“How Hard?” More Like “Hard How?” – A Response to The Meta


So I found these super cool guys at Kill Screen who run a column called The Meta. It’s the way they talk about games that caught my attention. Just look at the opening to a conversation they had about animation canceling in games:

“All animation cancels should be destroyed.”

Woah. I can count the people I know who make statements like that on one hand. I love that it’s such a strong position, because even if someone disagrees, the conversation that follows is bound to be interesting. You should definitely check it out.

The fundamental question of the conversation is this: “Is it okay for a game to have unintuitive, unexplained aspects of its technique in the name of skill?” Animation canceling, can of worms that it is, is only one aspect of the issue. And if you’re reading this and saying “WTF is animation canceling?” consider yourself a perfect demonstration of the problem.

But there’s one quote in particular that I want to talk about. It’s a quote that subtly colored the entire conversation, and it exposes some of the biases that we have about how to make games.

Justin Groot: “Aren’t mechanics that make things arbitrarily more difficult essentially the definition of any competitive game though?”

What a question. There’s a lot of baggage implied there, and I want to unpack it. It seems obvious that competitive games need skill, but it really matters what he means by “skill.” Not all skills are compatible with each other, and using conventional skills like dexterity and reflexes has a way of limiting a designer’s imagination, of cutting off all kinds of depth that other skills have to offer. Above all, designers should be purposeful and shrewd about which skills they put into their games instead of relying on whatever skills accidentally made it in.

Gotta Have Skill

Let’s dig into that quote. Groot suggests that for a game about competition to work, there have to be things in it that are hard to do. On the face of it, that’s not wrong. Without some capacity for players to outperform each other, there’s no room for competition. But remember the context for that quote: he’s referencing a Dota 2 technique called “orbwalking,” a janky, unintuitive fingering technique where you alternate your mouse clicks between an enemy to attack them, and next to them to walk toward them. All of this serves to interrupt attacking animations that would otherwise have you moving slower. In other words, it’s a little dexterity mini-game you always need to play with your hands in order to optimize your performance in the game.

In that context, Groot seems to be saying something like “Well, I mean you gotta put something weird and janky into the controls of the game, otherwise there’s no SKILL to it!” And to be fair, orbwalking is legitimately a skill. Drifting in racing games is a skill, and the infamous “snaking” of Mario Kart is non-stop drifting for people with caffeine jitters. Bunny hopping in Counter-Strike is a skill, and Counter-Strike surfers elegantly fly, glide, and vault themselves through digital skate parks that you navigate by sliding around on your butt. Rocket jumping in a first-person shooter is a skill, and Team Fortress Classic players played on conc maps, dedicated stunt tracks where players flung themselves across impossible jumps and onto tiny islands of safety by blowing themselves up with grenades. One against the wall, one in the hand, nothin’ but net.

All of these things are definitely skills, and every single one of them is definitely weird and unintuitive. You can find these things in a lot of our most popular competitive games. So it seems pretty reasonable to think that games need a certain jankiness to function, because you just don’t often play games in which there isn’t any.

This attitude is actually pretty common. We get so focused on adding skill to games that we kind of don’t even care what type of skill it is sometimes. And God help you if you ever suggest that maybe you should take some of those skills out. Canceling League of Legends animations with flash? Lag-stabbing in Dark Souls? It doesn’t seem to matter how bizarre or – let’s just say it – stupid the technique or controls may be, a lot of people are just happy the skill is there at all. So when someone suggests doing away with some impenetrable, oddball gameplay-from-a-bug that requires a tutorial to even know about, the inevitable response is “But if you take that OUT, the game sucks!” along with a salvo of “git gud.” They mastered the skill, so why can’t you? Basically, games gotta have skill.

A skill doesn’t even need to be that janky, confusing, or bug-induced to qualify as “arbitrarily more difficult.” The headshot in first-person shooters is a completely straightforward example. See that guy in front of you? If you put your pixels on his pixels before he puts his pixels on your pixels, you win! Whatever else is going on strategically in this particular encounter, pixels must be pixeled before anything else can happen. And if you can put your pixels on the head-shaped pixels? Extra doubleplusgood.

