A Comprehensive Game Design Methodology
From First Ideas to Spectacular Pitches and Proposals

The content of this website is licensed under a Creative Commons CC BY-NC-SA 4.0 Attribution–NonCommercial– ShareAlike 4.0 International License. You can freely use and share everything non-commercially and under the same license as long as you attribute it to and link to:

J. Martin | |

However, you can also buy the Ludotronics PDF edition
for an unreasonably moderate price at DriveThruRPG.
Learn here about the five excellent reasons to do so!

Why DriveThruRPG? It’s the largest tabletop RPG download store and you’ll probably end up buying much more than just your copy of Ludotronics. Which would benefit all game designers!

Why not Amazon? Ludotronics isn’t well-suited for the Kindle format. And at €14.99, Amazon’s cut amounts to €9.75. Well, no.

More to read: My papers at Research Gate, my blogs at between drafts and just drafts.

Level Five: Architectonics

Process Phase Level Five

Beat 3. Paradigms Lost

Mottos & Mantras

The nature of tasks, from typology to difficulty, and the Ludotronics task model based on it are discussed in-depth in Level Two: Interactivity. If you beat that level already, you know all about it.

If not, don’t worry. All you need for this level is a rough understanding of the model’s three fundamental types of tasks: physical tasks, cognitive tasks, and empathic tasks. Roughly, they correspond to action, puzzles, and interactive storytelling, respectively. In this and the following beats, we will primarily focus on empathic tasks.

Fig.4.25 Task Model
Fig.4.25 Task Model

To create a great game, you need to translate all its dramatic elements from the story development arc, the character development arc, and the player development arc into tasks that provide the player with the just-right amount of challenge and, which is even more important for this territory, meaningful choices. Translating dramatic elements into physical or cognitive tasks that advance plot, character development, and player proficiency (including skill, knowledge, understanding, and attitude) are the routine take-off and flight procedures of successful game design. The hard part are empathic tasks, which we find first and foremost in dramatically complete games. Translating dramatic elements from the story and character development arcs into empathic tasks is the crosswind landing of successful game design.

To approach this topic, we first need to examine, and eventually retire, two design paradigms with regard to interactive story mode and interactive storytelling, respectively: “show don’t tell” and “balancing story and gameplay.” Thus, let’s put both our three development arcs and our task model on the back-burner for a while. We travel light—let’s hunt some paradigms!

First, interactive story mode. Stories cannot be separated from the medium through which they are told. For stories that are told through words, either orally or in written form, everything’s in the telling. Dramatic dialogue and dramatic description, respectively, create emotions in the audience’s minds. For stories that are told through a visual medium, the telling paradigm doesn’t cut it. Here, it is all about visually compelling dramatic action that is shown to create emotions in the audience’s minds. Thus, for this medium, the Show, Don’t Tell paradigm is a great fit. (It is often claimed that the Show, Don’t Tell paradigm also applies to written media. But that’s a dubious proposition because it entraps fiction writers in a cinematographic metaphor. You sing, not show, the anger of Achilles.) For a story that is told in an interactive medium, finally, Show, Don’t Tell needs to make room for a new paradigm, just as telling had to make room for Show, Don’t Tell for the visual medium. Neither scripted NPC gamesplainers (telling) nor full motion video snorefests (showing) are adequate modes to advance the story, character, and player development arcs in interactive games.

Instead, everything should be designed to push the player through playable action toward plot resolution, character fulfillment, and the goal of the game. Therefore, our paradigm for games and similar interactive media will not be Show, Don’t Tell, but Push, Don’t Show.

Now, the argument can be made that the game shouldn’t “push,” but “pull” the player into the game like a reader is “pulled” into a story, as brought forward by Neal and Jana Hallford in Swords & Circuitry. Playable action and goals should be so enticing that the player cannot but be pulled toward it, without any need of being pushed. This would be convincing if it weren’t for the fact that games don’t work like stories. A story pulls the reader along by keeping them curious about what’s going to happen next, and this curiosity depends in no small part on the fact that the story can’t possibly be influenced or changed by the reader in any way. Accordingly, being pulled into a story means, at bottom, that you can only follow the story and, under favorable circumstances, cannot but follow the story. This is different from how gameplay works.

