Ludotronics

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 | betweendrafts.com | ludotronics.net

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 download store for role-playing stuff in existence and you’ll probably end up buying much more than just your copy of Ludotronics. Which would benefit game designers everywhere!

Why not Amazon? For one, illustrated non-fiction isn’t well-suited for the Kindle format. Also, at a €14.99 price point, Amazon’s cut amounts to €9.75. Well, no.

Level Two: Interactivity

Process Phase Level Two

Beat 4. Representations

From Rules to Realism

Now that we discussed what rules and game mechanics are, how interacting subsystems create individual and non-exhaustive playing experiences, and how compelling conflicts can be designed around mathematical models, let’s look at the representational aspects of video games associated with the Interactivity territory.

Each territory has its own context of perceived realism, from “realistic” design and graphics in the Plurimediality territory and “psychological realism” in the Narrativity territory to “realistic” stories in the Architectonics territory, each of which demand a range of design decisions that sharply impact the playing experience. The same is true for the Interactivity territory. Here, “representation” refers to the realism of the game world, as perceived by the player, in relation to game events on the one hand and player input on the other.

This beat’s first part is about the perceived realism of events in the game world. For that, we will discuss both discrete in-game player actions and accumulating game world events, the latter as a scenic walk along the mathematical probabilities of pen & paper role-playing rule systems.

The second part about the perceived realism of player input will lead us first to car racing games as a typical example for the contentious issue of input realism, and then to the equally contentious issue of quick time events as an intermediary between highly abstracted player input and hyperrealistic game world–output.

Thus, let’s turn to our first topic, in-game events. The rules of all the games we looked at so far fall into either of two categories: those that are completely abstract, like the rules of chess, and those that simulate aspects of physical reality, like the rules of Shroom!, among them running, jumping, shooting, or picking up an item from the ground.

For design decisions in the Interactivity territory with regard to realism, we can put abstract rules to the side, and with them games like chess, Skat, the game of Go, Tetris, Qix, match 3 games, and so on. That should leave us with rules that map actions from the real world into the game world.

For single actions, transitions are always vague and highly subjective, but there will be a point at which most players agree that, e.g., a human avatar’s allowed jump height or jump distance has stopped being realistic. Of course, our assessment itself of what’s realistic or not is often anything but realistic. A human avatar might be able to scale more than its own height in a standing vertical jump, which is pure fantasy, but the player might still “feel” that a trained person could accomplish this. Somewhere between that and three times the avatar’s height, though, that feeling will certainly have evaporated.

Around such contexts, design decisions need to be made. Should the rules for a given game be realistic (capping vertical standing jumps at around 40 inches, or 100 centimeters), feel realistic (maybe allowing up to 60 inches, or 150 centimeters), or should realism be just thrown overboard (because scaling enormous obstacles with vertical standing jumps is a blast)? Obviously, your design decision should depend on your type of game, its target audience, its value characteristics, and its theme. Your decisions will be different if you’re designing a training simulation, an arena shooter, or a beat ’em up. They will be different if you’re designing a contemporary detective mystery, a high-adventure survival thriller, or a superhero brawler. But whatever you’re working on, decisions need to be made. That is also true if your level of realism doesn’t correspond to your game type or story type on purpose.

So far, so good. But what about the perceived realism not of discrete actions, but how events are handled in general—i.e., the perceived realism of when, how, and how often events occur and how they play out over the course of the game?

To illustrate the seriousness of this question and its impact on the interactive playing experience, let’s have a very detailed look at the mathematical probabilities of pen & paper role-playing rule systems, a traditional and familiar topic to dedicated P&P designers and aficionados alike.

