Case Study: Ezreal in League of Legends

So this isn’t so much of a case study as it is a fantasy changelist, but oh well. Ezreal is my favorite champion by far, but if I had a chance to rework him, there are plenty of things I would want to try.

He’s one of those weird cases where he can become OP simply by existing. Look at Ezreal’s history and you’ll see that he has been buffed relatively few times compared to the number of times he has completely changed the meta. Remember when people started maxing W first for the attack speed debuff? Or remember how Ezreal pretty much single-handedly nerfed the Iceborn Gauntlet and the Runeglaive for ranged champions? In all of these cases, nothing happened to Ezreal to make him stronger, people just found new ways to use him. Who’s to say that there won’t be even more new ways to use him popping up as the game evolves? It’s not sustainable to just reach for the nerf hammer every time something new comes up. Out of all the champions, this phenomenon seems to happen most often with Ezreal.

Then there’s the whole issue with AD vs. AP Ezreal. If you go AD, your W is almost useless, and if you go AP, your passive is almost useless. Both versions of Ezreal are essentially balanced around the fact that they only have 4.5 abilities rather than 5. Right now, Riot is trying to encourage more aggressive Arcane Shifts by adding an AD ratio, but I feel that Arcane Shift is really powerful as an offensive AP spell if you build Ezreal as a mage. What if the next Ezreal meta becomes full AP and max E first in mid? It sounds silly, but remember that it also sounded silly to max W first before the Koreans started doing it.

But putting balance and game health aside, Ezreal has a lot of untapped potential to fill a unique role in the League: an explorer. There isn’t really an explorer character in the champion roster. I’m talking about a pathfinder, a scout, a ranger who clears the way head first. Teemo is kind of like an explorer, but he is what I would call a passive explorer: he places traps and watches them like wards. I think Ezreal could be an active explorer, someone who gets his hands dirty and puts himself in risky situations because he can.

The Core Theme: Ezreal’s Identity

In keeping with my three-part system of game design, we start by identifying the theme. So who exactly is Ezreal? What’s he like? How can his personality translate into his mechanics?

Ezreal is cocky and arrogant, and he likes having situations under control. He’s most comfortable when he’s setting his own pace, and when he’s in his element he thinks he’s invincible. But if something unexpected throws his rhythm off balance, he panics and retreats, looking for an opportunity to try again.

Ezreal is a loner. Other people slow him down, and he sees cooperation as babysitting. He’ll work together with other people if he really has to, but he’s more comfortable launching a small precision strike by himself rather than joining a coordinated siege. When other people help him, he takes it grudgingly: he thinks he would have been just fine by himself.

Ezreal is selective and picky about his targets. When he fights, he loves intense duels and skirmishes. Nothing makes him happier than a true display of skill. Enemies who rely on their teammates, towers, or minion waves just seem cowardly. Summoner spells, monster buffs, and level/gold advantages are all external factors, and Ezreal hates external factors.

Ezreal likes to think of himself as omniscient. He is aware of everything that happens around him. If something hostile is nearby, he’s forming a plan to deal with them before they even reach him. When he’s being suspicious and careful, it’s difficult to surprise him, and it’s nigh impossible to hide from him. But he has tunnel vision, and when he focuses too strongly on his target, he loses sight of other things he should be paying attention to.

The Unified Elements: Ezreal’s Skills

Again, I’m not a god of game design, so I’m going to try to avoid numbers wherever possible. I can’t come up with balance off the top of my head. But hopefully, these reworked skill suggestions point towards a direction that helps support what I think Ezreal should become.

Rising Spell Force: Reworked. Ezreal gains a stack of Rising Spell Force whenever he hits an enemy champion with a basic ability (max 1 stack per spell). When he reaches three stacks, he enters Rising Spell Force mode, where he gains 50% attack speed, his basic abilities refund half of their mana cost when they hit an enemy champion, and his basic abilities gain new bonuses. Rising Spell Force mode lasts as long as he has three stacks, and its duration is refreshed if Ezreal gains another stack while it’s active.

Mystic Shot: Basic mechanics remain unchanged. Mystic Shot no longer applies on-hit effects, but has higher AD and AP ratios. It has a higher base cooldown, but if Ezreal lands a Mystic Shot on an enemy champion while in Rising Spell Force mode, it reduces its own cooldown by an additional amount (important: only its own cooldown).

Essence Flux: Completely reworked. No more attack speed buff, lower cooldown, higher mana cost. It now acts as an instant ground line cast. Ezreal goes into a short casting animation, the line indicator is drawn in front of him, then after a brief delay he zaps everything along the line. Hitbox and behavior is about the same as a fully-charged Arcanopulse from Xerath. Damage is lowered drastically, I’m thinking something like 20/50/80/110/150 (0.5 AP). When Ezreal casts this, he gains vision along the line during the casting animation, and if he hits an enemy, they are revealed for a brief period of time (even if invisible). If Ezreal is in Rising Spell Force mode, Essence Flux does a large chunk of additional damage, I’m thinking something like 100/150/200/250/300 (1 AP).

Arcane Shift: Basic mechanics remain unchanged. If Ezreal is in Rising Spell Force mode, the arrow from Arcane Shift slows its target by 20% for 2 seconds.

Trueshot Barrage: Basic mechanics remain unchanged. Mystic Shot no longer reduces Trueshot Barrage’s cooldown. No longer generates Rising Spell Force stacks.

So what does this do? First of all, it decentralizes Ezreal’s power in Mystic Shot. Too much of Ezreal’s strength lies in Mystic Shot, which is a good thing and a bad thing. It’s a cool spell and it’s fun to use, but there’s a point where you’re going too far. Removing the on-hit effects on Mystic Shot is a really big nerf, and it’s alleviated a little bit by slight buffs to AD/AP ratios, but his power level is still lowered. Next, increasing Mystic Shot’s cooldown propagates a cooldown nerf on all his other abilities, because of how it reduces cooldowns when it lands.

In exchange, a lot of power is stuffed into Rising Spell Force mode. He is very weak outside of Rising Spell Force mode, but very strong when it’s active. This represents how Ezreal needs to build up his power through a prolonged display of skill (landing successive shots on the enemy). The three-hit mechanic generates a lot of interesting play: do you use three Mystic Shots over a long period of time, or do you unload all of your basic abilities at once to trigger RSF mode immediately? Note that Rising Spell Force stacks are much harder to gain now, since you only get them when you hit an enemy champion (and there’s an added limit of one stack per spell). Also, there’s no spectrum anymore, his passive is all or nothing.

RSF-enhanced Mystic Shot brings back the URF feeling of machine gun shots, but only if you can keep hitting your target. Even though it only reduces its own cooldown by a greater amount, that still means you can reduce the cooldowns of your other basic abilities faster by using more Mystic Shots. On the other hand, RSF-enhanced Essence Flux goes in the opposite direction that incentivizes players to be more accurate and deliberate in order to cash in on a more damaging payout. Finally, RSF-enhanced Arcane Shift is just a small buff to help Ezreal when he’s in the zone.

Trueshot Barrage is nerfed by no longer having its cooldown reduced by Mystic Shot. The enhanced Essence Flux is meant to act as Ezreal’s primary finisher, replacing Trueshot Barrage’s purpose in duels. With this, Trueshot Barrage is almost completely decentralized from the rest of his kit, which encourages him to save it for situations that are outside of Essence Flux’s reach.

The Ensured Delivery: Ezreal’s Dynamics

This rework would demand a lot of skills that people don’t usually associate with Ezreal. People are used to just forgetting about his passive and his Essence Flux, but now they are the most important parts of his kit. He will feel very weak, especially without the on-hits on his Mystic Shot. Ezreal is known for chopping out chunks of health with Mystic Shot, but a lot of that damage comes from Triforce or Lich Bane.

But under this system, Ezreal takes out larger chunks of health with RSF-enhanced Essence Fluxes. The Mystic Shots do some damage, but mostly they act as indicators to show how Ezreal is building up his power. He floats like a butterfly and stings like a bee depending on whether he’s in RSF mode or not.

Mystic Shot and Essence Flux serve two different purposes: damage and utility. Outside of RSF mode, Mystic Shot is for damage and Essence Flux is for utility. But when RSF mode becomes active, it suddenly switches. Mystic Shot gains greater utility as a tool for maintaining RSF mode because of its self-cooldown reduction, and Essence Flux gains a massive damage boost. Ezreal tracks enemies down with Essence Flux and whittles away at them with Mystic Shots until he enters RSF mode, at which point he uses Essence Fluxes for nukes and Mystic Shots to keep track of his prey. This creates an interesting dynamic because Mystic Shot is nowhere near as good at tracking targets as Essence Flux is, but you have to get good at predicting enemy movement with Mystic Shots in order to keep Essence Flux ready for a nuke, because if you use Essence Flux for tracking you will lose its potential for damage. Skilled Ezreal players already use Mystic Shot to check brushes.

Ezreal’s AP and AD builds are also separated distinctly. AD Ezreal is the safer build, the one with consistent damage. If you can’t enter RSF mode, it’s still okay because Mystic Shot’s damage doesn’t depend on it. Essence Flux and Arcane Shift are used primarily for utility and scouting, and autoattacks still become his primary source of damage. AP Ezreal is the risky build that relies heavily on RSF-enhanced Essence Fluxes for damage, and Mystic Shot just becomes a tool to enter RSF mode and reduce Essence Flux’s cooldown.

He’s not comfortable with direct confrontations, but he’s strongest in jungle skirmishes where neither combatant knows where the other is. In those situations, he probes with Mystic Shots and Essence Fluxes until he enters RSF mode, at which point he starts using Arcane Shift more aggressively to line up an enhanced Essence Flux finisher. If the battle doesn’t go his way, he probably doesn’t achieve RSF mode, and instead Arcane Shifts to safety while using the vision from Essence Flux to make sure his escape route is clear.