Consider this. What could it mean that a beat has to “push” the player toward an action, and finally toward the goal? A beat can, for example: provoke a reaction; trigger a movement; force a maneuver; demand a payment; present a puzzle; request an item; ask for a favor; call for help; give an order; require a decision; pose a problem; necessitate a response. Which doesn’t mean that there’s no pulling. A beat can also, for example: provide information; make a suggestion; create tension; arouse curiosity; raise expectations; build anticipation; press to reconsider or rethink; present a surprise. Yet even then, the player isn’t supposed to simply follow these pull elements, but to act on them, just like they would on push elements. Hence, Push, Don’t Show, not Pull, Don’t Show.

Is this paradigm valid for any game, always, and everywhere? Emphatically, no. Game elements that show, tell, and pull all have their place. One outstanding example for more tell- or show-oriented games are interactive visual novels and sound novels. Another outstanding example are interactive dramas like Heavy Rain or Fahrenheit (a.k.a. Indigo Prophecy) from Quantic Dream, or the complex cinematic narratives from Konami’s games from the Metal Gear series. They all develop their show, tell, and pull elements into impressive art forms to tell stories in innovative and unexpected ways.

But in games that are supposed to push the player toward action, tell and show elements—particularly non-interactive cinematics like cutscenes—are used to “solve” design issues for empathic tasks way too often.

Fig.4.40 Push, Don’t Show (Paradigm)
Fig.4.40 Push, Don’t Show (Paradigm)

Next, interactive storytelling. The corpus of research published on this topic has produced a broad range of insights, but for a number of years now, this insight curve has flattened out. The reasons, probably, are twofold. On the one hand, the most substantial observations have already been made and expressed by now. On the other hand, interactive storytelling probably reached its productivity limit because of its underlying paradigm of balancing story and gameplay.

To understand on a deeper level why we need to abandon this paradigm of balancing story and gameplay, we need to know a little bit about the historic development of dramatic structure and its implications.

In the beginning, dramatic structure was narrative structure. From Aristotle’s Poetics to Gustav Freytag’s Pyramid to Syd Fields Paradigm, philosophers and artists have tried to analyze, formulate, and refine the dramatic structure of non-interactive media to achieve the highest possible emotional impact on their audiences. From humble beginnings like exposition, climax, and dénouement with rising and falling action in between up to inciting incident, showdown, and resolution schemes peppered with intricate, precision-timed networks of plot points and pinches, dramatic structure has been used by artists to engage their audiences with relentlessly controlled experiences that enthrall and absorb. Dramatic structure became nothing short of a high-precision instrument without tolerance for tolerances—in mechanical parlance, there’s almost no room for “backlash” or “play.” Now, adding player agency to that system is like throwing Charlie Chaplin’s Modern Times character armed with his two wrenches into the works, and in games and comparable interactive media the whole dramatic machinery rumbles and crumbles and breaks apart.

Which is another way of saying that for games and interactive media, dramatic structure no longer equals narrative structure.

In “Beyond Myth and Metaphor: The Case of Narrative in Digital Media,” Marie-Laure Ryan discusses what is and isn’t desirable for emotional “first-person” experiences in games. One of her examples, Tolstoy’s Anna Karenina, reveals two momentous aspects. In contrast to non-interactive media, the dramatic structure can’t throw all its plot points of pain and misery and suffering on the player/player character without running into problems. Then, plot points that are essential for the dramatic structure to work can no longer be cast as player/player character choices and decisions.