Depending on the type and number of dice you’re using, playing experiences in pen & paper role-playing games differ vastly with regard to skill use and difficulty levels, brought about by different probability distributions. If players resolve tasks by rolling a D100, a D20, a D10, or a D6, the probability distribution is uniform for every die roll. That’s because every value has an equal chance of turning up. For a D100, each value has a 1% chance, for a D20, each value has a 5% chance, and so on. (In science, probabilities have values between 0 and 1, whereby an event with probability 1 always occurs, and an event with probability 0 never occurs. This has many advantages, but we’ll nevertheless stick with the percentage values most players are familiar with.)

If players roll any combination of dice, in contrast, the probability distribution is no longer uniform. Different values no longer have an equal chance of turning up.

Compare a D20 system against a 3D6 system, for example. That the probability distributions are markedly different is obvious. But that the respective interactive playing experiences will also be markedly different, that’s the important point.

Fig.4.19 Probability Distributions D20/3D6
Fig.4.19 Probability Distributions D20/3D6

To see how different playing experiences are constructed by different probability distributions, let’s drill down to the details. The principle behind such die rolls in general is that a player with an average skill level (in our example 10 in both systems, for simplicity’s sake) has an average chance (50%) to succeed at a task with an average difficulty level (again roughly 10 in both systems). That might or might not correspond to how the world works, but it’s a long-established convention where “average” difficulty is not equivalent to routine tasks.(Learning curves are typically not that steep from novice to average. But modeling player level and skill upgrade paths is a topic that, while related, warrants its own individual treatise.)

In systems where players must roll less than their target number, the player will succeed by rolling a 10 or less in both systems. That way, a character with average lock picking skills will successfully pick average locks about half of the time. But as soon as you leave that isolated sweet spot around these averages behind, things drift apart. In a D20 system, a one-point increase or decrease in skill level or difficulty level corresponds to a five-percent increase or decrease of the chance to succeed, proportionally at every point. In a 3D6 system, in contrast, the chance to succeed is disproportionately affected by one-point increases or decreases in skill level or difficulty level.

Here’s a simplified yet fairly typical example where the range of possible skill levels maxes out in the neighborhood of 20. If your character has a better-than-average skill level of 12 in the D20 system, i.e., 10% better than average, their chances to succeed at an average task will also be 10% better than average, i.e., 60%. Now, if you raise a character’s skill level in the 3D6 system by 10%, worth about 1.8 or 2 points, from 10 to 12, their chances to succeed at an average task are raised by a whooping 24%—that’s a 74% chance to succeed! And vice versa. At lower skill levels or with higher difficulty levels, the player’s chances to succeed at a task dwindle much more rapidly in the 3D6 system than in the D20 system. The same is true for an increase or decrease through modifiers. A modifier boosts or reduces skill levels (e.g., having studied and trained with that type of lock before versus being intoxicated) or raises or lowers the task’s difficulty (an inbuilt alarm system versus using a premium set of lock picking tools). Like skill levels and difficulty levels, modifiers affect chances proportionally in a D20 system and disproportionately in a 3D6 system.

What’s more, this also greatly affects competitive tasks between players or between players and NPCs. If two competing players have the same average skill level, each will succeed in a competitive task on average 50% of the time over time. If one player has a skill level two points above that (or a +2 modifier, which amounts to the same thing) in the D20 system, this player will gain an additional 10% advantage and succeed on average 60% of the time over time—a noticeable improvement. In a 3D6 system, though, the same player will gain 24% and will crush the other player by winning three out of four matches on average over time.

But wait! The probabilities of the improbable are also different. The most improbable event in a D20 system, a good one or a bad one, occurs at a roll of 2 or 20, respectively; with a 5% chance for each. Which, if you think about it, is immense. In a 3D6 system, the most improbable event, again a good one or a bad one, occurs at a roll of 3 or 18, respectively; each has a mere 0.46% chance to occur. That might just suffice to advance from morning coffee to night cap without getting yourself killed.