It might seem to you like aiming in first-person shooters doesn’t really qualify as “arbitrary.” If so, hats off to you, Mr. Golden Arm. But have some perspective! It may seem a simple matter to you, but to people who didn’t grow up riding that bike, it’s some otherwise meaningless sleight of hand. Trying to pixel that pixel can feel pretty frustrating: there’s really no reason why you wouldn’t do it, and the only thing stopping you is moving your muscles in not quite the right way, or – dare I say it – an arbitrary way.

It may feel natural to put headshots into a game already about aiming. After all, it allows people to distinguish their aiming skills in more fine-grained detail – it is a skill. Just don’t forget that this particular skill is not much more than something to do with your hands.

The Skill Smokescreen

So what’s wrong with adding skill to games? Well, nothing really. But you need to be careful about which skills you’re adding to a game, because the ones you choose can lock you into a design paradigm that’s hard to escape. A lot of games lean too heavily on old, familiar skills, and it limits their potential in subtle but tangible ways. A game that asks a lot of our hands often forgets to ask very much of anything else.

Let’s go back to headshots. The primary skill of a headshot game is the arbitrary flick of the wrist. It’s such an important skill that many players take great pains to have their mouse sensitivity just so, their desk space with ample enough room, their mouse precisely the one they’re used to. If they flick their wrist just a little bit wrong, too bad so sad.

Is head-clicking the only skill? In all likelihood, no it isn’t. Maybe there’s some team tactics going on where players have to push together. Maybe you have some kind of capture-the-flag dynamic going on where teams need to split their forces for adequate map coverage. But whatever else is going on, head-clicking tends to be the single most important skill. If you can’t head-click, you don’t get to play. You can be the most cunning, tricky bastard in the whole game, but none of that matters if you get your head clicked before you can do anything.

So what happens if you take that skill out completely?

Let’s imagine a head-clicking game where heads are clicked automatically. A person comes onto your screen, you look at him, and the game takes care of the rest. What does that game look like?

Probably pretty stupid, if we’re being honest. Players are getting shot down the moment anyone looks at them. Battlefronts of territorial control are locked in an eternal stalemate as players dispense with the enemies who showed up a moment before them, only to be similarly extinguished by the next generation of auto-head-clickers.

But as silly as this looks, let’s keep looking for a moment. This is what any player actually wants the game to look like, for just them anyway. This is head-clicking’s ultimate end-game, and every hour of practice is another step toward making this a personal reality. So if you suddenly give perfect head-clicking to all players, that axis of skill goes away. What’s left? How else are players going to distinguish themselves from all the other aimbots?

In order to have further avenues of skill, other interesting ways to deftly navigate a system to your advantage, there has to be a system there in the first place. And a designer who focused too strongly on the head-clicks can’t be sure that very much of a system is there at all. Maybe there’s a few weird tricks that ended up in the game by accident, but relying on accidents to make your game great is like standing in the middle of a storm and begging for lightning.

“You won’t BELIEVE how high you can jump with this one weird trick!”

This approach severely impacts the development of a game’s design. Part of designing a competitive game comes down to watching how people play it and seeing how your system holds up. If everyone starts doing the same few things every game, there just isn’t much there for players to explore and learn about. And the designer needs to actually see that in action to know it’s happening and intervene.

But there’s a cheap way out out of that. There are in fact many ways. And head-clicking is one of them. You see, if you just raise the execution skill-cap as high as you possibly can, you can get fine-grained skill differentiation right now. If you give players busy-work for their hands, that’s a concrete skill barrier you can see.

Head-clicking (any kind of aiming, really) and other dexterity challenges all serve as a giant smokescreen for system depth, and not just for the players – game designers can blind themselves to a system’s lack of depth elsewhere. Since players are all contending with their physical abilities, trying to push the limits of what they can do, there’s less room to contend with the system, to explore what they should do. And that part of the game quickly falls off the development radar. As long as the game’s already “got skill,” it’s easy to lose sight of just how interesting the system itself may or may not be.