In teamfights, Ezreal either tries to focus down a single target with Mystic Shots and autoattacks, or he lines up enemies for Essence Fluxes and Trueshot Barrages. He’s not a focused killer like Vayne, nor is he a widespread damage dealer like Miss Fortune, but he’s flexible enough to switch between them on the fly. But rather than being in the heat of teamfights, this Ezreal is much, much better during the start or end of one. Before a teamfight starts, he can scout ahead so he and his allies knows where the enemies are with consistent Essence Flux vision, making sure that they are never surprised. After a teamfight ends, he takes a pursuit role and hunts stragglers down with precise Essence Fluxes enhanced by RSF.

Principle of Charity

In keeping with the principle of charity, I have to ask why this hasn’t actually happened. This is the part where I tear my own suggestions to shreds, so I might not be the best person for this, but that’s the nature of criticism.

It’s really confusing and difficult to keep track of abilities that change uses. Abilities in League don’t usually change so dynamically between vision tools and damage nukes the way that my proposed Essence Flux does. Usually, abilities are straightforward and clear, and always do the same thing no matter what the context is. If you want to have an ability that does multiple different things, you split it up into two separate abilities and use some kind of transformation (Nidalee, Gnar, Jayce). However, my proposed changes are a lot more subtle than full-on transformations, so they force players to reevaluate their abilities in the heat of battle. Not a very easy thing to do.

Small bursts of vision have historically not been very well received. Just look at how the Warding Totem is so much more popular than the Scrying Orb. My proposed changes to Ezreal revolve around those small lines of vision from Essence Flux, but practically, they’re kind of difficult to work with. Vision in general is a source of power that isn’t very easily grokkable, in that it’s hard to know when you played your vision right. Long-lasting stationary vision just feels better than long-ranged temporary vision, and I don’t think balance can change that. Branching off of that train of thought, precisely aimed vision is even more troublesome. The new Essence Flux is basically like Ashe’s Hawkshot with a much shorter range, much shorter vision duration, and a much shorter cooldown.

Two long-ranged skillshots are kind of overkill and give too much zone control, especially when one of them (Essence Flux) doesn’t stop on collision. Currently, this is balanced by the fact that Essence Flux is useless on AD Ezreal, and it’s one of the reasons why AP Ezreal is so strong in the late game. Right now, my proposed changes mean that non-RSF Ezreal still gets a little bit of utility from Essence Flux, and RSF-mode Ezreal is basically as strong as late game AP Ezreal.

Finally, it’s very difficult for this Ezreal to fit inside a team composition. He’s not a strong sieger, he’s not a strong assassin, he’s not a strong split pusher. He is a strong explorer, and I’m not entirely sure when a team would want an explorer, at least in the current meta. In fact, I would see this Ezreal as a strong counter jungler, and he could possibly be pushed in the direction of an ADC jungler like Kindred (maybe he can charge Rising Spell Force on monster hits?). When you think about it, he’s not the type of guy to stay put in bottom lane anyway.

Case Study: FFXIII System Easing

Japanese AAA games seem to have a penchant for convoluted, complex combat systems. I mean, they all have such ridiculous names. FFXIII‘s “command synergy battle.” TWEWY‘s “stride cross battle.” Tales of the Abyss‘s “flex range linear motion battle system.” Don’t get me started on Kingdom Heart’s real-time menu.

It’s interesting because American/European AAA games seem to be leaning towards minimalism and simplicity. The freeflow combat pioneered by the Batman Arkham series has been coming through in many other titles like Shadow of Mordor or Mad Max because it’s so simple and intuitive for the player. Movement in games like Mirror’s Edge or Assassin’s Creed is often reduced to a very few amount of buttons that cover many different actions depending on the context. When you think minimalism, you still think indie, but it’s leaking a little bit into modern AAA titles.

But this article isn’t about the differences between Asian design philosophy and American/European design philosophy (which is probably an article I should get around to sometime). I may not personally agree with the Asian penchant for overly complicated control schemes, but I do have to agree that they’ve refined their craft to a point.

There was a certain boss fight in FFXIII that struck me as very well-designed, particularly in regards to how they used it to ease the player into the game’s more advanced nuances. This was the boss fight (or rather, series of boss fights) against the Ushumgal Subjugator. Apparently, they like to name their bosses as strangely as they like to name their battle systems.

For a little bit more context about FFXIII’s “command synergy battle,” check out my post on the mana system paradox.

Anyway, FFXIII’s combat is just very weird. You have three people in a party, but you only directly control one. As for the others, you can assign them roles and they’ll automatically do things based on their role. A healer will heal, a fighter will fight, a defender will defend, so on and so forth. So you get a really weird situation where the player isn’t sure where they should put their focus. On one hand, you can control your one main character, but on the other hand, enemies are also moving and attacking in real-time, and on the third hand (???) your ALLIES are also moving and attacking in real time. It’s a lot to take in at once.

For the first half of the game, you never really have any extended periods of gameplay where you have a full party of three people. Instead, everyone splits up into pairs and wanders around as separate groups, then they all come together and you can swap party members out. This helps reduce the burden, because a party of two is easier to manage than a party of three. Not only that, but the game is very careful to make sure that the two people who are paired together can work together and cover each other’s flaws. There will never, ever be a situation where your team doesn’t have a healer, for instance.

But that means that there’s a very interesting design challenge: how do you ease players from controlling a party of two, to a party of three? The game continues in groups of pairs for a very significant amount of time, and the player has probably developed habits and tactics revolving around their two characters. Fortunately, the Ushumgal Subjugator boss fight does it in a very smooth transition that trains players to rethink their strategies and feel really powerful while they’re doing it.

The Ushumgal Subjugator has three stages: an aerial form and a ground form that is fought over two battles. When you first fight the aerial form, you’re doing it with a party of two, just like any other boss fight. However, it helps prime the player for the upcoming shift by introducing a new mechanic: knockups. Periodically, the Ushumgal Subjugator will use an attack that launches any characters hit up in the air, which renders them completely unable to take any actions until they land and recover. Here, you can see an obvious connection: if a character has been knocked airborne, you have a few seconds where you don’t need to worry about controlling them. During the first boss fight with two people, it’s a major setback, but when you have a party of three, you can afford to lose someone for a moment, and it actually helps ease the transition a little.

Otherwise, the first Ushumgal Subjugator boss fight isn’t particularly interesting or noteworthy. The player defeats it quickly (it’s meant to be an easy boss) and moves on. But then it shows up again during a sob story cutscene, and one of the characters is too wounded to fight. Now, suddenly the tables are turned: for the first (and I think only) time in the entire game, you fight against a boss with only one character. This second fight is scripted, and the boss does nothing interesting except keep attacking. Whether you run out of health or survive for long enough, two allies run in to join the fight and the third stage starts.

Now, the player controls a full party of three against a boss. It’s really sudden, but there are a lot of subtle mechanics in play to make sure that the transition goes as smoothly as it can. First, two of the three characters can be healers, which means that there’s always an opportunity to restore health. Second, the boss uses many light attacks that damage everyone on the team, which means that healers will always be topping off their allies: there are almost no times when a healer is useless. The player is most likely using some combination of two attackers and one supporter, so that the two attackers can be thought of as a single entity and the supporter is constantly topping everyone off from the boss’s multi-target attacks.

Third, and most interestingly, the boss uses a variation of the lockdown mechanic that was seen earlier during its first stage. It doesn’t knock characters airborne, but instead, it will select a single character and launch a series of concentrated attacks against that character. This is telegraphed ahead of time (the boss will say that it is targeting someone) so the player has plenty of time to switch their playing style. Now that the boss is focusing attacks on a single target rather than multiple ones, the player needs to take healing and defense more seriously: topping everyone off evenly isn’t viable anymore.

Whether the boss is doing light damage to everyone or heavy damage to a single attack, it makes the appropriate counterattack obvious and accessible. The player doesn’t need to think of their team as three separate characters, but rather in two categories: high offense, or high defense. They can put three characters on offense, or two on offense and one on defense, or two on defense and one in offense, or three in defense. No other mechanics are necessary for this immediate boss fight.

This is interesting because with the parties of two that the player has been controlling up until now, there were many more mechanics in play than offense or defense. They had to worry about buffing allies, or debuffing enemies, or splitting their attacks across multiple enemies, or removing status effects, or all sorts of other things. But the boss fight against the Ushumgal Subjugator involves none of that. It is literally just offense or defense, and the player allocates their resources on a slider across that spectrum. As the player progresses, they slowly have to relearn all of these mechanics with three people instead of two. The Ushumgal Subjugator bossfight acts almost like a hard reset: it brings the player back to a simpler time, before they need to worry about invulnerability windows or elemental typings.

I think the fight could have been improved if the Ushumgal Subjugator’s knockup attack was brought back for the third stage. For example, when the Subjugator locks onto a single target, it should do something to suppress that target for an extended period of time, and in exchange do less damage to that target. That way, it gives the player an opportunity to rethink the combat scenario with two characters instead of three. This would also be empowering because the player still has access to a wide range of actions while a single member of their party is suppressed, as opposed to before when they only had parties of two people.

This boss fight is a really good way of introducing complex game mechanics through play, which is a good thing. Many games will introduce complex game mechanics by adding more mechanics on top, which is a little counterintuitive but happens more often than you’d think. For example, in Lethal League, there’s a whole world of parries and special attacks that form a precarious system of checks and balances. When you hit a ball, there’s a special move you can do to make sure that someone else can’t hit the ball out of your hands, and that special move feels like a bandaid design solution. It’s solving a problem by adding a new mechanic, and maybe for them it was inevitable, but in general I think a designer should hesitate to add new mechanics.

And that’s not to say that FFXIII doesn’t have a truckload of extraneous mechanics, because it does. Upgradeable weapons, tactical points, full ATB attacks, the list goes on. But the Ushumgal Subjugator boss fight is a nice exception, because it doesn’t add anything new to confuse the player. Instead, it gives the player all the exact same mechanics they had gotten used to previously, and throws a third party member into the mix while keeping the overall tension level low enough that the player can understand everything that’s going on. I think that’s good design and it’s something to be learned from.

Case Study: Emotion In Cave Story

Stab me with a pitchfork, but I’m really not on board with the whole nostalgia trip that’s been taking over the indie world recently. Too often, it just feels contrived and meaningless. If you want to use pixel graphics, cool, but is there any real reason why pixel graphics are good for the game, or are you just using them because you feel like it? That’s not a valid design reason for anything. There are too many games that are just nostalgic for nostalgia’s sake, which I guess is a potential target audience, but design decisions should be made for meaningful reasons within the product’s core theme.

