Super Mario Maker: Nintendo's essence, in your hands
30 years on from defining the concept of "platformer" with the original Super Mario Bros., Nintendo has cracked its star franchise wide open with its Wii U title Super Mario Maker. The toolset is robust; it contains everything you need to make a compelling Mario level. Most importantly, it's both incredibly simple and accessible, and also allows for rapid iteration and playtesting, which all developers know is key for the creation of great games.
I've been playing Mario games for just shy of 30 years myself. I got my NES for Christmas 1986, and Super Mario Bros. with it. Since then, I've played every iteration of the platforming franchise, 2D and 3D. But I'm not a game designer (and would never claim to be). So how's Super Mario Maker working out for me?
Again, I'm no game designer, but by taking my ideas through execution, I'm beginning to see in a concrete way all of the things I've heard and read about over my years at Gamasutra. I think, if anything, Mario Maker is a valuable tool for the appreciation of the craft of game-making.
Then again, I've been playing some of the gimmicky, thrown-together levels already up on Nintendo's servers from the press-only pre-release period, and I'm not so sure everyone sees it that way.
Observations from the trial period
I don't like a lot of what I'm seeing, and the ones I like least fall into two broad categories: "I have a clever idea" or "I want to make a really hard one." The latter is an understandable impulse. Everybody who's played a bunch of Mario has a dream of making a dreadfully difficult course; hell, "punishing Mario" is more or less the design doc for Super Meat Boy.
But while Super Meat Boy is very carefully designed, with great attention paid to the player's experience, bad Super Mario Maker levels commonly take a "more-is-more" approach that chokes out the player. And often, arbitrary and unpredictable punishment (enemies unavoidably falling from off-screen, for example) is passed off as "challenge."
I get it, though. After spending an enjoyable evening creating a stage with my husband, we were pretty satisfied with our design, so we showed it off to Tim Rogers and Michael Kerwin (Videoball). Tim calls levels like our first try "the programmer can do it" -- in other words, action game design that's considered "done" once the person who made it can complete it, with no consideration paid to anything beyond that. That's not okay.
He told us stories about balancing his games; since that conversation, I've managed to resist my own worst impulses by keeping in mind the design wisdom I've received over the years, both as a player and as a journalist who works closely with developers. That first stage has been through many iterations now, and it's very different.
Since that night, we've often referred to something Anna Anthropy said at GDC 2013. It went something like this: In Nintendo's signature boss battles, you figure out how to beat the boss, and then you have to execute the pattern several more times to win. Why?
As we worked to improve that course, we began to realize the challenges were, indeed, excessive. We worked to focus them, and made the level shorter instead of repeating the same idea over and over. Sure, you can place 10 flying red turtles, and then require the player to hop across all of them to make a safe landing. You can also place three, and then move on to a new idea.
And then there was the direct and honest advice of another professional game developer, Double Fine's Anna Kipnis (Broken Age), who playtested another course we made. After all, since she works at a studio, she's used to giving game design critique.
Observing her play and listening to her feedback was as useful as you'd expect in helping me understand what I should and shouldn't be doing with my level designs. Interesting, also, was seeing her commitment to surmounting the challenges I'd devised -- a great way to figure out which were fair and which unfair. And the most important thing was being able to use someone else's well-made observations as a jumping-off point for my own ideas on refinement.
A playthrough of the resulting course, "Another Castle," in its final iteration
"I have a clever idea," though… That's cultural, and much harder to work with. These are the worst levels, in my opinion, that pop up, yet some of the most popular, judging from the user ratings.
It's an affliction that's much bigger than Super Mario Maker, in fact. I think Mario Maker, like anything else in this choice-saturated era we live in, will reward those things which give the user a quick stab of emotion over the ones that reveal their possibilities after consideration and understanding; one "lol" is worth a million "oh, now I get it"s.
After all, even during the press preview, there are enough courses available that you can quit a level if you die and easily try another. The ones that make people go, "Woah, that's neat!" are what's so far rising to the top of the heap -- whether or not the execution of that novel idea actually works, sometimes even in a literal sense.
Anyone can see what percentage of players who've tried a level actually complete it. Even courses that have miserable, single-digit clear percentages are getting more stars than most thoughtful attempts to create an honest-to-goodness Mario level. And I say this as someone who's excited to spend time playing some great courses over the months and years to come.
What both the "hard" and "clever" approaches have in common is that they reject (whether deliberately or not) generally accepted tenets of good game design -- either imbibed directly from the Mario series or communicated in any form by any other source that I've ever paid attention to.
Even if this is a Mario game, after all, it's also a wide-open toolset; ideas need not mimic Nintendo's canon. But when I reflect on design insights that fit Super Mario Maker, the ones that stick with you -- there's Valve's advice on Portal, Fumito Ueda's on Ico's subtractive design -- their lessons aren't represented, either.
Mario Maker is proving that more is less. My husband is playing while I write this, and I'm hearing a truly awful cacophony from the TV. "What the hell is going on over there?" I ask. "I have no idea," he replies.
My curiosity roused, I wander over and try a different level. My only question after playing it was this: Did its creator playtest it and think, "This is fun" before uploading it? Or did they just get something up onto the screen and push the button?
In other words, Super Mario Maker is a microcosm of the struggles the game industry -- even our internet-enabled culture -- routinely faces, painted in bright primary colors. And that's kind of fascinating in its own way.
There's always the optimistic view, of course:
Let's hope that pans out.
When it comes to being analytical about Mario Maker itself, I think the twin keys to unlock its potential are: Playtest your levels with as many people as you can, and be thoughtful about the player. These are generally accepted principles of game design, so that's no revelation. But to make the game a success for players, from Nintendo's perspective, there needs to be strong curation.