What Else Is There?

The quest to put “skill” into games is often heavily biased toward more obvious and straightforward execution skills, and once you have skill in that one ultra-narrow channel, you no longer feel like you have to worry about other kinds of skill. The result is all too often a game that looks suspiciously like something we did last year, or last decade. What is it that makes designers lean so hard on those janky tropes? Simply put, it’s because they’re easy to understand, and they’re a pretty quick fix.

You see, coming up with ideas for adding other kinds of skill to games is actually really hard. Just imagine: a designer is trying to make a system where the answers to its problems aren’t obvious, one that’s hard for humans to see right through, and he’s using a human brain to do it. That is mind-fuckingly hard. But finger-dexterity mini-games? That’s way easier to get my brain around. Double-blind guessing and mind games? Heck, I could shoehorn that into any game! When you consider the difficulty of inventing from scratch a system that addresses a more cerebral skill set, using those kinds of shortcuts is comparatively easy. But in a way, any time you see a designer resort to that, that designer stared into the unmitigated truth of a shallow system – and blinked.

What would it look like if they didn’t? It’s hard to say, because so many designers give up. And why not? It doesn’t really feel like giving up, and it produces an acceptably playable result. What good could come of trying something new?

Well, let’s find out. Let’s look at Dungeon of the Endless as an example. I mean look at this game:

Here is a thing shooting some things.
Here is a thing shooting some things.

Look how easily this could’ve been a game about lining up your pixels with the enemy pixels and pushing a button called “GUN.” Look how easily it could’ve been about occupying your hands with twitchy fingerings. If you’re not familiar with this game, that’s probably the first thing that occurred to you – you’re saying “This looks like a game where you move with WASD, maybe a gamepad, and click to shoot at bad guys.” Your first thought was that this game is like all the others.

But moving and shooting isn’t what Dungeon of the Endless is about at all. For one thing, there’s no micromanagement to this game whatsoever. Battles happen practically automatically. You tell your gunners where to be, and… that’s it. They take care of the rest. If there are enemies in the room, they shoot at them all on their own. You can’t make them dodge. You can’t tell them what enemy to kill first. You can’t even tell them where to stand in the room! What gives? Why didn’t they put any SKILL into this game?

Because all the skill lies elsewhere.

The skill of the game comes down to putting your gunners into the right rooms, and moving them around as needed. It’s about building enough supporting defenses to keep them alive while managing a finite amount of resources. It’s about finding the best place in the level to put your resource generation devices so they’ll be safe, but putting them up early enough so they give you a good return on your investment. It’s about exploring the randomized room layout in a strategic order so you don’t expose a flank to your home base or overwhelm one of your battlefronts. It’s about manipulating where enemies are allowed to spawn to your advantage. It’s about knowing when you can keep exploring to get more resources, and when it’s time to GTFO. Yes, things move. Things shoot. But those kinds of details happen all by themselves and are beneath your notice – you have more important things to worry about.

...and here's the bigger picture.
…and here’s the bigger picture.

There’s a lot going on in this game despite the lack of any dexterity requirement, indeed because of the lack of dexterity requirement. It’s a system that would never have existed if the designer had settled for a simple 8-way shooter. Because now imagine a version of this game where you do have to headshot things. That’s a completely different headspace for a designer. Now that you expect the player to aim at things, your impulse is to make things harder to aim at, to raise that aiming skill cap. The design starts to coalesce around hand twitches. And since the player is already pretty occupied with all that aiming, it’s impractical for him to deal with much else in the moment-to-moment gameplay. So… you just don’t put any in. When all the player can reasonably deal with is what’s in this room right now, the game zooms in again, turning back into just a thing shooting some things.

Dungeon of the Endless didn’t settle for that. By formalizing headshots into “just a thing that happens without you having to worry about it,” the game frees the player up to consider the bigger picture. And since that big picture is now the focus for both player and designer, now there’s an impetus to create a whole system to govern it. The rejection of micromanagement exposes the macro level of the broader system so a designer can see just how engaging it really is. It focuses the development on making that system as interesting and fun as possible.