Let’s look at this more in-depth. On a lighter note, it has been quipped with reference to Marie-Laure Ryan’s essay that not many players would be willing to kill themselves to make the game more interesting. But one shouldn’t underestimate games. NieR’s third playthrough, for example, indeed presents the player with such a choice; and while it’s “just” an alternative ending, it erases the player character’s in-game existence along with the player’s save files. Then again, NieR is certainly an edge case. In most other cases, and that is the important point, the only way to guarantee that an indispensable dramatic event materializes is to take the associated dramatic decisions away from the player/player character and rob them of their agency—which, together with autonomy, is one of the four motivational drivers from our player motivation model.

That’s not a good sign, but it gets worse. For any given dramatic structure to work, the player can’t even be granted unrestricted agency with respect to conflict. Some players will try to defuse the conflict. Others will try to over-escalate the conflict. Barring an unlikely army of Goldilocks players, the amount of possible player choices has to be reduced considerably just to maintain an interesting conflict. (There are non-Western storytelling traditions that are purportedly “conflict-less,” but, at closer examination, such interpretations almost always turn out to be wishful thinking with or without a dash of benevolent exoticism.)

For all these reasons, it has become accepted that games can’t have a screenplay-like roller coaster ride and real player agency in dramatically complete games—because dramatically complete games, by definition, include not just a loose assemblage of story elements, but plot. Game designers have navigated this narrow channel of “balancing story and gameplay” for ages. And while this paradigm has carried us forward for a long time, it doesn’t carry us far enough.

Let’s briefly recapitulate some of the most common design strategies that have been put to use: some that don’t work so well, some that work a little better; and one that works really well as it abandons the paradigm.

By far the worst but also most popular strategy to negotiate this issue is to split the game into player action for the agency department on the one hand, and use cutscenes to convey anything and everything that presents even a whiff of plot or story on the other. Neither is this helpful, nor does it measure up to the potential of cutscenes for innovative storytelling as touched upon above. Another popular strategy is the illusion of decision-driven forking, with complex machineries of smoke and mirrors in the background to prevent the exponential growth of story lines that even a small number of true successive decision points beget. A third strategy to provide both plot points and player agency are serious attempts at designing procedural storytelling machines and plot point–generating algorithms—all of which didn’t exactly blossom, as especially the intrepid and indefatigable Chris Crawford can attest to.

Two strategies that work much better are soft-gating and combining linear with non-linear gameplay. Soft-gating has been employed in open world–games with miscellaneous techniques, from Gothic to S.T.A.L.K.E.R.: Shadow of Chernobyl to almost every modern action-adventure game and even racing games. Here, players can choose where to go and what to do—but, depending on their overall progress in the game and their corresponding skill and equipment levels, not every choice is a healthy one and life can terminate quickly. Games that combine linear with non-linear gameplay create an effectively linear course with a starting point, a handful of major plot points, and an endpoint that are all fixed, but with non-linear plot points and freedom of movement in between. (Some of the Ludotronics phases represent a smallish version of that strategy.) A close neighbor are games with a fixed storyline and plot points, but an abundant supply of side quests that can be pursued or ignored in any order.

What works best so far extends the linear/non-linear combination strategy in clever ways through intricate under-the-hood computing of player decisions, as it’s done in the games from the Mass Effect trilogy and especially in The Witcher 3: Wild Hunt and the forthcoming Cyberpunk 2077. These games give players an immense range of choices, especially the latter, that have consequences in the game world without forking into a googolplex of different story lines. In The Witcher 3: Wild Hunt, and most likely in Cyberpunk 2077 as well, a fact-tracking database under the hood tracks player decisions and its consequences, particularly as to how successive events will play out. (We will discuss the difference between what and how for player decisions and their game world ramifications soon.)

As you can see, even the strategies that work amount to bricolage and clever tinkering, and enormous amounts of handcrafted design work. But it’s bricolage, tinkering, and handcrafted design that pushes us ahead right now, and it already left the old “balance” paradigm behind, as noticed.