All that goes on under the hood. Now let’s see how it affects the playing experience in terms of perceived realism, so you can make meaningful design decisions. Over time, all other things being equal, a 3D6 system will have much more “expected” results than a D20 system because there’s a better chance to roll reasonably average results for average performances between, say, 7 and 13 for a combined 67,58% chance. Contrast this with the D20 system, where the corresponding “middle ground” only yields a 35% chance! That’s a dramatic difference. Over time and all other things being equal, a D20 system will produce many more “improbable” results through deviating from expected averages with surprising in-game effects, because the probabilities for more extreme results become progressively higher toward both ends of the scale.

All in all, over time, players will perceive a 3D6 system as being more realistic, but also more predictable, and a D20 system as less realistic, but maybe a little bit more exciting. (And more slapstickish, at times.) Granted, liberal use of steep modifiers can change these systems’ basic characteristics in both directions, and that’s what rule systems and game masters often do. But when your system produces your intended playing experience only in a Game of Modifiers, you should go back to the drawing board and rework your rules until modifiers are pruned back to do what they’re supposed to do, namely reflect contexts in terms of uncommon difficulties, advantages, or handicaps.

Which doesn’t mean that all this is identical for every system that has equal distribution! A D100 and a D20 system each have equal distribution, but the D20 system amplifies properties common to both. The chance to roll an extreme value, for example, grows from 1% in a D100 system to a whopping 5% in a D20 system. (In a 3D6 system, as you might recall, it’s a mere 0,46%.)

Likewise, different probability distributions also behave differently. To illustrate this, let’s very briefly compare our 3D6 system to a 2D10 system.

Fig.4.20 Probability Distributions 3D6/2D10
Fig.4.20 Probability Distributions 3D6/2D10

With 2D10, the extreme values 2 and 20 each have a probability of 1%, similar to a D100 system, and the chance to roll reasonably average results for average performances is around 58%. That is considerably closer to the 3D6 system’s 67.58% than to the D20 system’s 35% (or the D100 system’s 37.5%). Thus, over time and all other things being equal, players will perceive a 2D10 system as more realistic than a D20 system, but it will give them perceptively more surprise results with interesting consequences than a 3D6 system.

This role-playing probability example should make you attentive to how probabilities affect perceived realism, and how perceived realism affects the interactive playing experience. And, just like our Grime Dice example from the preceding beat, it should give you ideas for designing mathematical underpinnings that shape your intended playing experience very precisely.

Before we proceed, a few remarks on probabilities in digital RPGs. Probabilities of any kind, in most cases, raise some difficulties as the player can just save and reload until the desired “roll” occurs. Some games just roll with it, so to speak, but most games don’t. For more action-oriented tasks like combat, large-scale strategic as well as up-close-and-personal, it’s less of a problem. Here, the probabilities are not only fed by player performance and player decisions, but by a substantial number of actions and maneuvers from all parties involved. That way, it’s easier for the player to ramp up and develop their combat skills than to fall back on saving and reloading the game until the stars are right. It wouldn’t go away as easily with tasks like lock picking, of course, except you turn it into a performance-driven task like, e.g., a minigame.

At this point, we will pass the baton to this beat’s second major aspect, the perceived realism of player input and the contentious issue of input realism. If you design your game’s rules in ways that incorporate realistic player performance, what could be gained and what could be lost with respect to its interactive playing experience?

In Rules of Play, Katie Salen and Eric Zimmerman relate an anecdote shared by pen & paper RPG legend Steve Jackson of GURPS fame at the GDC 1998. According to Jackson, his team had built a driving simulation prototype that was so realistic that only a professional race car driver could handle it. But, importantly, Jackson referred not to a “regular” racing video game but to his aborted Car Wars gaming center project, so he probably tried to wrap an entire car around the player. In other cases, though, when you hear about a video game project X or Y or Z that built a prototype for a jet fighter game or a car racing game or a powerboat simulation game that was so realistic that only a fighter pilot or a Formula One driver or an offshore racing veteran was able to play it, chances are that a good number of tall tale elements are involved.