That puts a whole new spin on the earlier assumption that taking out arbitrarily difficult skills makes the game suck. Maybe taking the weird, janky skill out of a game and making it suck is exactly what you have to do. Maybe you need to take your head-clicking game, give everybody an automatic head-clicking aimbot, and start over from there. By making the game suck, you’re exposing that badness instead of burying it in quick-fixes. Only by confronting your boring game head-on, face-first can you see clearly enough to lay a new foundation and truly fix the system on its fundamental level. And as a bonus, by not relying on traditional skills, you’re a lot more likely to be exploring new and fresh kinds of skill in your game, stuff no one’s seen before. That’s pretty rad.

On top of that, you also take that fresh, new skill and make it more accessible. If you don’t take out those janky fingering tricks, they become barriers for people who may like the premise of your game, but are unable or unwilling to figure out the jank. Making a game is not a matter of dumping as many kinds of skill into the thing as you can. Inevitably some skills become more important than others, and fiddly hand-janks always top that list. Players need to be able to, y’know, operate the game in order to explore what it’s actually about. Is your game only about hand-janks? Take them out and you’ll find out pretty quickly.

Lose the Crutch

So is Dungeon of the Endless super amazing? I dunno. I’ve only played it for a few hours or so. If you want to tell me that it doesn’t have the strategic depth of your dotalike of choice, I’ll understand. But what caught my attention was how everything that makes it interesting comes entirely from what the system asks of you rather than the controls. It sits on a solid systemic foundation, and it’s a pretty compelling proof of concept that game skill doesn’t need to resort to muscle memory or janky tricks.

Now contrast that with Groot’s suggestion of how to add depth:

Justin Groot: “Like in Gears of War, they replicated the reload cancel mechanic with a visual minigame that let you reload faster if you timed it right.”

Pixel that pixel!
Pixel that pixel!

…really? The height of design is inserting little quick-time events into gameplay every few seconds? Sounds more like the mother of all quick-fixes to me.

Let’s not go back down that road, not when games have come quite a way toward minimizing wrestling with inputs. Dota 2 does pretty well on the strategy front, and it definitely doesn’t need to rely on silly stuff like orbwalking for depth. Whatever strategic nuance that it adds, a clever designer can preserve it while ditching the fiddly hand-janks. In Smite, for example, all players automatically orbwalk all the time, and attacking slows your movement speed to compensate, all on its own. In Overwatch, Pharah doesn’t need to fumble around with manually aiming rocket-jumps by twisting all over the place – the simple push of a button executes a perfect “rocket-jump” every time. No weird tricks, no dexterity mini-games, just the gameplay as intended.

I think it’s no mistake that over the years, as video games have gotten easier to control, the potential for systemic complexity has gone up correspondingly. When the controls occupy less of a player’s headspace, there’s more room for considering a system, the game proper. While the potential is there, designers need to be careful to not squander it by taking all those old shortcuts. Stare at that system, do whatever it takes, and don’t blink. It will be tougher to design that way, for sure. But any artist knows that limitation is what drives creativity. If you can’t use “arbitrarily more difficult” controls to prop up your game, who knows what you might come up with instead?

7 thoughts on ““How Hard?” More Like “Hard How?” – A Response to The Meta

  1. Remy77077 November 27, 2019 / 4:41 pm

    Based on your comments, as well as the article, I wonder if I’m reading it correctly that you think all current fighting games aren’t worthwhile games, as behind their “skill smokescreen” you’d consider they “break down” to RPS quite quickly?

    • The Ludite November 27, 2019 / 10:48 pm

      That’s an article I hope to write sometime soon, but it’s a difficult article to write, primarily because it’s a difficult pill for fighting-game fans to get down.