That’s why I like Cave Story. It’s one of those nostalgia games, and usually I would just groan and roll my eyes at the pixel graphics. And that’s exactly what I did, up until one specific moment in the game that made me go “holy shit,” so that’s what I’m gonna talk about now. I started off thinking that this was gonna be a level design analysis, but this specific moment in Cave Story is much more than just level design. It’s aesthetics, it’s expectations, it’s controls, all leading to one giant punch in the gut. So, get ready for spoilers, because there’s gonna be a lot.

So Cave Story is a story about a cave. “Wow, brilliant observation skills Kenneth, you get a Nobel Prize in title deciphering.” Alright, more specifically, Cave Story is a story about a dude who wakes up in a cave. You play as this dude and you have adventures in the cave and you meet rabbit people and you get guns and you shoot things.

There’s a big giant evil dude who you have to kill to save the world, of course. It’s not a nostalgia game for nothing (although that’s not really a valid thing to say, because the “kill the dude to save the world” trope isn’t just nostalgia, it’s everywhere). But I’m not gonna talk about that. I’m gonna talk about the most strangely named character you will ever cry over: Curly Brace {?}.

Curly Brace is a robot, like you (first spoiler alert). You both have guns and you both kill things. The first time you meet Curly Brace, she’s protecting a small group of rabbit people and she thinks you’re a murderer trying to kill them, so she fights you. Later, you patch things up and become best robot buddies.

The moment I keep referring to starts with a completely unprecedented event: Curly Brace comes along with you on your way to fight some big evil boss. This is the first time in the game you have ever fought alongside a partner, and it’s empowering. Curly follows you around and she has a machine gun and she shoots everything. Her AI is a little wonky, but it’s okay, she doesn’t take damage. You and Curly have a brief segment where you fight through a bunch of bugs and things on your way to the boss.

During this segment, you notice something: Curly is immune to water. There are little puddles of water interspersed throughout the environment, and when she invariably steps into one (because of her wonky AI), a bubble surrounds her like a shield protecting her from water. This is new and interesting, because you have no such bubble. In fact, water is a very dangerous threat to you. When you enter water, an air meter starts counting down from 100, and if it reaches 0 you die. So it feels kind of unfair that Curly Brace is completely immune to water, right? But you can chalk it up to retro-style game design: she has the bubble because her AI couldn’t be improved. It’s just a lazy workaround that the game developer put in so people don’t whine about Curly being able to stay underwater: she can do it because she has a deus ex machina bubble shield you’ve never seen before, so deal with it.

Alright, you shrug and move on (or you ragequit, which you probably don’t). You go through the area together with Curly and her magical bubble, fighting enemies in water and solving water puzzles and doing so many water things that you’d think you were at a waterpark. Eventually, you reach the core of the floating island (second spoiler alert, the cave is a floating island). Then you start fighting the core for some reason I can’t remember. But the interesting thing about this boss fight is that it uses water in a way you’ve never seen before. The boss is in a pretty large chamber, but it’s a fully enclosed chamber, and for its attacks the core will sometimes flood the whole chamber. Since you can’t survive very long in water, you need to reach high ground and hold out until the waters recede, but sometimes the flood will last so long that you will die no matter what. When your air reaches 0, you die instantly, you don’t start losing health gradually like you might come to expect, so it gets frustrating. And while you’re being frustrated, Curly Brace is sitting there with her machine gun and her bubble shield. I don’t know if the designer’s intentionally trying to make you feel jealous of Curly’s bubble, but you definitely end up that way.

The boss’s flood attack, and the player’s dwindling air meter.

Then you defeat the boss. Once the boss is dead, something crazy happens: the whole room gets flooded. Every inch of the room, and there’s no sign that the flood is gonna recede like it did during the boss fight. You try to go back the way you came, but the door is shut tight. You go to the opposite side of the room but there’s nothing. You try to jump up, but there’s no air left. All you can do is fall down and watch your air go to zero. It’s disempowering and frustrating, and the whole time Curly is just standing there in her little bubble. Your air runs out, the screen fades to black, you’re notified that you can’t breathe anymore.

But the game doesn’t end. A text box pops up: “……? You can breathe.” Huh? What? How? The screen fades back into the game view, and you see Curly lying lifeless on the floor next to you in the flooded room. Her bubble shield is now centered around you. Even though all the water’s still there, the air meter is missing. Curly gave you her bubble shield so that you could live on.

Everything up until now has been leading up to this point. All the water, all the puzzles, all the boss attacks, everything revolved around water. And the whole time, you couldn’t help but be a little jealous of Curly’s bubble shield. It was easy to rationalize it because the whole game has a nostalgia tint around it, so you can just say that the bubble shield is a cheap workaround solution so that Curly’s AI didn’t have to take water into effect. But it wasn’t actually like that: the bubble shield was a real in-game element that she just gave to you in exchange for her own life. I don’t know about anyone else, but I felt pretty horrible about it. Here I was, blaming Curly for having a bubble shield purely because she was a dumb little AI companion. When I was treading water and managing air, I casually thought to myself that it would be nice if I had a shield like hers, and then she gave it to me.

Now that you have the air bubble, you continue on through your journey. You travel through aqueducts and you flow through underwater currents and you dodge submerged spikes and you battle giant fish. The whole time, there’s no air meter anymore, because you have the bubble. Even after the boss battle, the very next area you have to move through just continues to reinforce the water/theme and accentuate Curly’s sacrifice. For the rest of the game, you never have to worry about water ever again.

I don’t know if this is the same reaction that the average gamer will have when playing Cave Story. For me, most of this was because I had a special eye out on the nostalgia bias, and I was deconstructing the rationale behind Curly’s AI as I was playing. But there are so many level design elements at play to draw your attention to Curly’s shield. Everything is there to increase the tension at the moment of the final flood, right before she gives you the bubble.

Then, in true old-timey-JRPG fashion, there’s an obscure way to save Curly that you’ll never find out unless you read about it online. I am definitely not a fan of that design style, I think all options should be fairly presented as choices with consequences and that the game should revolve around what the player does with those choices rather than how the player finds out about those choices. Cave Story isn’t a perfect game by Chen-ian standards, but I think it’s good to break down and analyze specific pieces of games and appreciate how they worked in the larger context. There are tidbits of good design in bad games, and there are lessons of things gone wrong in good games. That’s why I never do game reviews, I do case studies.

And Cave Story is a nice game to study. It’s free, it can be completed in one sitting, and it’s got plenty of variety built into its mechanics. But more than anything, it’s one of the few games (rather, one of the few ANYTHINGS) that has made me feel sad over a character’s death. Level design is typically thought of in terms of mechanics and skill curves and learning over time, but it’s also important to recognize that level design can also carry emotions and feelings. As I get better at level design and design in general, I hope to use my skills to give players good experiences, both mechanically and emotionally.

Case Study: Advanced Warfare Looks Promising

I’ve already said before that I’m not interested in doing straight game reviews. That goes for game previews too. So it’s pretty rare for me to be writing about my thoughts on Call of Duty: Advanced Warfare. Hell, I’m not even a Call of Duty player in the first place. And considering how much bad rep the CoD series gets from the design community, it’s really weird for me to be paying so much attention to this one. But I’m really excited about this game from a design perspective, and now I’m going to break down why.

The reason why I heard about CoD:AW in the first place was because I’m a big fan of Titanfall, and I was reading a lot of internet rage about how CoD:AW was just a copycat. And if you watch some of the preliminary trailers, it really seems that way. They both have jetpacks, they both have cloak abilities, they both have lateral dashes. They both even use the name “Atlas” as a major pivot point (in Titanfall it’s a well-rounded Titan model, and in CoD:AW it’s the name of a giant world-dominating PMC).

But as a game designer, I don’t think it’s good to criticize games for copying mechanics. A lot of industry people always say “ideas don’t matter, execution is everything,” so if a game steals a mechanic and executes it brilliantly, it’s better than a game with a radical idea that fails on execution. For me, I take a different approach: I think copying mechanics is fine as long as it connects to the game’s core theme. My whole crusade against mana has been about the overuse of resource systems in games where they don’t belong. So if another game has done something that really fits into the core theme of what you’re making, I say go for it.

With CoD specifically, that kind of philosophy gets really muddy, because CoD‘s problem has always been a lack of a core theme. I think it’s really important for a game to have a solid core theme that summarizes everything the game is trying to deliver in a short phrase. Once that’s done, as you continue through development you can check every element and see whether or not it helps your core theme. But CoD as a series has always been criticized for just haphazardly dumping random elements into their games for the sake of spectacle. Remember all the controversy about their airport level? If it was a significant, meaningful part of the game’s intended experience, that level could have been a Spec Ops kind of deal, and it would have reached out to gamers in a way they haven’t seen before. Considering all the public opinion backlash we see about it now, that obviously didn’t happen. The other CoD games are filled with similar spectacle moments that never consolidate into a solid takeaway.

On the other hand, Titanfall, the game everyone says CoD:AW is copying, actually has a really solid core theme. Their core theme is “scale, verticality, and story,” and every single Respawn developer interview always includes one or more of those words. Every mechanic in Titanfall plays a part in delivering scale, verticality, and/or story. The double jump gives pilots the ability to traverse terrain on a greater scale. The wallrunning and walljumping allows pilots to wield verticality in their combat engagements. The titans break up the flow of a match and introduce “oh shit!” moments to turn every battle into a dynamic story. Everything is about scale, verticality, and story.

So if CoD:AW was trying to mimic Titanfall exactly, their core theme would be “scale, verticality, and story” too, which would be cool but not interesting enough to get my attention. But CoD:AW is actually lifting game mechanics from Titanfall and applying them to a different core theme, which really got me paying attention. It’s risky because these games were developed with different experiences in mind, and it would be a lot easier for CoD:AW to just copy everything about Titanfall down to a T, including their core theme. Since they’re not doing that, it really seems like this is the first time in recent history that a CoD game gets designed around its own unique core theme.