To gain a better understanding of player input–realism, let’s first have a look at prevalent categories of car racing games to illustrate this topic.

There are three major categories of car racing games on the market right now. With two of them, simulation and arcade, we’re all familiar. The third category can be called upgrader, of which The Crew is a good example. The Crew features a highly realistic open-world background in the Plurimediality territory, a tortuously ludicrous storyline in the Architectonics territory, and specific gameplay moments in small towns and “little places” that feel just right, and realistic, in the Narrativity territory. In the Interactivity territory, though, something seems to be amiss. The Crew isn’t realistic, but it isn’t arcade either. In both simulation- and arcade-flavored racing games, the point is to become a better player (but not a better driver in real life, we’ll come to that). In The Crew and other upgraders, players do not primarily become better players through improving and mastering input controls and mechanics. What they’re really getting better at is upgrading their car through a steady stream of grinding. And as soon as a player has a new car, or needs a different upgrade kit for a different environment, the illusion of improvement and mastery up and dies. As Austin Walker expressed it in “The Crew Review: Postcard America”: “It feels like, at some point during The Crew’s development, its designers lost sight of the fact that people like to drive. [...] The Crew is not a good game for driving.”

Granted, upgrading cars with fancy stuff is what people do in the real world too. It’s realistic all right. But an experienced race car driver will nevertheless be able to hurl any off-the-shelf automobile down the track with miraculous precision. Vice versa, if you’re an average driver, see how much your driving will improve if you find yourself behind the ultra-responsive steering wheel of a race-tuned high-performance engine. Good luck with that.

The point being, you hear surprisingly often that there’s a trade-off between “fun” and “realism” with regard to player input, controls, and game physics because the latter is more competitive and the former more enjoyable. That doesn’t make sense at all. Not only are things more complicated, as upgraders show. Enjoyable experiences, and that is equally true for players who enjoy The Crew, do not correlate with how realistic the player input is supposed to be, or how realistic or unrealistic the game’s controls and physics react to this input.

If your primary target audience consists of simulation game aficionados, you’d better deliver on the realism part for an enjoyable experience. If your primary target audience is dedicated to arcade-style gaming, it makes perfect sense to support phantasmagoric maneuvers that would make Newton’s head spin in his grave. If your primary target audience is more tuned toward role-playing experiences, then upgrading will be their thing. All, if done right, will deliver the intended experience to your audience.

But we’re not done yet. Do players who get better at playing realistic racing games become better drivers? What does it mean to design a realistic car racing game that would make that possible? That’s an important question with regard to realism in the Interactivity territory.

To answer this question, we need to go on a little detour and explore a number of aspects related to media psychology and the debate around violence in video games. There is a popular argument that goes like this. The more realistic the controls, physics, input details, and accompanying game graphics for an activity in a video game are, the more the player will learn about that activity and become better at it in real life. Assuming that a player can become a better driver with the help of realistic driving simulators, a player can become a better shooter with the help of realistic shooting games. (In interviews and articles, researcher and author Dave Grossman even refers to such games as “Murder Simulators.”) And because shooting games’ dominant operational mode for overcoming obstacles and solving problems is to apply violence, the player also develops the corresponding mindset.

This bundle is quite tricky and sneaky because the second argument has some validity and cannot be easily dismissed without discussion, lending unwarranted credence to the former.