But the question will naturally be, then: What do players actually want to see? That will be fascinating to find out. And whether that will have implications for the direction Nintendo takes the mainline franchise may sound unlikely but is hardly impossible.
The Mario Maker experience
And then I inevitably come to discussing Super Mario Maker more concretely. The thing about Mario Maker is this: It's more or less flawless. Mario grognards like me can easily point out the small variances from canon (that enemy isn't in Super Mario World, and that object didn't work that way in 1990, and Nintendo's designers would never do that in any of the real Mario games), but indulging in such commentary is to miss the old-growth forest for the trees.
It's clear the collection of tools was selected for simplicity, completeness, and, from a design perspective, extensibility. Not everything works just like it did in 1985, 1988, or 1990. But so what? Discovering just what you can do with these tools is a game in itself. To play Super Mario Maker is to experiment, and to envision new possibilities one after the other, continuously -- I've experienced it myself.
Ideas that have never been in the Mario franchise before are, in fact, essential -- just as essential as Mario Maker's tweaked but convincing reproduction of the controls and aesthetics of the franchise's star installments. They're essential because they are a road to unlocking Super Mario Maker's true potential. What Super Mario Maker unquestionably contains is a universe of possibility with a perfect UI for tapping into it.
It's not just that it's easily understandable and responsive. There's another layer, one that's characteristically Nintendo: You will discover new ideas and interactions by playing what you've created; to change the functions of objects, you have to shake them. You can extend their functions by stacking or combining them. You need to experiment.
It's very Nintendo to make the act of creation so tactile. The game taps directly into the company's heritage of creative projects, most recognizably 1992's remarkable Mario Paint, to great effect.
Mario Paint. Image from MobyGames
There was never a question of whether Mario Maker was a good idea. We've all had it before. Nor was there a question whether Nintendo would understand what tools people need to make good levels. The only salient question was whether it could be packaged intelligently and presented in a way that's both easy to use and powerful; the answer is yes.
Making "P is for Penance"
I took some screencaps of a stage my husband and I made. Here's a quick rundown of the iterative process we went through with the level; you can watch a playthrough above.
Having noticed that the P-switch item (which turns coins to blocks, and vice-versa, for a limited time) also halts conveyor belts, I had a basic idea for a stage that relied on that mechanic: platforming across stopped belts in the sky.
The idea is born, in a limited way. I started out using the Super Mario Bros. 3 creative palette, but we switched to Super Mario Bros. so players could not lift and carry the P-switch.
An early execution of the basic premise: Climb the conveyors and make it to the next P-switch before time runs out.
At some point we switched to the idea of making the level a Ghost House. That idea did not stick.
Undead koopas did, though -- and we added challenge with homing Bullet Bills. They're not as scary as they look.
This is what the level ended up like: P-switch challenges and obstacles to avoid. Fire Mario's fireballs are powerless here, but that's deliberate: it's so you get two hits to make it through the level.
Thinking back to making that level, the final concept for the stage emerged in about an hour, and another hour was devoted to polish. Again, Mario Maker quickly teaches you, in a hands-on way, game design lessons you'll hear again and again.
The iterative process is so simple, and the Wii U GamePad -- an answer so often in search of a question -- is ideal for Super Mario Maker.
It's been particularly interesting to play with this unfettered creative tool while reading Nathan Altice's "I Am Error," a book (part of Ian Bogost and Nick Montfort's Platform Studies series) about the development of the NES and its quintessential games. To illustrate the hardware's capabilities from both a technical and game design standpoint, Altice focuses on the development of Super Mario Bros., the title that utterly defined it.
Consequently, as I'm offered this remarkably simple, convenient, and flexible way to create Mario levels with little restriction on scope, I'm learning about how the original programmers of the game coded levels in tightly-packed assembly, and how the game squeezed out every last drop of potential in the NES' initial, unaugmented cartridge format.
I've read about how the NES' tile-based design effectively divided the screen into a grid, much like the one on the screen below. The console drew both sprites and background layers with specific priorities under severe constraints. It's a fascinating contrast to ponder as I drag and drop enemies into place, and the only constraints are the ones I impose on myself.
Super Mario Bros.' dev team worked in a grid, but definitely didn't have the handy tool you see here, which records Mario's path. It lets you easily tweak object placement.
"The data for the levels was entered manually. We'd sketch a design for the level on a piece of paper then pass it on to [the coding team], saying: 'Please make it look like this.'" These are the words of Nintendo's Takashi Tezuka, who worked on the original game alongside Shigeru Miyamoto, and still helms the series' 2D installments.
I'm well aware that there's nothing technologically remarkable about Super Mario Maker, just like anyone who's played a current-gen game is aware there's nothing technologically remarkable about the last good 2D Mario game Nintendo released, New Super Luigi U.
There is something intrinsically remarkable, however, about the compellingness of this franchise; Nintendo's willingness to create and release a toolset that breaks its secrets wide open for anybody to construct, rearrange, and remix however they like is a significant step for game culture.
30 years ago, this is how Nintendo handled development of Super Mario Bros. (as discussed here). Now, a flick of the fingers is all that's needed to create a Mario course.
A colleague recently asked me if Super Mario Maker is a "must-buy." We throw that term around a lot in the game industry. Years of hype and marketing have rendered hyperbole banal. This month's "must-buy" is next month's trade-in, and last year's vague memory. Of course Super Mario Maker is a "must-buy," but in a much truer sense of the phrase. To see the underpinnings of the console game as a form laid bare like this, and to be able to take them into your hands and manipulate them? There's nothing that could be more worthy of attention.