For CoD:AW, their core theme is “power changes everything.” And it really feels like they’re intentionally orienting their game around this theme. One of their trailers opens up with a quote from Abraham Lincoln: “Nearly all men can handle adversity, but if you want to test a man’s character, give him power.” From the story to the mechanics to the aesthetics, everything is about what happens when you’re given more power than you can handle. A test of your character as a player, and as a person.

One thing I’m really interested in is CoD:AW‘s battery powered weapons. The idea is that they regenerate power over time, and power is essentially their version of ammunition. That means these guns can potentially have unlimited firepower, but you can’t expend too much at once. This is really exciting for me personally, because I’ve written about resource systems in my mana essays and it sounds like the battery weapons are a way of getting around that problem. But more importantly, the battery weapons are conceptually harmonious with the core theme. You have infinite power, and it changes the way you approach situations. Do you get drunk on power and overspend recklessly, or do you back down out of fear and undershoot your full potential? Power changes everything. In this case, power has changed the reload/ammunition system that every other FPS game was driven by. They embedded the theme into the mechanics.

It really reminds me of the Smart Pistol system in Titanfall. Not mechanically, but conceptually. In Titanfall, the Smart Pistol is a weapon that automatically locks onto targets that you’re looking at, but it takes time to lock on. That means it takes emphasis away from precision aiming, so instead you need to focus on parkour, mobility, and wallrunning, which is exactly what Titanfall is all about. The Smart Pistol got a lot of attention for being something new and innovative in the FPS genre, and it really works in Titanfall‘s context. Obviously, the Smart Pistol would never work in Sniper Elite or Gears of War or CoD:AW, but it’s a perfect fit for Titanfall. Likewise, the battery powered weapons sound like they align with CoD:AW‘s core theme.

But that leads me into a discussion about why I probably shouldn’t be so hopeful. In Titanfall, no one uses the Smart Pistol. It’s outclassed by pretty much every other weapon there is, and they’re all your standard rifle/SMG/sniper FPS guns. Titanfall tried to innovate, but it chickened out halfway through, and as a result it never became the groundbreaking trailblazer it was meant to be. Nowadays, Titanfall can be considered in the same category as CoD and Halo and Battlefield, which really isn’t a bad thing. Still, I wish Titanfall had delivered on its promise to revolutionize FPS gameplay.

Likewise, CoD:AW will probably not be the amazing revolutionary FPS it sounds like it will be, either. Energy weapons are cool, having a core theme is cool, but in the end it’s still a triple A game. That means they’ve got big wig higher-ups who want senseless spectacles because they think that’s what sells games. So in all likelihood, CoD:AW will not be as good as I’m imagining. But it’s still nice to see that big developers are starting to change up the monotony we’re used to seeing as consumers. I hope CoD:AW delivers.

Case Study: Gender In League of Legends

League of Legends (LoL) is an online competitive video game, and as a digital space one might assume it to be relatively free of gender norms. However, the game is filled with many instances of identity normativity ranging from artistic direction to gameplay dynamics to the player community.

In LoL, players form teams of five and fight against another team (making ten people per match). The game is won when one team successfully invades the other team’s stronghold. Each individual player selects a champion to play as, and each champion has different strengths and weaknesses. Some champions can turn invisible and sneak up behind enemies, some champions can launch attacks from a very long distance, some champions can grab enemies and pull them closer. Teams need to carefully decide what champion each player picks, because the enemy team also gets to pick five champions of their choosing. All five players on a team need to be aware of their champion’s role and how it relates to the roles that their teammates picked.

The stereotype is that women primarily play as supports. A support in LoL is a champion who specializes in protecting their teammates and enhancing their abilities. Supports are commonly able to heal their allies and confer buffs upon them (a buff is a temporary boost in combat abilities such as movement speed). They usually need to stay as far away from combat as possible, because if they die they are no longer able to help their teammates. A support is typically paired up with a “carry,” which is LoL terminology for a champion who can do lots of damage but is also very fragile: champions like this must be “carried” by their teammates early on because they are weak, but later in the match they become powerful and “carry” their teammates to victory.

This stereotype (women primarily play supports) has several nuances. First of all, it perpetuates the idea that women are caretakers and homeowners, the benevolent mothers who act as enablers for others rather than taking action themselves. Likewise, the counter stereotype is that men primarily play as carries, because carries are the ones tearing enemy teams apart and claiming all the glory (obviously, thanks to the help of their supports and tanks). In terms of gameplay, supports and carries have a relationship similar to that of the traditional homemaker wife and breadwinner husband.

However, this runs deeper than just a social norm: the stereotype is further strengthened by the idea that supports are easier to play as, so women play as supports because they’re not good enough to play a more difficult role (such as carry). This stereotype is particularly damaging because it is completely untrue at higher levels of play (professional LoL depends very much on each team’s supports), but as a beginner it feels true: all a support needs to do is heal their allies, and they can sit back away from the action and let their teammates fight in actual combat. The problem is exacerbated by ragers in the community: a rager is a person who will harass and belittle a teammate who isn’t doing well, and if a teammate isn’t doing well it drops their chances of winning the match. If a player tries a new role and does a poor job, their teammates might rage at them, and they’ll feel bad about it. It’s very tempting to stay away from the action so you don’t draw ragers to yourself, especially for new players, and support is (initially) the best way to do that.

If someone starting to play LoL for the first time decides to play support because support is the easiest role, they will become accustomed to that role. Maybe they’ll switch roles later on as they get better and try out different champions, but there are just as many players who prefer playing the first role they picked out of habit. From this, it is easy to see “women play supports” as a self-perpetuating stereotype: women playing LoL for the first time are incentivized to play support because it’s easier, and then they continue to play support because they’re used to it, and then other people deduce that women play support.

Riot Games (the company that makes LoL) has made efforts to reduce the stigma around support role with its recent “Teamwork OP” campaign. “OP” is game terminology for “overpowered,” which refers to a mechanic that is so strong that it is unfair to play against (if an enemy champion singlehandedly defeats my whole team, that champion would be OP). In this case, Riot’s campaign raised awareness that having good teamwork is just as powerful, if not more powerful than having strong individual skill.

Riot’s Teamwork OP campaign.

Traditionally, supports were supposed to buy wards, which are items that grant vision of an area and alert teammates when an enemy is nearby, which prevents enemies from sneaking up on allied characters. Buying wards costs gold that could be spent on buying other items that increase the amount of damage you do, so carries typically never bought wards and instead saved all of their money to buy damage-boosting items instead, and supports had to spend all their income on wards. Ever since Riot’s Teamwork OP campaign, it is no longer rare to see all five champions on a team buying wards, rather than having all of them offload ward duty to the support.

LoL is almost a microcosm of society: many modern day interpersonal connections can be interpreted as a variation on the support/carry relationship. However, that is not to say that supporting is wrong and carrying is right, nor to say that women are supporters while men are carriers. People can be more nuanced than that, and LoL proved it with its recent Teamwork OP campaign. We can develop a better understanding of gender roles in our society by analyzing gender roles in LoL.

Case Study: Arguing Against Mana in League of Legends

Ryze, the living embodiment of mana in LoL.

This essay takes the concepts from my previous essay The Mana System Paradox and applies it specifically to League of Legends. Knowledge of both of these is preferable, but if you haven’t read “The Mana System Paradox”, its core argument is that resource systems in action games are bad because they punish players for taking actions in a game that’s all about taking actions.

Thus the main question is this: is LoL one of those kinds of games? I believe it is, and that it would benefit from rethinking its mana systems.

In my original essay, I took apart a few games that use mana and analyzed what the resulting effects were. There is no blanket statement on mana systems: they work great for some kinds of games, but are horrible in others. Copying a mana system directly from one game to another can have drastic consequences, because every game is different and reacts differently depending on how its mechanics communicate. Just imagine if Magicka had a traditional mana system.

A game like LoL has a lot of subtle nuance: many factors go into every interaction in every match, and mana governs over a lot of those interactions. Judging mana in LoL requires a breakdown of mana’s consequences and its effect on the game as a whole.

A Quick Foreword

This essay does not argue for the immediate abolishment of mana. Simply snapping a finger and instantly making every champion manaless would have disastrous effects. Many of the game’s core systems are built off of mana, balanced around mana, and deliver engaging experiences through mana. There are many good points about mana, and the bad points aren’t quite as obvious. Personal mana control, resource management, and long-term tactics are all strong mana-based aspects of the game that I don’t intend to demean.

For example, the argument “Mana isn’t a problem if you buy an Unholy Grail and run regen seals” lies within the issue itself. I am not talking about mana on a personal scale, but rather on a game-wide level. Athene’s Unholy Grail and mana regeneration seals fall under the scope of mana that I seek to address.

Champions like Blitzcrank, Kassadin, Kog’Maw, and Ryze who depend heavily on their dynamics with mana would also need to be reconsidered in this new direction. I do not mean to say that these champions should just have all of their mana costs removed: of course more care would be necessary for cases like those.

Rethinking mana in LoL would be an extraordinarily massive shift in design philosophy, possibly one of the largest in the history of the game. Large tasks are to be approached in small steps, and I hope to make one of those small steps with this essay.

Problems With Mana: Hoarding Encourages Passive Turtling

LoL‘s design philosophy is all about interactive play between opponents, and mana doesn’t allow much of that to happen. The problem with a statement like this is that “mana” is often connected to “abilities,” but they’re two separate concepts that are considered separately in battle and need to be considered separately in design as well.

There’s no way for players to intentionally cause their enemy to run out of mana. It’s bad for you when your enemy has a lot of mana, because that means they can use a lot of abilities on you. But the only way for you to reduce their mana (and thus reduce their threat level) is to let them use abilities on you. Gameplay around the actual usage of abilities has plenty of interaction, but there’s a distinction between using an ability versus deciding whether or not you should use an ability.

Let’s say that I’m playing as Ezreal and I want to hit you with a few Mystic Shots. There are two decisions that I need to factor into the equation before I fire an ability off: how it will impact my current mana pool, and how it will impact my enemy. The net worth of my ability depends on the latter, because even though two casts of the ability cost the same amount of mana, it’s worth more to me if it hits than if it’s a complete miss. However, enemies still don’t have the ability to do anything to my mana pool: they can only change what happens after my ability is fired, if they react fast enough to dodge. All resource-consuming abilities have this dual nature between deciding when to use the ability (mana) versus deciding how to use the ability (counterplay).