      The problem that I have with Street Fighter clones is that they don’t really have an arc. They’re not much beyond a simple “best of X matches.” This goes beyond the obvious “2 out of 3” that most games use – each one of those three bouts is really a series of mini-contests, each lasting less than a second. Both players approach each other for a game of RPS, they both push a button, and the game is immediately over and ready for the next one. It’s RPS turned up to lightning speed. You play maybe 10-15 of these individual RPS games until someone runs out of life. You play 2-3 of these life bars, and you have a set. The format of the game is just more and more elaborate ways to group the RPS together, where the game takes 80 or 90 isolated mini-contests and recontextualizes them all into one “match.”

      The reason Street Fighter clones struggle to create an arc is because the possibility space of the game is extremely constrained. The things that you do in Street Fighter are very direct. You push a button to immediately do damage, faster than anyone can react, and everything else in the game revolves around that sort of move. Some moves do damage slow enough to react to, but you use them against faster moves that they counter. You block in anticipation of those fast moves. And the natural counter to blocking is the throw, a damaging move with no startup time that’s impossible to react to.

      All of these moves in the game happen immediately and resolve immediately before returning the situation largely to neutral. There isn’t much in the way of lasting consequence that alters the conditions of the game going forward. Ostensibly, there’s a positioning-based game where the players try to maintain an optimal range, or try to put the opponent in the corner, or try to set up projectile traps and such. But no matter what’s going on, all you really have to do is win the next RPS exchange. Just do the move that counters your opponent’s move, and you can resolve any situation.

      The most important thing is doing the explicit counter to your opponent’s next move – if you can do that, all the other stuff in the game becomes irrelevant. So there really can’t be any sort of long-term arc to the game because previous actions from a minute ago or even five seconds ago don’t really echo forward in time to alter the conditions of the game and change up the strategic space. Even the life bar isn’t properly understood so much as an arc as it is a score.

      In practice, not all fighting games are perfectly like this. Super meters and other similar resources do have a small impact on decision-making. Sometimes a character can temporarily alter or augment their moveset, which makes a situation appreciably different. Certain games like Fantasy Strike have toyed with the idea of indirect moves that do no damage but alter the play space. But the importance of out-RPSing your opponent has a way of choking out whatever temporary arc they might create. And that’s because all those things, no matter how indirect, have to exist alongside the primary verb of the game: instantly damage someone.

      The fighting game that comes closest to breaking out of this paradigm is Smash Bros. In that game (and its clones), each time a player is hit, the situation becomes different. Players are gradually knocked farther and farther away, which alters the moveset that players want to use. More than that, the goal isn’t to win a series of RPS exchanges. Winning them helps, but it doesn’t necessarily get you closer to actually winning the game. It isn’t until you hit them with the correct move in the correct situation and orientation to knock them off the stage that the game awards a point. Until then, you’re simply spinning your wheels and possibly allowing the opponent to close the gap!

      This gets even more complicated in a 4-player free-for-all. While most players view this format as a silly, casual form of play, one not as deserving of competitive respect as 1v1, 3-stock, no items, Final Destination, I believe it has more strategic depth than the 1v1 version of the game. It opens up the possibility space and asks for more nuanced decisions that play into the long-term. Who exactly do you try to bump off the stage? How do you go about doing so? You can score a lot of “damage” on another player and get absolutely no points because someone else got the final hit. You need to carefully balance playing for damage and going for knock-outs while also hopefully keeping yourself as safe from damage as you can. Because even if you avoid the battle and let others do the dirty work of increasing damage on opponents, you run the risk of missing a point that happened before it “should” have.

      And this gets even more tricky when you have to think about the fact that eventually you’ll be at risk of being knocked out yourself. When that happens, who do you allow to claim the knock-out? Who do you perceive to be winning so as to deny them points? When do you allow the knock-out? It takes a little bit of time to respawn, and you might miss a knock-out or two in the meantime! But if you wait too long, you might end up in a situation where the leader not only has opportunity to score on you, but has that much easier a time due to your level of damage.

