September 12, 2014

Against the cult of simplicity

If you’ve spent any significant time going through my archive of design-focused articles, you probably know that I am allergic to dogmatic declarations about how to design games. Over the past four years, I’ve responded to calls to remove text from games, calls to remove story and attractive visual art from games, assertions that video games ought to be all about action and reflexes, and calls to remove leveling and character progression from games.

Still, it has been a while since I last wrote one of these–one might be forgiven for thinking I had dealt with all of the unjustifiable dogmas floating around the indie scene. Unfortunately, my participation in Indiecade last October reminded me that that is far from true.

While at Indiecade in 2013, I had the pleasure of listening to Brenda Romero give an inspirational talk (one which she has evidently delivered elsewhere since) themed after the movie Hiro Dreams of Sushi. It was a talk about seeking perfection in game design. She described a triangle with one corner labeled on time, another labeled on budget, and a third labeled great. She exhorted the audience to disregard the “on time” and “on budget” sides of the game development triangle, and instead aim for a game that is truly great at all costs.

Brenda is a funny and dynamic speaker, and it made for a very entertaining talk. I would have enjoyed it without reservation but for one moment where she exhorted the audience to design games centered around a single core mechanic. Romero indicated that that was the only way to design something truly perfect. I considered asking her to defend that position in the Q&A that followed, but I hesitated. “I already have a bit of a reputation as a gadfly in the indie community,” I thought to myself. “And I like Brenda. Is this really a battle worth fighting?” I chose to let it go–but the memory of that moment continued to nag at me.

Later on in the weekend, I dropped by a tent where Jeremy Gibson was giving a talk on game design. I don’t recall the name of the talk, but it struck me as an intro-level lecture for folks who had not been making games for very long. He, too, apparently felt obliged to spend some time telling the audience to make games with only a single core mechanic. He did not give a reason; he did not limit his statement to new designers who are just finding their feet. He simply said that game designers should do it. Full stop.

Now, here’s the thing: I genuinely enjoy games that employ only a single core mechanic. But I also enjoy highly complex games that leverage many different systems, and I admit that I am deeply uncomfortable with the thought that leaders in the indie community are running around telling everybody that one of these is somehow better than the other.

Before we get into the repercussions of telling people that a game with a single core mechanic is preferable, I want to take a step back and examine the merits of the contention itself: is it, in fact, actually better to have a single core mechanic?

Is using a single core mechanic better?

I’m going to go ahead and answer this question up-front: it depends on the game. Some games do, in fact, benefit from sticking to a single core mechanic. For others, however, sticking to only one core mechanic would be an unmitigated disaster.

Let’s talk about the benefits. I can think of three big advantages that result from focusing on a single core mechanic in one’s games:

1. Reduced time and expense: these are two sides in the game development triangle that Brenda Romero mentioned in her talk. If you’re an indie developer, you are bound to have limited resources. Coding and bug testing new systems means a longer and more expensive development cycle. New systems will probably also mean more art assets to give the player appropriate visual feedback, which will likewise add to your time and expense developing the game. Keeping your game to a single core mechanic shrinks the time and money required to get everything done right, leaving you more latitude to polish the heck out of whatever makes it into the game.

2. Easier balancing: having multiple core mechanics poses a design challenge. As a general rule, the fewer mechanics you have in your game, the easier it will be to balance them and ensure that every mechanic is polished and fully utilized in the game. In some ways, we can analogize this to the dilemma that skill-based character creation poses to RPG designers:

[Skill-based character creation] presents a serious design challenge for any developer. Each new skill requires a new system, or–at the very least–a new branch or skill check in the game’s conversation trees. In turn, the more systems and skill checks one adds to the game, the harder it becomes for the developer to ensure that each skill is more-or-less equally important to the game.

To restate this in more traditional terms: the more core mechanics you add in, the bigger the risk you take of introducing strategic dominance, which will render strategic player choices illusory and neuter the player’s satisfaction from interacting with the game. If there’s only one core mechanic in play, this cannot happen.