Thus we run into a problem: what happens if I just decide not to use any abilities? I can maintain high amounts of mana (which means a higher threat level) and there’s nothing that my enemy can do about it. Sure, I’m not using abilities, but I’m not losing mana either. It’s beneficial for me to do damage to my enemy, but it’s also beneficial for me to have lots of mana, and I can only do one or the other because damaging my enemy costs mana. If I’m able to do enough damage to my enemy to kill him, the benefits outweigh the loss of mana. But on the other hand, if my enemy outplays me, now I’ve done no damage and I’m out of mana, which is a bad situation to be in. If I just don’t use any mana in the first place, neither of those outcomes will happen, and I’ll be in an unremarkable but safe place.

If you’re at 100% mana, you’re probably throwing out some harass now and then. If you’re at 50% mana, you’re a little bit more hesitant about using your abilities. If you’re at 10% mana, you’re saving every last drop for your escape mechanism. The degree to which you need to save mana scales with how little of it is left.

It’s a decision between the high-stakes path versus the consistent path, and the latter is a pretty popular option. LoL has struggled with passive play for a long time, most notably with the support sustain meta. But it’s important to notice that mana systems also contribute to passive play. No one wants to get yelled at for being killed, and it’s a lot easier to avoid being killed when you have mana than when you don’t.

One solution might be to strengthen the connection from “using the ability” to “deciding when to use the ability.” Ezreal is a good example of this, because when he hits an enemy with his Mystic Shot it reduces his ability cooldowns by a second. His decision to use Mystic Shot is reliant on how well he can land Mystic Shots. For every other champion, it’s the other way around: their decisions to use their abilities is reliant on their available resources, whether it’s mana or cooldowns. Enemies can reduce the rate at which Ezreal uses Mystic Shots by dodging the ones he fires. This places emphasis within the action of the ability rather than the decision behind its usage. Olaf’s Undertow and Reckless Swing abilities were recently reworked to use similar structures.

Another possible solution would be to allow interaction with mana, but LoL decided not to go in that direction. If enemies could somehow reduce the amount of mana you have, they could actively prevent you from using your abilities. In the MOBA realm this concept is called mana burn, and LoL intentionally removed it. Having mana burn would solve this current problem, but it still feeds into the next problem I’m about to describe.

Problems With Mana: Being OOM Sucks

Naturally, resource systems cause things to run out. The paradox is that the gameplay experience depends on having resources that are consumed by the gameplay experience itself. Why does playing the game as it’s intended to be played cause you to be unable to play the game anymore? If the core gameplay experience of LoL is to use abilities, and using abilities causes you to run out of mana and be unable to use abilities, then there’s a problem. It’s antithetical and confusing for the game’s core experience to prevent you from fulfilling the game’s core experience.

But it’s a matter of understanding what exactly LoL‘s experience is supposed to be. Is it to outsmart your opponent? Get the most gold? Last-hit the enemy nexus? Kill Teemo as many times as possible? Have as much fun as possible? You could take each of these random ideas I just threw out and design a whole game around it. Implementing a mana system makes perfect sense for some of these ideas, and is a horrible idea in others. If you went all-in with a ton of mana and your opponent outsmarted you, that’s a proper punishment in a game about outsmarting your opponent. On the other hand, if you run out of mana because you killed Teemo too many times, and the only reason why you play the game is so you can kill Teemo, it’s a total buzzkill. Of course, the core gameplay experience of LoL isn’t actually to kill Teemo. What exactly is the core gameplay experience, and why don’t mana systems play into it?

Many of Riot’s prominent designers have said that counterplay and player interaction is a major focal point for LoL‘s experience, and you can feel this in the game itself. If you look at supports in LoL, traditional healers were nerfed because they didn’t provide strong counterplay: there’s not a whole lot you can do when the enemy Soraka heals your lane opponent back to full health every ten seconds. However, recent supports have been designed with counterplay in mind. Nami’s Ebb and Tide ability is a weak heal outside of combat, but gains strength when used within an engagement, so it becomes an active skill used in the heat of battle rather than a way to patch up allies after the fight. Thresh has no heals whatsoever but is able to reposition his allies, so enemies must coordinate in order to catch an enemy being assisted by a Thresh. Other trends like the rising amount of dodgeable skillshots have also pointed towards this direction.

A large part of counterplay lies in abilities. If I’m Ezreal and I fire a Mystic Shot at you, your counterplay is to dodge. If I’m Yasuo and I place a Wind Wall in front of you, your counterplay is to navigate around it. If I’m Zyra and I place a seed on the ground, your counterplay is to trample it. Other than abilities, counterplay also happens through item purchases and through metagame. But counterplay through abilities is the main hook, the event that happens most often during a match, the part you remember after an awesome play, the moment when crowds get up and cheer at tournaments. Nobody gets up to cheer when the tank buys a Randuin’s Omen.

In order for counterplay through abilities to happen, it’s obvious that abilities need to be used. However, the existence of mana presents two cases which prevent the use of abilities: holding back in order to save mana, or being out of mana. Either way, someone is unable to interact meaningfully with their opponent, and it’s frustrating.

These concepts are even used thematically in champion design. Rumble’s overheated state is functionally the same thing as being out of mana. For him, the frustration of being unable to unleash all of his abilities is a part of his character. He has to fight against himself as much as he has to fight against enemies, which is why he needs Danger Zone benefits: the more he fights against himself, the easier it is to fight against enemies and vice versa. On the other hand, manaless champions like Katarina can divert focus away from herself and more towards her enemies, like an assassin concentrating on her prey. As thematic mechanics, both of these champion’s resource systems work fine. But giving a champion mana is basically giving it a toned-down version of Rumble’s heat. It’s not for everyone, and it shouldn’t be the default. Ziggs is a “blow-up-everything” kind of guy, but since he has mana there are times when he has to stop himself from blowing things up.

There are plenty of other situations that prevent abilities, but those situations tie into LoL‘s core experience. If you’re silenced or stunned or dead, you’re unable to use abilities because the enemy outmaneuvered you in some way. The punishment is reasonable because your counterplay was not as strong as the enemy’s. However, being out of mana isn’t a matter of being outplayed, it’s a matter of using the abilities you were supposed to use. Mana systems inevitably lead towards OOM situations, but in a game about interactive counterplay it’s more fitting to focus on dealing with your opponent rather than struggle to balance your own expenses.

Problems With Mana: Balance Revolves Around It

Regardless of anything I say about mana, it’s still a valid point that LoL was built from the ground up with mana, and it’s a core part of how the game is balanced. However, mana is a balancing point for the decision to use an ability, not the action of using an ability. Attaching high post-usage costs feeds into the passive turtling and OOM scenarios described above, and doesn’t deal with the abilities themselves.

Currently, nearly all ranged champions are gated by mana costs to limit their harass potential, and manaless champions generally have higher ability cooldowns, but mana and cooldown gates both shift too much focus into the decision to act rather than the act itself. Mana and cooldowns serve very similar purposes: both of them limit ability usage by forcing the user to wait for a certain duration. An individual ability’s cooldown can be thought of as a miniature mana bar, with its own regeneration rate. Xerath’s 20-second cooldown on Locus of Power can be reimagined as a mana bar that’s capped at 100, regenerates 5 per second, and costs 100 to use. Thinking this way, it’s as if all champions have five different mana bars, one for each separate ability and one to govern them all.

Many of my arguments against mana also apply to high-cooldown abilities like ultimates and summoner spells. If you don’t use an ability, it won’t go on cooldown, and it’ll be ready to use if you’re suddenly caught in a bad situation like a gank. Offensive flashes are rarely used except when a kill is almost finished and it just needs to be secured, but defensive flashes happen all the time (in fact, offensive flashes are generally used to follow up after defensive ones). The similarities between cooldowns and mana mean that both systems carry similar effects on gameplay. High cooldowns should be condemned as much as high mana costs.

Abilities need to provide windows for counterattacks, but mana and cooldown costs present those windows outside of the ability itself. Sion’s Cryptic Gaze has a high mana cost, but it doesn’t lend much counterplay. This has historically turned Sion into a one-trick pony: once he stuns you, he either succeeds at bursting you down with his shield, or he fails and he’s unable to take other actions. In the former case, his enemy has no meaningful actions to respond with because the stun is inevitable. In the latter case, Sion himself is left out of mana and has no meaningful actions to use against enemies. The counterattack case is clear (if Sion fails to kill me, attack him), but that’s an external cost that lies outside of the ability itself. If, for example, Cryptic Gaze was a skillshot, it would provide counterplay in and of itself because it would be dodgeable. Lissandra’s Frozen Tomb is the cooldown equivalent, and presents a similar gameplay pattern (but unlike Sion she can still function without her on-demand stun).

Up until several months ago, every champion has been gated by either mana or cooldowns, but recently abilities have been gated by windups instead. Mana and cooldowns are both post-usage costs, which makes them feel more like punishments for taking action. However, windups shift the cost to pre-usage, so players need to properly set up a scene before using abilities to their full potential. Pre-usage costs are paid in preparation rather than in mana, and the effort invested in pulling off a perfect ability results in a positive payoff. Lucian’s Culling is a strong example of this: by itself, it’s rather difficult to land and doesn’t do much damage. However, when Lucian first hits an enemy with Ardent Blaze, the movement speed allows him to keep pace during The Culling, so more of his shots will land and he does more damage. If in addition he also times his Relentless Pursuits well, he gains even more potential out of his ultimate. Thanks to all of these elements, The Culling is able to have relatively low post-usage costs, because so much of its cost is shifted to pre-usage preparation.