And that’s what we need to do for good, abandon the design paradigm of balancing story and gameplay cognitively and technologically. In its place, we suggest a different design paradigm altogether, which we already introduced in Beat 1. Partitioning and put on the back-burner at the beginning of this beat. Instead of balancing story and gameplay, our design paradigm for dramatically complete games consists of integrating story, character, and player development.

At first glance, this seems even more difficult because there’s more to negotiate than ever. Suddenly, there are not two, but three parties sitting at the table! But the important point is that it’s no longer about compromises. Consider this. Balancing story and gameplay is a paradigm focused on trade-offs, on making concessions, on finding compromises. In other words, it’s what the Harvard Program on Negotiation calls “positional bargaining” in a “fixed pie” situation. Both sides take their positions, argue for it, and make concessions to reach a compromise, i.e., give up something they want less in order to get something they want more. But for an integrated dramatic structure for games, something altogether different is needed: a design perspective that natively includes player development and character development and player decisions, in the sense of “native to the medium.” To negotiate all three parties, we need to “broaden the options,” as the Harvard model calls it, and realize that the number of shared and compatible interests between them is substantially larger than the number of opposed interests on which the paradigm of balancing story and gameplay focuses on.

Thus, we also need a different terminology. Balancing story and gameplay served the design paradigm of “interactive storytelling.” Integrating the three development arcs will serve our brand-new design paradigm of—ta-da!—interactive narrative!

Now, you probably expected something more spectacular, and indeed, there were some interesting terminological candidates. But for a good number of excellent reasons, we call it “interactive narrative” and not something else. First, the term is familiar and not exotic. It has been used by others who pushed into similar directions, from Mark Stephen Meadows’s Pause and Effect and Marie-Laure Ryan’s “Interactive Narrative, Plot Types, and Interpersonal Relations” to Grant Tavinor’s The Art of Videogames and Tynan Sylvester’s Designing Games. What’s more, the term is used by CD Projekt Red designers, e.g., by Miles Tost and Nikolas Kolm in their talk “The Witcher 3: Crafting a Compelling Narrative in a Believable Open World,” which is probably the best association you can get in this department at this point in time. (Historically, the term was also used, and continues to be used, for AI approaches to algorithmic storytelling; recent examples are Brian O’Neill and Mark Riedl’s “Emotion-Driven Narrative Generation” or Mark O. Riedl and Vadim Bulitko’s “Interactive Narrative: An Intelligent Systems Approach.”) Then, “interactive narrative” is a perfect sibling for “world narrative,” Tynan Sylvester’s term adopted for the Ludotronics paradigm to replace “environmental storytelling,” which is discussed in Level Four: Narrativity.

Finally, “story” and “narrative” have different characteristics that really make a difference, also discussed in Level Four: Narrativity. “Story” is primarily about what is told, the content; all that’s not part of a story belongs to different stories; and stories are robustly impervious to audience participation. “Narrative,” in contrast, is primarily about how something is told, the structure; it is open-ended and can be complemented with new elements; it can accommodate not just one, but many stories; and, above all, it can accommodate any number of player narratives.

Fig.4.41 Interactive Narrative
Fig.4.41 Interactive Narrative

Now that we have established interactive narrative and its underlying design paradigm of integrating the three development arcs for the story, the character, and the player, we can come back to our tools, also left on the back-burner at the beginning of this beat: the constituents of the Ludotronics task model, especially empathic tasks.

Empathic tasks, according to our model, are primarily associated with understanding and attitude. Both, in turn, have to do with learning. To accomplish an empathic task shouldn’t simply boil down to making a decision, but to making the right decision. Thus, the first thing we need to do under this design perspective is to treat empathic tasks the way we treat physical tasks and cognitive tasks: the player can fail because their action or decision was either wrong or not good enough.