3. Improved accessibility: as you add more core mechanics, you increase the number of discrete vectors along which the game’s state can change. This steepens your game’s learning curve, thereby increasing the barrier to entry for new players. Sticking to a single core mechanic means maximizing your game’s accessibility.

These are significant advantages, and all of them are good reasons to stick to a single core mechanic.

However, choosing to use a single core mechanic isn’t quite the no-brainer that this list might seem to suggest. After all, there are more things to value in games than just manageable scope and accessibility: player choice, for instance. Restricting a game to a single core mechanic drastically limits a player’s ability to choose how to approach the game.

Restricting player choice can be good for purely reflex-based games (see e.g. Super Hexagon), or for games which don’t offer a balanced or appealing set of alternative approaches in the first place. For games which do offer such alternatives, however, removing all but a single core mechanic is crippling to the experience.

For example, imagine Deus Ex with stealth and dialog stripped out of the game. It would become nothing more than a mediocre first-person shooter. Imagine Gunpoint without the hacking mechanics: nothing would remain but a fairly simplistic stealth platformer. The breadth of choices these games offer to the player by dint of multiple core mechanics is absolutely central to their appeal.

Multiple core mechanics can also lend depth to games by interacting to create new and unexpected game states. Consider titles like Dwarf Fortress, Nethack or Spelunky. Their popularity with gamers stems from their very complexity. Combining numerous core mechanics and randomized starting states all but ensures that each play session will offer something legitimately new to experience. Games with only one core mechanic can offer up different playthroughs as well, but they have comparatively little power to produce situations where the operation of the game’s mechanics genuinely surprises the player.

Multiple mechanics can be important to the overall experience in other ways as well. Tactical combat may be at the heart of X-COM, but X-COM wouldn’t be X-COM without player-driven research, mission selection and base management mechanics. Increasing resource scarcity combined with escalating encounters on the macro scale are central to the game’s pacing; it would be nigh impossible to accomplish that with only combat in the game. A similar observation can be made about FTL: ship-to-ship battles may be at the heart of the game, but exploration and resource management are equally important to the experience, generating both context and tension that heightens the significance of the individual encounters. Both of these games are almost inarguably better for having multiple core mechanics.

In short: multiple core mechanics can give a game a breadth of available playing styles, it can create a great deal of mechanical depth, it can generate tension, and it can help pace a game.

There are entire genres that the “single core mechanic” approach does not lend itself to. To name a few examples:

  • RPGs require multiple core mechanics. Multiple core mechanics are necessary for creating meaningful alternative approaches to the game; but even the most bare-bones of RPGs, the ones which lack the “role-playing” part, must still have both (1) a core mechanic for handling conflict and (2) a core mechanic governing character creation and advancement.
  • Roguelikes in particular require multiple core mechanics, moreso than most RPGs. They thrive by producing ever-changing, unpredictable environments and forcing the player to painstakingly learn strategies for navigating them. In order to make the various features of each environment consequential, there need to be mechanics to govern the player’s interaction with them. One core mechanic won’t cut it.
  • Simulation games require multiple core mechanics. Reality is complicated, with numerous causes for every effect–to simulate any system that exists in the real world is to attempt to mirror that same complexity. That means multiple core mechanics. Attempting to simulate a complex system with a single core mechanic would not only fail, it would be terminally dull to boot.
  • Survival games require multiple core mechanics. These are titles whose primary appeal depends upon the player struggling to survive in a hostile environment through planning and resource management. To be of any interest at all, there has to be more than one type of resource to manage; and thus, core mechanics to govern each resource’s behavior in the game world.

And so on. Telling all developers to make games with a single core mechanic is effectively asking developers to either ignore vast swathes of the gaming universe, or else to exclusively produce less compelling versions of games that already exist in those spaces.