      All four players are making these decisions the whole game, and every decision, every success or failure echoes forward in time to build an evolving situation. Each time it evolves, the decision space alters. Different moves become more desirable. Different targets become more valuable. Different opponents become more dangerous or more vulnerable. You start to waste more and more time returning to the stage from a long hit, and you start considering who gets to knock you out. All of this is happening externally to the RPS “mind-reading” game, and in many ways supersedes it. Many of the “proper” moves that one might learn to do in 1v1 are quite sub-optimal in the free-for-all, and players who train themselves in 1v1 will find that they have many bad habits that don’t work anymore. If you try to approach it with the mentality of a 1v1 player and give no strategic concern to the broader arc created by the four-player chaos, you will lose.

      But players who don’t consider that strategic space often dismiss free-for-all as “too random” or not “skill-based” enough. They might even reject the idea of intentionally dying to particular players as a form of cheating. That’s because in their mind, endless RPS mini-contests is the entire game, indeed the game at its most pure. They try to play free-for-all the same way, and when they don’t win, they assume it must because there’s no skill to it. In reality, the skill of the game has shifted to something else. Something has finally trumped “counter the opponent’s next attack” as the most important thing. And as much as they might be loathe to admit it, there’s something inherently random about fighting games’ RPS purity. Because if it were all skill and completely deterministic, why would we need more than one exchange? What purpose could having a hundred or so consecutive RPS moments serve? Could it be to simply smooth out the inherent randomness of any given one of them? The very format of the game says more about what’s going on under the hood than a lot of players probably want to admit.

      Thanks for reading!

  2. Zera June 22, 2018 / 1:35 pm

    Great article, separating controls complexity from system complexity.

  3. DestroyedArkana December 29, 2016 / 3:19 am

    It’s really tough to nail down “difficulty” when it comes to knowledge, skill, dexterity, and muscle memory. What I think you’re really distinguishing is between a players tactics (execution) and strategy (overall plan). I think there’s definitely a place for both kinds to exist on that sliding scale from just about decision making, to just about dexterity.

    You talked briefly about inserting mind games and double-buff guessing into a game. I would love to hear you talk more about that specifically. I consider that part of the “strategic” side of things. In rock paper scissors it might be a bit too simplistic, but in a fighting game, knowing your opponents tendency to back off at specific times is a huge advantage, or being able to fool your opponent into thinking you have a weakness like that. In another article you mentioned the “kingmaker scenario” and that’s a similar kind of strategic decision as well. I think I prefer gameplay systems that encourage mind games rather than simple puzzles to be solved once you get the right knowledge.

    In a single player game you can test either two things: The players knowledge, or their execution. Once a player gets past the final hurdle that either of those provides, then there’s nothing left to test. Either they can do it and win, or they can’t and have to improve or give up.

    Although if it’s multiplayer, you are pitting your intelligence and/or execution against your opponent’s. The reason this is engaging is because of the imperfections people have. You don’t know how smart or skilled your opponent is until you fight them. If you know the difference, the gap between you and them can be exploited. If it was just two TAS bots fighting against each other, the match would effectively be decided before it began. That’s why I value mind games quite a lot. There isn’t really a “peak” of intelligence you can get, it’s just about play and counter-play. This usually results in the person with the most planning and knowledge to be the victor, rather than the person who could just twitch their fingers the best.

    • The Ludite December 29, 2016 / 9:24 pm

      If you like mind games and bluffing games, I have good news for you: almost every game, no matter what else is going on, eventually breaks down into rock paper scissors. A game that does not will instead break down into Trivial Pursuit or darts. In fact, all games are on some level a combination of these. This is pretty well understood at this point, and is the source of untold amounts of existential hand-wringing by anxious designers. There are even people who laud the rock paper scissors, indeed bask in it, declaring it to be not a problem at all, but rather the solution!

      However there’s a reason I chose the phrase “break down” into rock paper scissors – because it is, in fact, the game breaking. All the things the game was previously about are no longer true. Playing chess is a game of controlling the board with an economy of moving individual pieces. But as solutions start to emerge, the game starts getting chunked into less and less actual decisions. Chess openings are the classic example of this – many moves go by with no decisions being made on either side, so the whole thing is practically one move. Eventually these chess chunks get so big that there isn’t a whole lot of game left, especially in comparison to what it looks like when beginners play.