Here’s an example. In a game, the player/player character tries to cross a superhighway on foot instead of walking up to the nearest footbridge because they’re pressed for time. That doesn’t work out because at this point, the player isn’t fast enough to dodge the approaching cars (or needs to earn a speed upgrade first). What happens next is that the player character winds up under a car. Rare is the game that forks the story every time the player messes up like that, has the player character picked up by an ambulance, driven to the hospital, taken care of, and then arrested for public endangerment. Now, in the same game, the player has to find a way to break into a closed supermarket without attracting attention. The player/player character kicks in the door, an alarm goes off, a dog begins to bark, and the siren of a nearby police cruiser approaches fast. Again, rare is the game that forks the story every time the player messes up like that, has the player character put into custody and scheduled for a court hearing. Then, in the same game, the player character is approached and questioned by the police. Here, suddenly, we have the misconception that we have to offer the player a set of choices each of which needs to play out in different ways with different consequences that require at least some moderate forking—instead of treating this task exactly the way we treated the physical and the cognitive task. (There are games that try to do it the other way around, to treat failure from physical and cognitive tasks like empathic tasks and accommodate all non-fatal player failure. But these are rare and usually quite idiosyncratic—as, for example, Beyond: Two Souls to a large extend.)

For every empathic task, there should be a small range of right decisions and a greater range of wrong decisions, and the player can succeed through understanding the situation or modifying their attitude by correctly judging what the non-player characters in question are up to. Just as with physical and cognitive tasks, not every outcome of an empathic task needs to be accommodated. And just like physical and cognitive tasks, empathic tasks should become more challenging over time, across the whole range of cognitive empathy, emotional empathy, and compassionate empathy.

What’s more, physical tasks and cognitive tasks are often designed to be solvable in different ways. Equally, empathic tasks can be designed to be solvable in different ways. For a physical task, the footbridge takes more time, but is less dangerous. For a cognitive task, trying to enter the supermarket by cutting a hole through the roof might be as feasible as knocking out the night watch on duty and take the keys, but both courses of action need different skills and different temperaments. For an empathic task, when the character is questioned by the police, they can get along with the officers by acting polite or by acting tired or maybe even snappy, which makes for different experiences in the situation (and might even make life for the player easier or harder somewhere down the road), but there are options like acting blustering or aggressive or dumb or drunk which have the character promptly removed into the can, and that was it. All this leads us to our first rule of empathic tasks.

  • The first rule of empathic tasks: decisions can be right or wrong, there should be more than one right decision, and there should be more wrong decisions than there are right decisions. And the game should provide exactly as many right decisions as the game can handle without forking into a code version of Yggdrasil.

(As an aside and an outlook, our examples so far are all based on the premise that failure in games is handled in traditional fail–reload–repeat fashion, some aspects of which are discussed in Level Two: Interactivity. How failure can be handled in different ways will be discussed in Level Six: Integral Perspectives II.)

Our first rule of empathic tasks seems to leave us with one big, hairy problem. How do we get the player to make the effort to understand a situation or change their attitude if they can simply proceed by trial and error?

Physical tasks rarely have that problem. If the player isn’t good enough to perform a task, they have to try again and again until they’re good enough. Cognitive tasks are more susceptible to that kind of problem. But it’s quite conventional, even acceptable, to solve cognitive tasks through trial and error, even though we’re supposed to use memory, analysis, and evaluation to find a solution. (Also, patience is in short supply, and second screen solutions always seductively close by.) For empathic tasks, the solution to this problem isn’t obvious, but already embedded in our first rule of empathic tasks: a small range of right solutions, at least two, and a broader range of wrong solutions.