Simplicity is not elegance

It is something of a truism in the indie game space that elegance is a desirable quality in game design. I seldom see people actually specify what they mean by “elegance,” though. If we are to talk of the desirability of elegance, then we must mean something more than just “simplicity”–otherwise, we are merely making another thinly veiled attempt to privilege simple games and squeeze complex ones out of the universe of titles that receive recognition from the community.

When we talk of elegance, we should not speak of a game’s raw simplicity or complexity: we should be talking about how much complexity the game creates relative to the number of elements used to build it. Elegance should be about achieving a lopsided ratio of possibilities to rules, about generating a huge possibility space out of a comparatively small set of pieces. Achieving this lopsided possibilities-to-rules ratio means nestling into a sort of localized maximum of complexity and freedom relative to the time and approachability cost it incurs. It is, in a word, about efficiency, not about simply making the “right” kind of game.

Simplicity alone is not elegance. Suppose we create a game where two players fly around in rockets and fire bullets at each other–and that is it. No use of secondary objectives or special terrain. A game like this is simple, but not elegant; its elements do not add up to a possibility space larger than its minimal parts suggest. Or, to name a more extreme (and even clearer) example: consider the hypothetical game “Button Clicker RPG” from my old RPG definition article, a game in which we simply click a button and a number increases onscreen. That is not an elegant game. It is devastatingly simple, but the possibility space is also vanishingly small; we both use little and accomplish little. That is not anything particularly praiseworthy.

If elegance is efficiency, then simplicity is something else entirely. Limiting a game to a single core mechanic is ripping the fat off its bones; it is turning a light bulb into a laser; it is driving the experience relentlessly forward in a singular, closely defined, tightly controlled loop. This can be beautiful in much the same way as a minimalist painting or a Philip Glass piece: which is to say, beautiful, but there are alternatives that deserve at least as much respect.

Why we should stop advocating for just one mechanic

“Okay Craig,” you might say. “So not every game should be limited to a single core mechanic. So what? What’s so bad about people telling developers to make games with a single core mechanic?” A few things.

To begin with, it’s incorrect. The “single core mechanic” approach reflects a spartan design philosophy that has long been popular in certain corners of the indie community, one that can perhaps best be summed up as “take a single core mechanic and do everything remotely interesting that you can with it, then end the game.” And hey, that’s a perfectly fine approach to game design–there’s nothing wrong with it–but it’s not the only valid approach.

By elevating the simplistic approach above all others, we serve to diminish the diversity of possible game types that receive attention and praise in the indie game community. Compounding this problem is the privileged position of the people spreading this message. To take just our examples from above, both Brenda Romero and Jeremy Gibson are Indiecade festival co-chairs. At a bare minimum, when festival organizers come out and frankly state “we want you to make games with a single core mechanic,” it gives the appearance that the festival will favor games that meet those criteria.

And when it comes to festivals, this particular bit of advice feeds into a pre-existing problem: judges are volunteers, they are busy, and consequently they are already predisposed to favor games that they can pick up and understand immediately. This means that relatively simple games with a single core mechanic are already at an advantage!

The nature of curated events is that some games will be chosen and others excluded. When that exclusion consistently keeps certain portions of the spectrum of gaming experiences from receiving proper consideration, however, something has gone awry. When celebrated developers tell other developers that using only a single core mechanic is good design, they provide a pseudo-intellectual veneer in which to justify their favoring more simplistic games. If anything, we should be giving festival judges the opposite message: slow down and take some time to appreciate the depth that more mechanically complex games have to offer.

I write all of this not to criticize Indiecade (which is a worthwhile event), nor to embarrass Brenda Romero or Jeremy Gibson (both of whom I have nothing but respect for). I write this piece because I love diversity in the indie community. I love the enormous spectrum of entirely different experiences that games can give us. And yet, I see these events consistently failing a significant chunk of our community year after year.

Bit by bit, I have seen our community growing, broadening, opening. For years, narrative titles fought a long and bitter battle to be included in the indie scene, and now we honor games like Gone Home and Depression Quest. We fought about whether games needed to have challenges and goals in order to be games, and now we honor titles like Panoramical, Dear Esther, and The Stanley Parable.