Contrast Lucian’s Culling to Miss Fortune’s Bullet Time, which doesn’t benefit from prior preparation and is instead balanced by longer cooldowns. Let us imagine for the sake of example that Bullet Time has a 60 second cooldown and The Culling has a 50 second cooldown. The Culling just by itself is drastically weaker than Bullet Time, but Lucian can invest ten seconds to properly prepare a perfect scene before ulting. This means that each ability needs 60 seconds to reach its full potential. However, those ten seconds that Lucian is spending on preparation means  a ten second window for enemies to respond. His enemy can dodge Ardent Blaze, or try to force him to reposition with Relentless Pursuit, so that The Culling’s potential drops. In this case, enemies have ten seconds to exert force and try to outplay Lucian to reduce his total power. On the other hand, there’s nothing that enemies can do about Miss Fortune’s Bullet Time except dodge it and spend the next 60 seconds on offense.

Having lower post-usage costs is good because it weakens the mana-hoarding strategy and reduces OOM punishments. In this over-simplified hypothetical case, Lucian’s The Culling is able to have lower post-usage costs than Bullet Time because it’s weaker. To compensate, Lucian can invest time and effort to make The Culling stronger, but likewise enemies can interfere with his efforts. This way, both abilities can be balanced without running into the problems of high post-usage costs. Of course, there are so many other elements that play into the actual game, so I’m not trying to say that The Culling is the paragon of balance. Still, many recent champions have taken advantage of pre-usage preparation.

Mana costs work well enough at balancing abilities, but they do so on a game-wide scale rather than fixing the dynamics of individual combat scenarios. Costs are a necessary element for counterplay to exist, but they do not necessarily have to be traditional post-usage costs. Abilities can be balanced by attaching high mana or cooldown costs, but that doesn’t give the ability counterplay and runs into the mana hoarding/OOM problems. On the other hand, abilities with pre-usage costs can have windows for counterplay, low post-usage costs, and balanced power levels.

A New Direction: Externalized Mana Design

Being unable to act should be a punishment. Taking actions is what makes LoL fun, and if you’re rendered unable to do that, it should be because you did something wrong and your opponent rightfully bested you. And yet, mana systems bring you closer and closer to being unable to act every time you use an ability. When you lose mana, you’re being punished for trying to play the game.

Rewards and punishments should be justified. Punishing players for using an ability is not justified in a game where you’re supposed to use abilities. Instead, reward players for their successes and punish them for their failures. This allows opponents to focus on dealing with each other, rather than worrying about themselves. If your opponent is unable to take action, it should be because you intentionally rendered them unable to take action, not because you curled up into a ball and let them rain blows upon you until they got tired (unless you are Rammus).

Windups create anticipation, whereas cooldowns create dread. Ability gates need to exist so that opponents can have a meaningful opportunity to counterattack, but the timing of where those gates go changes how the ability is experienced. Time spent preparing for an attack makes the payoff a reward, whereas time spent waiting after an attack makes the result a punishment. This also gives enemies opportunities to disrupt a powerful attack during its preparation phase, so the outcome of the battle depends on how well the combatants play within the battle itself rather than how many costs they incurred before the fight.

I call this direction “externalized mana design (EMD),” as opposed to LoL’s current “internalized mana design (IMD).” With these definitions, internal refers to the self, whereas external refers to how the self interacts with the others around it. Internal refers to how many mistakes you make, whereas external refers to how many times your opponents best you. Internal refers to sticking to your guns and maintaining your own path, whereas external refers to how you change dynamically in response to your enemies.

Mana and ability usage gates are too much of an internal struggle and not enough of an external one. In a game about counterplay, the tools that we use to outplay our opponents are limited by personal consumption. There are still many, many benefits to IMD, but externalism and internalism are not mutually exclusive: it’s possible to get the best of both worlds.

LoL already has many examples of EMD. Cho’Gath’s and Swain’s high costs are balanced by their regeneration passives, so they’re rewarded for getting kills and punished when their enemies successfully starve them. The ninja’s energy systems are great ways to focus on in-combat dynamics: you can starve a ninja’s resources by avoiding their energy regeneration abilities, which likewise makes it a priority for them to land those abilities, but energy refills after a fight so there are no crippling long-term mana costs.

I feel that Yasuo channels the general concept of EMD most strongly. His Q and R put emphasis into windup, so they’re able to function with plenty of counterplay without having crippling cooldown costs. If his ultimate wasn’t restricted to airborne enemies, it would essentially be a 1300-range AoE stun nuke, and its cooldown would have to be ridiculously long to compensate. Since it has that prerequisite restriction, Yasuo and his enemies are able to have interactive, exciting battles where the most skillful player wins. All of his kit works on interactive reward systems that both the player and the enemy can build off of.

On the other hand, Rumble is the epitome of IMD. His internal struggle against his own resource system is his central character trait, his game balancing point, and ultimately one of the largest parts of the Rumble experience. That’s not to say that Rumble is a bad champion, nor that IMD is a bad design philosophy. There are pros and cons to both IMD and EMD, and a careful mixture of both can create uniquely engaging experiences.

Illustrating The Concept Of Externalized Mana Design

These charts are here to further illustrate what exactly EMD is and how it compares to IMD. Let’s take Varus’s Piercing Arrow as an example and analyze it using these graphs.

First, the top chart analyzes windups and cooldowns, the outside factors that go into ability uses. Windups are the parts where you try to ensure that your ability will land (or deal more damage) once you use it, but before actually using it. For Piercing Arrow, the windup is applying Blight stacks: Varus isn’t actually using Piercing Arrow, but the Blight stacks exist outside of the ability and can make it stronger. Other windup factors would include positioning, timing, and lining up so that enemy minions aren’t in the way.

The more effort Varus puts into the windup, the higher his threat level becomes. However, after he unleashes Piercing Arrow, his threat level suddenly drops: he loses mana, his ability goes on cooldown, he is vulnerable to counterattack. As he recovers mana and his cooldown continues counting down, his threat level slowly increases again as his opponent recognizes that he’ll soon be able to fire another Piercing Arrow.

EMD puts emphasis on the first portion, whereas IMD puts emphasis in the latter. For example, Karthus’s Requiem has very little windup: there is nothing he can do to increase the damage it does. To balance the lack of windup, it needs to have massive cooldown and mana costs.

While the top chart focuses on the outside elements influencing abilities, the bottom chart focuses specifically on what happens when an ability is fired.  Once again, let’s use Piercing Arrow as an example. As soon as he presses the key and starts drawing back his arrow, the mana and cooldown costs are applied, dropping the ability’s net worth. Enemies may then take actions in response to Varus’s ability. If Piercing Arrow lands, the net worth will go up and the damage dealt to the enemy will compensate for the costs incurred. However, if Piercing Arrow misses, the net worth doesn’t change: the initial cost was already paid at the start of the ability.

There are two directions that EMD can take within an ability’s specific dynamics, and both of them attempt to exaggerate the final effects. Notice the degree of the angle that occurs at “the actions that are taken in response.” Under IMD, this angle is quite low: a hit is better than a miss, but not drastically so. However, under both directions of EMD the angle is much wider, representing a much wider difference between a hit and a miss. Both directions of EMD attempt to make landing a hit drastically better than a miss (and conversely, they make missing a hit drastically more punishing).

Under EMD 1, Varus’s Piercing Arrow would cost more mana, but refund mana if it hits an enemy. Under EMD 2, Varus’s Piercing Arrow would cost less mana than it does now, but drain additional mana if it expires without hitting an enemy. Either way accomplishes the same goal of exaggerating the difference between a success and a failure. An ability’s net worth from its specific dynamics will propagate further up to the surrounding influences depicted in the first chart.

The point of EMD is to put more emphasis on “the actions that are taken in response,” in other words, counterplay. Punishments and rewards can be made to be an interactive experience rather than an individual one. Effects from these individual abilities can influence the flow of battle as a whole by strengthening the connection between long-term costs and short-term dynamics.

Applying Externalized Mana Design

Before I start, I must mention: I do not presume to be the god of design and balance. All of these are just ideas and nothing more: actually implementing these in any kind of useful fashion requires far more effort and resources than I have available. However, I hope to illustrate the general direction that EMD would take.

Riven – Increase the cooldowns on her basic abilities, but add an effect to her passive: if she attacks an enemy champion and consumes a charge of Runic Blade, it reduces her basic ability cooldowns by a flat amount. As an enemy, this helps clarify that Riven is a threat when she is engaged in a combo, and that you can win by breaking her combo. As Riven, this means that you have to be more careful before you rush in with a combo, because if the enemy manages to disengage, you don’t get the cooldown restoration and you are more vulnerable to counterattacks as a result. However, if she manages to get her full combo off, she’s rewarded for it with reduced cooldowns and thus higher threat potential.

Ezreal – I like the positive feedback attached to his Mystic Shot on-hit cooldown reduction, but I feel that it could be pushed further. EMD could be emphasized in his kit by increasing the base cooldown on Mystic Shot by a certain amount, but also increasing the amount that Mystic Shot reduces its own cooldown (important: only its own cooldown) by an equivalent amount. If EMD’d Ezreal lands a Mystic Shot, he is functionally identical to the current Ezreal. However, if EMD’d Ezreal misses a Mystic Shot, he is punished with a longer cooldown and enemies have a greater window for counterattack.

Nidalee – Reduce the overall damage of her spears, but if she lands a spear on an enemy currently revealed by one of her traps, it does bonus damage. I think it’s fine that she runs on mana, because her cougar form is manaless and it creates an interesting duality between human restraint versus bestial ruthlessness. However, tweaking her damage around this way gives Nidalee’s spear a preparation cost that it didn’t have, and reduces its uncounterable consistent poke damage. With this, the danger case is clear: if you’re revealed by a Nidalee trap, you better be really careful for spears coming your way, because not only can she see you, her Javelin Toss does extra damage to you. On the other hand, if she’s throwing spears randomly, she’s wasting mana on attacks that won’t make much of an impact even if they hit.

Draven – Every time he catches a Spinning Axe, he regains a bit of mana. If he misses an axe, he loses mana in addition to the mana cost he paid when first activating the ability. He’s the kind of guy who brags on and on when he catches an axe, but beats himself up when he misses one. This way, there’s even more incentive for both Draven and his enemy to focus on his axes (and of course Draven wants everyone to focus on his axes). As long as he continues catching axes, he gets rewarded with more mana to cast abilities with, but if an enemy forces him to drop an axe the enemy is rewarded for their play by actively hurting Draven’s mana pool (and ego).