      And when these chunks start forming natural counters to each other, it starts to look more and more like rock paper scissors. The horror!

      There’s a concept in game theory known as Nash equilibrium. It’s basically a decision matrix solved down to a few optimal moves, each adjusted for their success rate. In a game where you make only one move, the optimal one is the highest win-rate. In a game where you make this choice of moves many times, the success rates of each move become what we call a “mixed solution,” where you employ all the moves at rates that correspond to their success-rates. A great example of the latter is Street Fighter. An even better one is David Sirlin’s Yomi, a card-based Street Fighter game where not even the execution aspect can alter the mixed solution.

      Wait a minute! “Solution?” I know that word – that’s the word for when the game isn’t fun anymore!

      When everyone falls into the same old habits, the same few moves, the game gets stale and we get restless. We want something novel to happen! But the Nash equilibrium calls, and any deviation from it is to our detriment. Sometimes we get excited when it just so happens that two people’s decisions lined up with one person landing a really devastating move, but it wasn’t any different from any other moment. Two people continued to cycle through their NE percentages, each in their semi-random way. When the payoffs for any given move are widely varied, you’re bound to have a big moment where one person’s big payoff slips through.

      None of that changes the fact that the game is largely, if not entirely solved, and people are just going through the motions. Despite Sirlin’s assertion that Yomi wouldn’t be solved in our lifetime, players did solve it. Then they went on to break the game entirely by playing a pure NE mixed solution using an RNG to determine the cycle of their moves! These unreadable players were so unbeatable that they forced everyone to do it and defeated the spirit of the game. The whole thing even resulted in Sirlin banning players from the game, claiming that using an RNG was cheating.

      But the problem here wasn’t players breaking the game with RNGs – the problem was that the game was far too reliant on rock-paper-scissors mind-games in the first place. Rock-paper-scissors isn’t a “safe” design, as Sirlin claims. It is in fact unavoidable. And using it as a safety net for the game, or even as the primary content of the game as Yomi did, is only going to trick you into designing a weak system that falls right into a solution. Rock-paper-scissors isn’t something we should be celebrating either, as Frank Lantz claims. It’s a solution state that a designer should try to hold off for as long as possible!

      We love games, though, and so in our attempt to rationalize our love, we embraced rock paper scissors because that’s a shocking amount of content in so many games. It doesn’t have to be, though. Just think about Dotalikes – the concept of the “lane” is sort of a counter to the formulation of Nash equilibrium. NE matrices are essentially isolated guesses of what the other guy is going to do and what the counter ought to be. But Dota lanes remove a lot of the guesswork because you do get to see your opponent. Street Fighter is kind of like a version of Dota that’s all jungle and you could run into anyone at any time or even the whole team, and the winner is the one with the most kills. Think about how dumb that would be, how much of the systemic strategy completely evaporates. Now put lanes and objectives back in – even though you can see what your opponent is doing, the game isn’t obvious and solved. You can’t fit Dota onto a little 4-square decision matrix.

      Single-player games and multi-player games are far more alike than we like to admit. Multi-player games tend to focus on this “reading of the opponent” thing very heavily, and as a result the systems tend to suffer. Ultimately both single-player and multi-player games test your knowledge and (more importantly) judgement, but we think they’re different because they don’t look the same when they reach their solution states. The single-player game usually ends abruptly while the multi-player game stretches out playability with the unpredictability of a human opponent. Of course some single-player games also add unpredictability in the form of procedural generation, aka randomness, which stretches out its playability. Which means that adding a human opponent is actually the equivalent of adding randomness and… head explodes

      To sum up, any time I start to run up against Nash equilibrium being the way forward, I drop the game. There’s no need to play a dead game now that I’ve already extracted all the fun content out of the game, and I’m definitely not interested in playing rock-paper-scissors in any of its thousands of forms. There’s always a new game to play.

      Thanks for reading!

