CONCLUSION

There is an eBook version of this document with bonus features. You can find that book, and the entire reverse design series here.

The first two things that ought to be addressed in the conclusion are two nagging questions that may have crept into the minds of many readers. The first question is "Does this apply to anything other than Nintendo platformers?" The answer to that is a resounding yes. It doesn't apply to all games, but even a cursory glance into other games will reveal the prevalence of this system. Sonic the Hedgehog, for example, employs the same basic structure. The challenges in a Sonic title are often much wider because they are frequently completed at a very high speed, and the player needs more space and time to react at that speed. But it's nevertheless obvious that Sonic games do contain evolutions, expansions, mutations and probably even skill themes. Mega Man, Prince of Persia, Castlevania and even Half-Life all show considerable evidence of them too. Looking at those games through the same lens of challenge, cadence and theme would probably have very different results than what we see in this document, but the tools are still up to the task.

The second question is, "Does this system apply to modern 3D games?" The answer is yes, although again, it's definitely not true for all games. But an observation struck me during the writing of this book that I think shows that the pattern of challenges, cadences and skill themes works for more than just 2D titles. After writing the majority of this book, I had the chance to play a number of 3D Mario titles again. It struck me that Mario 64 has not aged very well, whereas the first Mario Galaxy is just as replayable today as it was when it came out. Naturally, eleven years separate those two titles and that means something. Indeed, the same amount of time passed between Super Mario 64 and Super Mario Galaxy as did between the original Super Mario Brothers and Super Mario 64. That difference is meaningful, but not in the way one might first assume. The great weakness of SM64 was one of organization. The big, open and awkwardly-shaped levels can be confusing even to players who once beat the game. Having seven overlapping objectives strewn across an irregular, open level is no way to construct a coherent game—as the camera clearly shows! Certainly, some of the Lakitu-cam's dreadful mechanics were a result of Nintendo's inexperience in 3D; after all, there weren't any 3D games quite like SM64 whose example they could follow at that time. But much of the camera's bad behavior can be blamed on the level design, too. Programming a camera that can follow Mario through the weird, experimental geography of these levels would be a truly onerous task. Super Mario Sunshine, which used a very similar design scheme, did a little bit better with its camera but it still has a reputation for having levels which are hard to see.

Super Mario Galaxy solves many of the problems with level organization and camera problems by changing its design. The purpose of the "galaxy" formation is to create discrete challenges in a 3D environment. Any time the designers can use discrete challenges, evolutions, expansions and mutations suddenly become very useful and fairly easy to execute. Yes, there are still some larger, open sections of levels that have wonky camera movements. But, in essence, Super Mario Galaxy manages to reclaim the best design aspects of the 2D Mario games and reiterate them (successfully!) in a 3D game. Other 3D games actually started employing that strategy of dividing their larger 3D environments into discrete content sections quite a bit earlier. Both Doom and Goldeneye divide most of their levels into rooms which contain obviously discrete challenges. Measuring the cadences for either of these games would reveal a pattern probably quite different than the one in Super Mario World. That pattern, however, would be a useful thing to know when studying the design of shooters, so the cadence structure is probably a useful idea for a variety of videogame genres.

With that in mind, let's look at the essential lessons from Super Mario World.

(1) Standardize your sizes and distances.

It is immensely helpful to both the player and the designer for every size and distance in the game to be divisible into obvious, standardized increments. Super Mario World does this by measuring everything in whole coin-block-lengths. Because all platforms, distances and obstacles correspond to the coin-block size, it's much easier for the player to intuit how difficult any given jump event should be, at least as far as the size of the jump affects that difficulty. It's also easier for the designer to calculate and modify a jump by increasing or reducing it by a block length. This reliance on unambiguous whole-block lengths avoids one of the greatest flaws in a platformer, which when a player asks in frustration, "Am I supposed to be able to make that jump?" Recently, I played through the original Donkey Kong Country, and found myself asking that question quite a bit, even as I got deeper into the game. For various technological and/or aesthetic reasons, there are many points in the game where the terrain is structured so that it's unclear whether a given jump is possible, or at what spot the player is supposed to start a jump. This ambiguity is problematic. It's hard to know whether a player is failing in these situations because he or she is messing up the jump, or whether the jump isn't supposed to be possible at all. Maybe there was another path somewhere in the level that led to the platform, or a missed tire or mountable animal along the way. The player gets ambiguous feedback when positive or negative feedback would have both been better.

Slopes like these in Super Mario World are the closest the game comes to truly ambiguous distances. The slope in Donut Secret 2 slides through block lengths, but there's still no ambiguity because the player can figure out the sizes of things from landmarks on either side of the slope. From this clearly measured context the player can intuit the plausibility of any given jump, and then deduce if there's a missed path or some required technique not immediately apparent. Trial and error may still be necessary, but at least the clues are there. What's more, it's not as though the designers had to go through and make sure every slope was surrounded by an obvious context; standardization of sizes takes care of all of that naturally.