Think about it. Even with the most realistic shooter, what you’re getting good at is handling a controller or a keyboard-mouse combination, not a knife, an M4 carbine, or a sniper rifle. Similarly, will you become a better driver in real life and on the track by playing Forza Motorsport, NR2003, or even Grand Prix Legends? No. What if we threw in a full set of peripherals, from a racing seat with gear-shifter mount and foot pedals to a force feedback steering wheel? Still, no. If you’re a professional race car driver, you might be able to use outstandingly realistic racing simulators to get familiar with a track or practice a certain maneuver. But that’s about as far as it gets. For shooters, likewise, even an M4 carbine gaming accessory will not make you a better shooter in real-life situations, let alone a better soldier or police officer (or mob enforcer, for that matter). You will get much better mileage out of visiting the shooting range a few times, or playing airsoft or MilSim. Military or law-enforcement professionals, just like race car drivers, can use simulation video games to prepare for certain situations, but that’s again how far as it gets. (And even that is disputable; see Ed Smith’s illuminating Vice article “In the Army Now: The Making of Full Spectrum Warrior.”) Barring extraordinary contexts and circumstances, simulators can only complement, maintain, and support real-world proficiency levels, not create them.

What it all boils down to is a misconception. Car racing games, even simulation-style games, don’t simulate driving. What these games simulate is the experience of driving, each game in its own distinct, uniquely crafted way.

Which is not only true for car racing games but for any type of simulation games. That’s what you, as a game designer, should aim at. If you’re designing a realistic game about the activity X, your question shouldn’t be: How can we simulate X best? Instead, your question must be: How can we simulate the experience of doing X best? This is a completely different question that will be conducive to making great games, and distinctive and highly differentiated games, that players can enjoy.

Fig.4.21 Realism in Games (Definition)
Fig.4.21 Realism in Games (Definition)

The second aspect of the perceived realism of player input and our final aspect in this beat is the often debated issue of quick time events, or QTEs.

Let’s start from scratch. Players often step in the shoes of professionals in games, like that ex-marine turned mercenary turned headhunter, a gladiator, a pilot, a swordfighter, a talented street kid, or Batman. Rarely does that correspond to these players’ personal real-life pursuits. Likewise, in most cases, these players control their in-game actions with a regular controller, mouse and keyboard, joystick, touchscreen, and similar, not with real guns, swords, combat aircraft, or roundhouse kicks. Now, how do you make these player actions look good on the screen and cool and not amateurish and silly? By way of abstraction!

Commonly, player input is translated into impressive on-screen maneuvers through abstractions that map the latter to the former and create a system of demanding combinations of keystrokes or button presses or gestures that the player has to learn, unlock, or both. This is great if your game addresses a target audience with “core” attributes (as discussed in the Preparation phase’s Level Four: An Army of Avatars) who are not only used to, but expect, tough challenges and steep skill barriers (more on that in the upcoming Beat 5. Readiness). But what if you have a different target audience, one without core attributes? How can your game make these players’ input look cool on the screen nevertheless, and impart a sense, and the joy, of dazzling accomplishment?

A great way to translate less demanding player input into impressive on-screen action are Quick Time Events. Sadly, QTEs have gotten a bad rap for “mash this unexpected button to pulp within a split-second in order to not die,” which even games like Tomb Raider sometimes fail to deem below their station. Another atrocious use of QTEs has been to give the player some flimsy semblance of control in cutscenes that stretch out forever, as in Resident Evil 4, which prevents players from skipping them.

This is unfortunate. Such practices make games less accessible while QTEs offer terrific abstraction techniques to make games indeed more accessible and even inclusive, capable of delivering compelling experiences in the Interactivity territory to non-core players and, potentially, players with disabilities. From its inception in the original Shenmue to games like God of War and Heavy Rain to mobile games like Revolution 60, QTEs have well evolved beyond “quick” and “time,” and they’re not done. Another great example for the use of QTEs is Batman: Arkham Asylum, demonstrated by Scott Rogers in Level Up! The Guide to Great Video Game Design. QTEs offer a whole range of options to map highly abstracted but accessible player input to spectacular maneuvers, convincing combat sequences, or satisfying finishing moves. You can even map them to moral decisions—as, somewhat heavy-handedly, games from the Mass Effect trilogy do here and there. QTEs can do all that without sacrificing player agency, without cutting the perceived cord of causality between player actions and on-screen events, and without severing even once the player’s continuous impression, however illusive, of agency, authorship, and control.

up