Karthus – Change Defile’s mana restoration on kill to be mana restoration upon dealing magic damage instead, like a form of spell vamp that heals mana. Make Defile scale with and consume a percentage of his maximum mana. Currently, Karthus relies on getting kills to restore mana, which is how he sustains through the laning phase, and after the laning phase he hardly needs to worry about mana at all because he builds tear. By switching “kills” to damage in general, mana loss and restoration stays relevant throughout the game.

Ziggs – Create some kind of interaction from Short Fuse to his basic abilities. The interaction from basic abilities to Short Fuse is already there (using an ability reduces Short Fuse’s cooldown by 4 seconds), but it could go the other way around as well. Increase Ziggs’s basic ability cooldowns, and make Short Fuse reduce their cooldowns when it damages an enemy. Alternatively, reduce Ziggs’s basic ability damages, but Short Fuse applies a debuff to enemies that increases the damage they take from Ziggs. There are lots of things that can be done to Short Fuse that would make Ziggs more interesting than a max-range Bouncing Bomb spammer.

Yorick – Increase the cost it takes to summon a ghoul, but have all of them restore mana back to Yorick every time they strike an enemy. His ghouls are way too spammable, and enemies should have the opportunity to fight back. This way, if Yorick summons a ghoul, the enemy has to decide between going about his previous business (farming, harassing, etc) or focusing down the ghoul before it replenishes Yorick’s mana pool. However, if Yorick is smart with his ghouls, he’s rewarded with more mana.

Sivir – She’s interesting because her Spell Shield was actually changed in the opposite direction: it used to have a mana cost and restore more mana, but then the mana cost was chopped out so she can cast it free. On one hand, this is necessary to balance the high cost on Boomerang Blade, so the interaction of these two abilities creates a unified whole system that welcomes smart counterplay from both sides. On the other hand, I wonder what would happen if Boomerang Blade’s power and mana costs were reduced so it wouldn’t need an auxiliary balancing point. Either way is fine.

Brand – This is getting into personal opinion territory, but I feel like Pyroclasm’s Blaze effect is a little underwhelming compared to the Blaze effects of his basic abilities.  Shifting some of Pyroclasm’s power into its Blaze effect wouldn’t change a massive amount because it’s the last step of Brand’s combo anyway, but it’s a step towards more preparation-based abilities.

Kassadin and Kog’Maw – I’m able to lump these two together because the primary issue is their simultaneous damage/mana cost increases with consecutive uses of their ultimates. For them, I would suggest a new type of resource, Void Energy. Void Energy would work similarly to Rumble’s Heat: starts at 0, caps at 100, decreases over time (although Void Energy would decrease at a slower pace than Heat), silences the champion for a duration when 100 is reached. All of Kassadin’s and Kog’Maw’s abilities would add a certain amount of Void Energy to their pool, depending on the ability level, and thus bringing them closer to hitting 100 and becoming silenced. However, dealing damage to enemies decreases Void Energy proportional to the amount of damage dealt (reduced to a third for AoE and DoT damage, doubled against champions). This way, they are naturally capped from spamming their abilities (namely, their short cooldown ults), but they can strive to make the most out of their Void Energy before it silences them. Likewise, enemies have to shut them down before they’re able to refund Void Energy and continue their assaults. Again, I am not the god of game design, I can’t whip out a perfect new resource system out of thin air, but it’s just to illustrate the concept of EMD.

Blue buff – Change its mana regeneration component from a constant passive aura to an active effort. Something like a stronger version of Doran’s Ring’s restore mana on kill would be suitable, or restore mana upon damaging an enemy, or restore mana proportional to damage dealt. This could possibly be applied to Baron buff as well, but considering Baron buff’s rarity it’s probably not an urgent change.

Ultimates and summoner spells – I think that summoner spells are fine because they’re external forces, like deus ex machinas that you call upon to save your champion. But with long cooldown ultimates, it’s really a case by case basis. For example, Shen’s Stand United is just so powerful that there’s no way it could function without a massive cost. If some of its power was shifted to preparation cost instead, that could mitigate its strength and it would be possible to reduce its post-usage costs. On the other hand, LoL benefits from the diversity. There’s no hard line stance I can take on long cooldown ultimates, as long as they aren’t a core integral part of what makes the character engaging and fun to play. If they are, they should be integrated like Yasuo’s and Lucian’s ults.

A general EMD trend is an overall nerf on long-range poke, and I think that’s a good thing. Just like the pure healer support playstyle that was nerfed long ago, the long-range poke playstyle is frustrating and not very fun for the victim. Attaching long-range poke to in-combat dynamics reduces the effectiveness of poking outside of battle.

Consequences Of Externalized Mana Design

One major consequence that I can already tell is lane snowballing. Positive feedback is essentially another word for snowballing, and EMD is all about positive feedback where it’s due. I talked about balancing abilities with mana earlier, but mana is also used as a long-term balance to prevent snowball cases.

On a theoretical level, this is exactly what EMD is supposed to do in the first place. If one player is more skillful than the other player in a scenario where both players are given equal opportunities to demonstrate skill, then the better player should win. The whole point of preparation cost is to give those equal opportunities to both players. However, the snowball situation could spiral out of control very quickly.

The core pieces that go into snowballing a 1v1 matchup are levels, gold, and skill. EMD would place a greater emphasis on skill, and I believe that out of those three factors skill should indeed be the defining one. If you’re able to consistently outplay an enemy three levels higher than you, you should win. Likewise, if you’re three levels higher than your enemy, you shouldn’t be able to roll your face on your keyboard and still win.

In order to balance the system as a whole, emphasis on levels and gold would have to be lowered in order to compensate for the increased emphasis on skill. That would entail reducing stat gains per level and item stats across the board. This is part of what makes EMD such an enormously massive change in direction.

Another problem with EMD is the reduction/elimination of long-term consequences. Just like with the ninja’s energy systems, the goal of EMD is to allow players to interact meaningfully every time they clash, but many clashes can happen while waiting for a long cooldown to end. A lot of gameplay centers around these long-term effects, the abilities that trigger them, and the events that happen during the cooldown period.

EMD itself is adverse to long-term effects. Abilities that are powerful enough to warrant such costs can’t be easily shifted to a preparation-cost format, because they would have to require a ridiculous amount of preparation. In addition, if enemies are able to counterplay against these powerful abilities, the punishment for the original caster feels unproportional to the effort required by the enemy. It’s like if you were playing Magic: The Gathering and you tried to summon an 8-mana creature, but your enemy prevented it with a 2-mana counterspell. A more relevant example would be stunning Karthus as he channels Requiem, putting his ult on a 3-minute cooldown with little effort (in comparison) on your end. There’s no clean way to attach long-term effects to EMD.

This is why I am not arguing for LoL to completely convert to EMD. IMD is far better at making these long-term effects balanced and engaging. A MOBA like Awesomenauts fully embraces EMD because its design goal is very different from LoL’s, and the end result is a palpable contrast in the game experience. Awesomenauts has no mana, short cooldowns, and absolutely everything is manually aimed (and is thus dodgeable). In exchange for letting players do awesome things whenever and wherever they want, Awesomenauts loses the long-term tactical depth that LoL gets out of IMD. And likewise, the opposite is true: LoL cannot subsist solely on IMD, for the reasons I wrote this essay to argue for.

IMD and EMD can coexist, and many LoL champions are already proof of that. Yasuo can fight Rumble fairly and cleanly. Manaless champions and mana-based champions battle side by side. The effect of EMD is already there: we need only be aware of its existence, and we can reap its benefits.


Externalized mana design is a tool just like internalized mana design is, and like all tools, it has its uses and its flaws. Some situations are undeniably better with IMD, and some with EMD. Some situations will call for a completely different direction altogether. But it’s important to recognize that these different directions exist, and weigh their pros and cons against each other rather than defaulting to one. Right now, LoL defaults towards IMD.

It may be presumptuous for me to say this, but I want to see a game where every player strains their hardest to win. I want to see a game where people clash with absolutely everything they’ve got. I want to see a game where whether I win or lose, I understand that the result was fairly and soundly earned, and I can be satisfied thinking about how hard I tried. LoL is very, very close to being that game, and EMD is a tool to help it get there.

Case Studies: The Importance Of A Core Theme

I’ve been looking at a few games, and I’ve been thinking something about design: the closer that every designed element is to the core theme of the object in question, the better the overall design is. Good design is about making sure that every aspect points to the core theme in some way, shape, or form. The following five games are ones I’ve analyzed with this concept in mind: first determining what the game’s core theme is, and then identifying mechanics that don’t tie into it sufficiently (or in worse cases, directly go against it). Hopefully, identifying and focusing on a core theme will help me design better games.

1. The World Ends With You: Solo Remix (Square Enix, 2012)

The new battle system implemented in Solo Remix.

TWEWY is a video game that mainly revolves around the theme of cooperation and teamwork. You play as the antisocial character Neku Sakuraba, who’s suddenly thrown into a confusing but life-threatening game that absolutely must be played in pairs. The most interesting thing about TWEWY was that you had to control both characters simultaneously: Neku Sakuraba, and the person he’s partnered with. This was obviously very challenging, because most video games only involve controlling one character at a time, and humans are not very good at multitasking in the first place. During the early stages of the game, players would normally focus only on controlling Neku and devote very little attention to controlling his partner, so the partner’s usefulness was extremely limited. Players at this stage would come to the conclusion that the partner was useless baggage and that it would be so much easier to play only as Neku (which is the exact same conclusion that Neku himself comes to).

However, as the player continues the game and gets better at it, they start being good enough to devote focus to the secondary character as well. After improving this way, players would be able to seamlessly control both characters simultaneously, gaining massive advantages that could not have been achieved by focusing on only one at a time. This system tied in symbolically with the game’s core theme, and was an example of pretty decent design. Many people complained that the game had a very steep learning curve, but I interpreted it as a way of forcing players to understand just how difficult it is for Neku to cooperate with other people. It definitely could have been improved, but it really drove home the narrative of TWEWY.

