too).

  62. Heh… CBS News Online put up a Mail Bag thread on their website about SWG and its iterations….

    Clicky.

  63. […] Hopefully this provides a bit of insight into why developers fuss about topics like “What is a game?” There are some important issues we need to consider, and sometimes the only way to really get to the root of the answer is to start thinking about things in a more “academic” way. Our industry has run on gut feeling and old wives’ tales for many years. It’s time for us to start really thinking about game development in more serious terms. This is one reason why Raph shouldn’t be afraid to be known for his theories. 🙂 […]

Sorry, the comment form is closed at this time.