First off, it isn’t particularly enjoyable to play a scene over and over, grinding through a number of wrong solutions just because you’re too lazy to pause and think about it for a moment. Then, if there are at least two right solutions that each have their advantages and their drawbacks, it’s much more enjoyable to try and make a decision that not only provides a solution, but the solution that has the most advantages and the least drawbacks attached to it. This can only be accomplished by gaining an understanding of the situation. On top of that, solving an empathic task without trial and error becomes even more interesting if these advantages and drawbacks are player-specific, connecting with wants and needs that vary from player to player in the context of their respective player narrative. That way, the player has a whole set of incentives to eschew trial and error, especially when one of the solutions can give them exactly what they want. (That there will always be players who try to “maximize” the outcome shouldn’t bother us at all—it doesn’t bother us with physical or cognitive tasks; there are no indications that maximizers don’t enjoy maximizing; and it can even boost the replay value of a game.)

In the example above, dealing successfully with the police officers by being polite and helpful can give the player character some advantage in dealing with the police later in the game. Dealing successfully with the police officers by being tough and snappy but with the just-right dose of respect can give the player character some street cred instead. Or, staying firm and knowing their rights might also be successful, but actually not endear the player character to anyone. That brings us to our second rule of empathic tasks.

  • The second rule of empathic tasks: the respective outcomes of right decisions should be different enough to be perceived by players as being more or less advantageous, depending on each player’s individual player narrative.

Let’s proceed and include our emotional landscape for dramatically complete games that we developed in Level One: Integral Perspectives I. Empathic tasks become much more interesting if they are attached to emotions, from both the fear class and the pity class. (The fiero class will kick in after the player has solved or not solved the task, depending on how well they succeeded or how badly they messed up.) Then, every solution of an empathic task, but especially every right solution, should have its own set of emotions attached to it. In our example of the police interview, both classes are involved, and different emotions are attached to different decisions.

Let’s start with the pity class as emotions from non-player characters the player can empathize with. One right solution could leave the officers happy, another could leave them annoyed, and for a third they might feel contempt. The palette of wrong solutions could make them angry or frustrated or even resentful. (To develop an idea about these emotions and what might trigger them is indeed part of the empathic task the player has to solve.) Now the fear class, for the dramatic situation the player can see themselves in. Again, different solutions would trigger different emotions. The edgy solution would trigger fear or at least some anxiety; the polite solution possibly joy or delight to have gained new friends; the third could simply come down to resolve and relief. (Not to forget, the emotional state of the player at the beginning of that situation should not be a matter of chance. In Level Four: Narrativity, design techniques like agenda-setting or priming are discussed that are cut out for this.) Now we can define our third rule of empathic tasks.

  • The third rule of empathic tasks: every decision on offer should have its own emotions from the pity class and the fear class attached to it, as well as follow-up emotions from the fiero class.

Finally, how do we deal with an empathic task that is set up as a plot point or a journey stage where something very specific has to happen, otherwise it would shred the story development arc or the character development arc to pieces, respectively?

If you think about it, the three rules of empathic tasks that we’ve defined so far are already quite sufficient to deal with it most of the time. Even the imperative that something has to happen rarely proscribes that it has to happen in exactly one manner. But okay, let’s one-up that. You have a dramatic solution in mind that has to happen to a T, exactly as designed or written, to get the most dramatic and emotional mileage out of your plot point or journey stage. Well, we have to amend the first rule then: in situations of paramount dramatic importance, there is only one right solution, and all other solutions end in failure. But designing such one-solution tasks is hard and exacting—the “right” solution must be even more convincing and even better prepared than any “regular” solution to an empathic task! Both the right solution and all the wrong solutions, in this case, must be impeccable in terms of plausibility and inevitability. Instead of an amendment to our first rule, through, let’s rather make it our extra rule of empathic tasks.

  • The extra rule of empathic tasks: a truly climactic choice is allowed to have one, and only one, right solution that then must excel in plausibility and inevitability.
Fig.4.42 Rule Set for Empathic Tasks
Fig.4.42 Rule Set for Empathic Tasks

To flesh out this new perspective with interactive narrative as its underlying design paradigm and empathic tasks as one of its tools, and to dig a little deeper into its potential application for different types of games, we need to take a closer look at the nature of decisions, which we will do in the following beat.