  4. Juho December 10, 2016 / 5:40 am

    Digging deeper into the Dungeon of the Endless example would actually reveal something interesting. Everything about the game is telling you that it’s about two things: resource management, and strategic positioning. All you do is decide what to build in which room, and which rooms to move your heroes to. And the game is balanced such that it’s possible to win the game by just making decisions at that level.

    But turns out there’s tons of room for effective micromanagement. Even though you can’t directly control the location of your heroes inside a room, you can reposition them by ordering them to another room and canceling. This can be used to kite monsters around, making your defensive buildings a lot more effective. Equipment can be moved between characters instantly even if they’re in different parts of the map. So when fighting a two front battle, you can constantly rejig the equipment between the two groups. When low on dust, you can build defenses in rooms unpowered. Wait for a wave to start, then power down some rooms on the other side of the map and temporarily power up the defensive rooms just for the duration of the wave. And probably more that I haven’t discovered.

    I can’t believe any of this micro was intentional design, but there it still is. The players will always find a way to extract more skill from the game, no matter how tedious it is.

    • The Ludite December 10, 2016 / 6:47 pm

      The difference between this kind of micro and the micro you see in other kinds of games is the input resolution. There’s a lot of little ways to play with DotE’s system, many little decisions that can tip the scales one way or the other. But compare that with the kind of micromanagement you need in a game where you control one character, something like Binding of Isaac. If you push no buttons, your character literally does nothing. It’s not uncommon to switch movement and shooting directions faster than once per second, especially as your speed goes up. And that’s nothing compared to the input resolution of something like Super Smash Bros. Melee, where displaying inputs visually reveals a stream of button pushes rivaling the APM of any Starcraft player.

      It’s useful to consider the input resolution of a game in terms of “necessary button pushes per decision.” Most “control one character” games require input CONSTANTLY. Even the decision to move in a direction requires a continual input – if you stop pushing a button, everything halts. When you make the player responsible for every last little action a character takes, the density of player inputs starts to really stack up. Fine motor-control makes all kinds of complicated movements possible, and before you know it, the player and the designer are in kind of an arms race – the player keeps meeting the execution bar, and the designer keeps raising it. This design paradigm happens so readily because it’s super easy to make little sprites colliding on a screen, and only slightly harder to come up with weird ways for them to move. And it feels really right because the player is constantly engaged with movement buttons, and that’s good right?

      DotE’s input resolution is comparatively low. Maybe you decided to switch up the battle lines and move four characters around, requiring four keystrokes and four clicks. There’s nothing requiring you to even WATCH the decision play out – indeed it’s often unnecessary. Some of the micromanagement you describe might require more button presses than simply watching a wave play out, but the decision density rose to match. The game only asks you to click when a new thing has to happen, when a change in your strategy has occurred. The closest that DotE gets to the input resolution of more fiddly games is the monster-kiting trick you mentioned (I agree, it seems unintentional). But even that is toned down by a simple pause button. Indeed, the pause button goes a long way towards reducing the input requirements in general. The element of “timing” a weapon swap is trivialized by a thumb on the space bar – it doesn’t require good reflexes even.

      The micromangement of DotE may in some cases be tedious, and that’s maybe its own separate issue. But it’s rarely close to the levels of input resolution other games demand.

      Also, let it not be said that all “control one character” games are inherently going to be about aiming and dodging. In a delightful subversion of this stereotype, Pocketwatch Games is making a very intriguing game called Tooth & Tail. It’s an RTS game where you control only one character: a field general who leads troops into battle by issuing commands in an area. This one simple limitation has all kinds of ramifications on the ways you interact with it: to interact with your base, you have to walk back to it, making it a strategic consideration. Troops will only go where you are currently standing, and enemy players who don’t see you know you can’t make them advance. Most importantly, it takes a genre famous for its extremely high input density and puts a harsh cap on it, making it accessible for the rest of us. And the way it does that is so sensible and easy to swallow! Here, it’s actually the avatar-based game that’s easier to operate than the arm-chair general game.

      So really the issue is less about micromanagement and more about keeping a tight reign on the player’s potential to convert input density into power.

      Thanks for reading!

Comment on this article

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s