We’ve made strides–but there is more to be done. We have gotten to the point where we now honor architectural installations and games with literally no interaction at all–and yet, I cannot think of even a single video gaming event anywhere in the world that can be bothered to validate the type of complex, long-form works that I’ve spent my life playing, loving, and creating. It pains me to visit indie festivals and see hardly any strategy games selected for inclusion. It pains me to hear people considered thought leaders in our community publicly elevate minimalist games above all others. And I am very, very tired of supporting events that have rendered themselves structurally incapable of supporting us back.

I want to see games from all parts of the spectrum honored, not just the titles that are easy to grasp, fast to play, or which reflect a prescribed approach to design. Complex games with numerous mechanics are wildly popular with the gaming public, and they have been for decades. When will our legions of would-be indie taste makers catch up?

  • Pack Ling

    It’s nice you acknowledge a problem many of us have identified for a while, where the narrativist approach to games (Raph Koster reference) has taken over the gamist and simulationist space. Not a week ago we were into a discussion where someone pointed how everyone should only play games where they were “intellectually challenged”, which was his misdefinition of being put on a different social/emotional challege instead of a mechanical or logical one.

    Those of us complaining about this trend in social media, the only contact point between creators and customers, we are dismissed as problematic, womanhaters (?!), mouthbreathers, basement dwellers or manchildren. It’s gotten tiresome and some of us just stopped visiting the sites, lost faith in several developers (Double Fine in my case) and kept playing old games instead. It still baffles me how today there are only a handful of games being made with the aim of matching the complexity of several DOS classics.

  • John Doe

    I’ve been ‘complaining’ about it for ages, and have never been called any of those things. If you’re referring to the mob spawned by the Sarkeesian debacle, please refrain from connecting normal sensible people that like complex games to that rabble. They’re despicable.

  • Pingback: Is Simplicity Overrated?()

  • DaveyD

    What the hell does a game with a “single core mechanic” even look like? Visual novels, and some old school platformers, I suppose (or is running and jumping two core mechanics?). But even a lot of platformers add have added many different mechanics (see Flashback: Quest for Identity and Out of this World).

    I would generally agree with someone saying that you should make gameplay as simple as possible, while still allowing the player enough options to make for an interesting and enjoyable gameplay. That is to say, needlessly complex mechanics are a problem, but not all complexity in games is undesirable / unavoidable.

  • Francisco

    I think the real message the peeps should be giving should go more along the lines of “if youre thinking of addind a feature, you either do it right or skip it altogheter, but do not do something half assed”
    In general, anyone that tries to tell you that there only one valid answer to something its full of it
    Lastly, in your examples you carelessly forgot to mention 4x games, shame on you.

  • CraigStern

    Uh, sorry? That was not meant to be an exhaustive list–just a few genres that came to mind. (Hence, the “And so on.”)

  • Francisco

    Hahaha, its okay I was just pulling your leg with that one, the irony its that I neglected tu put a 😉 or something to make it clear

  • Kain Yusanagi

    “needlessly” is subjective.

  • DaveyD

    Sure, but almost everything is.

  • Ike

    “A similar observation can be made about FTL: ship-to-ship battles may be
    at the heart of the game, but exploration and resource management are
    equally important to the experience, generating both context and tension
    that heightens the significance of the individual encounters. Both of
    these games are almost inarguably better for having multiple core
    mechanics.”

    FTL is so fun for a simple reason: it’s so freaking random! There is so much in the game that is out of your control, at the hand of a analogical dice roll.

    Those who have perseverance in the game will see it as more than a luckfest, that there are ways to beat the game and fairly early with a little bit of a luck and a lot of good decision-making, prepared for all possible outcomes.

    But it’s so fun and so replayable because the incredible level of randomization combined with the feeling that you still have enough control makes each playthrough a unique experience in ways that could not possibly exist without the randomness.