Some extra guidelines:

-Super Mario World standardizes height and width to the same measure, which is one coin-block length. Standardizing to two different units of height and width might have unforeseen complications, but we know the strengths of one standard.

-The standard unit must be large enough to be intelligible. Players aren't going to be using a ruler on every jump, but they will internalize the size of jumps if the unit is large enough that the difference between a jumpable obstacle and a non-jumpable obstacle is clear. One thing that Super Mario World does is it makes the standard unit of measure for the game the same width as the player character. It's easy enough for a player to quickly estimate the size of a jump by the size of their character.

(2) Don’t use real physics, especially not real gravity

Real physics are complex. Consider that we pay millions of dollars to professional athletes who make it their business to manipulate the physics of their own bodies to score points, in one way or another. Yes, those athletes are also in incredible physical condition, and they work hard to become so. A videogame can grant you an approximation of a professional athlete’s abilities. But there is no controller in the world sophisticated enough to grant you control over all those faculties. Instead, designers simplify the physics in even the most realistic simulations. For Mario games, this simplification is enhanced even more, but this is a good thing. By simplifying the physics of Super Mario World, the design team allowed players to focus not on mastering a finicky set of baseline mechanics but rather on larger, more engaging challenges that combine mechanics and skills in interesting ways. This is especially true for gravity in platformers; players don’t need to be doing calculus all the time. Super Mario World accomplishes all of its design goals with two falling speeds in dry levels, and two swim/walk speeds in water levels. Intricate physics might work in some games, but they’re by no means necessary, and might even be a hindrance.

(3) Use composite game design techniques, especially when working on mechanics.

The basic idea behind a composite game is framed by the question, "how does the player use the mechanics/skills of x genre to solve the problems of y genre?" In Super Mario World, it's about how Mario uses platforming skills and mechanics to solve action problems, and vice versa. That vice versa is pretty important, because it's responsible for the back-and-forth flow between declensions. The four skill themes all show how momentum, reflexes, precision and timing can solve a variety of both platforming and action problems. It's not hard to see how this could extend to other games. The simple part is simply reframing the question of how x genre mechanics solve y genre problems (and vice versa) in the game you want to analyze or the game you want to create. The hard part is figuring out if there's another division like timing/speed that works for the mechanics you're interested in. To some degree, steps four and five will help figure out that problem. The next two design lessons (below) will also help with the most important thing: always move back and forth between composited genres. That back-and-forth rhythm is what makes composite games so enjoyable. No game is perfect, but by making each level just different enough, players will be frustrated less often, and will more roundly learn the skills the game has to teach them, without even knowing that they have.

Do you have to design a composite game in order to make it good? Of course not—there are alternatives in the arcade and set-piece styles, although the latter of those two styles can overlap with composite design significantly. But if nothing else, this book should make it clear how composite design makes it easy to create novel, coherent content. It also makes it tremendously easy to organize that content and view your own content critically, as some of the other lessons (below) will illustrate.

(5) Iterate and then accumulate

All of the skill themes have one thing in common: they begin with iteration and end with accumulation. They all differ in the relative proportions and exact executions, but it's clear that this process does take place naturally. How could it not? If a designer is working on several levels based on the same core ideas (i.e. an incidental or intentional skill theme), he or she must iterate before accumulating. How can two ideas stack up in an accumulation if there has only been one idea?

But the point of this piece of advice is to begin the design process with iteration and accumulation in mind. All of the ideas in this book are supposed to be tools for designers to use in their games or for students/hobbyists to look for in the games they study. The games in question don't have to be defined by iteration and accumulation. It can, however, be useful to look at that material in terms of how it iterates and then accumulates. And really, this doesn't limit a designer's creativity by much, if at all. Consider how differently the periodic enemies theme and the moving targets theme in Super Mario World handle iteration and accumulation, with the former having two multi-level phases of accumulation and the latter having only one level of it.

Every theme in every game is going to be different, but at the same time, understanding this pattern may help to make greater sense out of the games we study and make.

Final Thoughts

Thank you to everyone who made this book possible, especially to Amanda Lange who helped with research and editing (a lot of work!), and Nicole Kline who scrutinized every word especially hard. Thank you to all the Kickstarter backers too, as this book would not have happened without the funding. We certainly hope you enjoy it! This book raises many questions about game design which it does not universally answer. If you are interested in a particular point not covered thoroughly enough somewhere in the text, you can always drop a line on Twitter to @tgdfweb or send a question to submissions@thegamedesignforum.com