All of that was for the original TWEWY, which was released in 2008. In 2012, Square Enix released an updated version of TWEWY, subtitled as Solo Remix to differentiate it from the 2008 version. Solo Remix completely eliminates the simultaneous control system: your partner in Solo Remix is controlled by the game’s AI (1). Suddenly, the message was gone. Players no longer had to focus on their partner character as much as they did in the original version. It certainly made the game easier for new players to pick up, but it lost the symbolic meaning that the original TWEWY‘s system carried.

TWEWY had to find a balance between accessibility and symbolism. The original version chose strong symbolism, while Solo Remix went with accessibility. But there had to have been a way to get the best of both worlds. Square Enix had several ways to go about doing this: they could have given players more options that affect both characters simultaneously rather than force them to input commands separately for each character, or they could have equalized the complexity difference between controlling Neku versus controlling his partner, or they could have locked certain abilities so that only a specific character could use them. All of these could have made the game more accessible to play, but still maintain the symbolic nature of cooperation and teamwork.

2. Infinity Blade 3 (Chair Entertainment, 2013)

The Potionmaster at his cauldron.

Infinity Blade 3 is an action game where you play as a guy who fights lots of big monsters. It doesn’t have a deep narrative theme in the same way that TWEWY did, but the game did have a solid core experience: constant growth. In order to progress through the game, you grew. You would get stronger after every battle, and you would get money to buy stronger swords, and you just repeated that over and over until you were strong enough to overcome whatever hurdle the game threw at you. And then the game would throw an even more difficult hurdle at you, and you would have to get even stronger than you were before.

This was the philosophy behind the first two Infinity Blade games. However, Infinity Blade 3 introduced a new element: potions that you could drink before a battle to increase your combat strength. Some potions would let you deal more damage, other potions would increase the amount of damage you could take, and still more potions would do things like boost the amount of money you got after winning battles. Potion effects would only last for one fight.

Suddenly, players were pulled out of the “constant growth” mentality. You could just stock up on tons of potions, but unless you actually used them they wouldn’t do anything for you. And once you used them, you wouldn’t have them anymore, so the strength you experienced was only temporary, which directly contradicts the nature of constant growth. Players had to find a balance between using a potion as an easy way to win a fight, versus gritting their teeth and trying to win through sheer power of will. Whichever option they picked, their decision wouldn’t feed into “constant growth” anymore.

Not only that, but Infinity Blade 3 had to balance around potions. When the game didn’t have potions, Chair’s designers had a rough estimate of what the player’s power level was at any given time, and they could present challenges that were reasonable and scaled to give players a task that wasn’t too easy to be trivial, nor too difficult to be overwhelming. But potions suddenly threw this system out of whack: the designers had no idea when players were using potions and when they weren’t, so they didn’t know how difficult to make enemies. If an enemy was moderately difficult to defeat normally, players could just down a power-up potion and that enemy would pose no threat whatsoever. On the other hand, if an enemy was moderately difficult to defeat while under the effect of a potion, players without potions would stand no chance because their power level was significantly lower. Infinity Blade 3‘s designers had three ways they could deal with this: they could balance around potions and make the game obscenely difficult for people without them, they could balance around normal combat and make the game obscenely easy for people with potions, or they could make potion effects so trivial that there was no difference.

It’s not a difficult problem to solve at all: just remove potions. They weren’t in either of the first two Infinity Blade games, and there was absolutely no reason to include them for Infinity Blade 3. If they absolutely had to stay, Chair should have at least put in some kind of mechanic to give them lasting potential: what if after drinking a potion, it regenerates after a certain amount of time. As little sense as that would make, it would at least give players the sense that by collecting potions, they are slowly but steadily increasing their strength over time, which is the whole point of Infinity Blade.

3. FTL: Faster Than Light (Subset Games, 2012)

FTL’s ship upgrading mechanic.

FTL is a science-fiction game set in a world where humanity and other sentient alien races have developed faster-than-light travel, enabling them to navigate among the stars. You play as the commander of a space ship on a very important mission, and you must use FTL transportation to reach your destination while avoiding the rebels that want to destroy you. The core philosophy of the game, as described by its designer Justin Ma, was to make players “feel like they were Captain Picard yelling at engineers to get the shields back online.” Tasks like piloting the ship or aiming weapons at enemy ships are all handled by the crew you control: instead, the player has to manage the ship’s energy resources and decide what systems need to be powered on or off. Add more power to the engines and increase your chance of evading enemy shots, or add more power to your weapons so you can fire more shots at the enemy?

Ultimately, I don’t feel that FTL fully accomplished that vision, for several reasons. The game is very much centered around resource management, but you are always able to upgrade your ship’s energy resource supply. Powering up your shields costs energy, and if you’re in a situation where you want stronger shields, your primary decision should have been “what system do I divert energy away from so I can add that energy to my ship’s shields?” However, since you can increase your total energy supply almost at will, you can just do that and have all of your systems functioning at full power. I wonder what would have happened if you were unable to increase your ship’s total energy supply manually, but rather upgrades were purchasable as merchandise in stores, or received as rewards for helping other people, or found as loot from destroyed enemy ships. That way, the player’s energy supply would be capped with no easy way of increasing it, so they would have to figure out what to do with the limited resources they have at the time.

The other problem that I think limits FTL‘s core vision is the variety of systems at the player’s disposal. Shields, engines, and weapons turn out to be the most useful out of the five systems to put energy into. However, the ship has two more systems, oxygen and health, which refills the oxygen supply in the ship and heals crewmembers respectively. These systems can often be left with very little energy to continue running, and if no one in the crew is injured, there’s no reason to power your health system in the first place. Since these two are so simple and low-cost, players rarely have to make decisions based around them. Sometimes, you may want to depower your oxygen supply so you can power on another weapon and hope that you’ll end a battle before your ship runs out of oxygen, and that’s an exciting decision to face. But with how little energy is required to keep the oxygen system running, it doesn’t happen very often. If the oxygen and health systems were a little more engaging,  players would have had to juggle their energy supply around five systems rather than just three. What if your crew would move faster and repair items more quickly if the oxygen supply was high? I don’t know, but it’s just an idea.

4. League of Legends (Riot Games, 2009-present)

A support’s role in LoL.

The online competitive game League of Legends pits players in head-to-head combat against each other in teams of five. Each player picks a unique champion and enters an arena alongside the four other champions picked by their allies, and they fight over objectives against the enemy’s five champions. LoL has always had a core design philosophy, which was to encourage player interaction as much as possible. People should be incentivized to launch attacks against the enemy, defend against and counter incoming attacks, and support their teammates. However, LoL‘s design was shaken for a long period when its designers mistook what “supporting their teammates” meant in the context of their game.

Certain champions in LoL are wholly dedicated to supporting their allies by healing them so that they don’t die from enemy attacks. The designers at Riot Games believed that players wanted to fulfill a healer archetype on their team by playing as these champions, so those champions are given very little tools other than the ability to heal their allies. However, this directly contradicted the core design of player interaction, because there was no motivation to try to attack your enemies. Being aggressive and trying to defeat the enemy champion is a risky undertaking, but healing effectively erases all combative endeavors. If an enemy deals damage to you, and your ally uses a spell to heal you, it’s as if the interaction never happened at all. Why bother trying to fight against your enemies at all, if their healer and yours will just restore all the health lost during an engagement?

LoL was becoming very stale and boring because of the presence of dedicated support characters (2), and the fact that they directly contradicted the core design direction of the game. However, the fact remained that people still wanted to fulfill the fantasy archetype of being the one that saves, the one who helps everyone on his team and leads them all to victory. The designers at Riot Games needed to find a way to make healing and supporting feel active and engaging, rather than defensive and reactionary, so that players could still take on a support role without erasing player interaction from the game.

If it were me, I would have wanted to play more with movement and mobility: a supportive character could help allies escape from enemy attacks, or push enemies away before they can attack his allies. And in a way, this was part of the design philosophy that Riot Games adopted to create more engaging support characters: many of them revolve around undermining enemy mobility, enhancing ally mobility, or both. This way, a support character can help allies avoid taking damage, but if they take damage anyway then it sticks. Helping allies using active abilities keeps the game’s core philosophy intact, while still allowing players to feel like they are rallying their team together as powerful leaders.

5. Mirror’s Edge (Dice, 2008)

Faith about to punch a cop.

Mirror’s Edge is a game where you play as a Runner, an illegal information courier in a dystopian city where all communications are monitored. As a Runner, you must physically travel to your destination using parkour and freerunning techniques such as climbing up walls or jumping between rooftops. The game’s core meaning is about speed and momentum: you need to manage how well you can maintain them in order to progress through the game, especially when enemy cops are chasing after you.

But the problem is how those cops are used in the game. A feature I found unnecessary was the aspect of combat: your character is able to attack cops, take their guns, and shoot at other cops. This completely goes against the game’s inherent speed-based flow, especially because your character moves slower while carrying weapons. Mirror’s Edge is most fun when you’re moving really fast, but in order to fight you need to slow down and confront your enemies, and the phrase “slow down” shouldn’t belong in this game.

This problem is a strange one because it can be completely ignored in the first place. Players are never forced to fight against enemies, and are always able to run past them or otherwise find another way to navigate around without being in danger. However, the fact that enemies exist create several problems for the game: there’s always the unfun possibility that a random stray bullet will kill you, and you’re unable to fully explore areas with enemies in them because they will keep shooting at you until you defeat them. Even if you do try to avoid enemies and focus on the speed/momentum gameplay, they still end up being a hindrance to your ability to move quickly.

A potential fix might have been to remove all guns and ranged weapons from the game, and limit all enemies to melee only. This would have kept them as a threat, but would have given players more of an opportunity to plan around them, rather than being forced to take cover and hide from bullet fire. Another fix might have been to remove enemies altogether, but they are still necessary to give areas a sense of urgency.


(1) I was actually very, very wrong about Solo Remix, but that doesn’t actually change my point so I’m adding this as an endnote. Your partner isn’t controlled by AI: you actually summon your partner in battle to do attacks alongside with you. That still doesn’t fix the problem, because now your partner is like a tool you use alongside all your other ones, rather than an actual person you have to coordinate around. It’s like a picture-perfect example of direct control versus indirect control (more on that here). Still, goes to show that I shouldn’t have assumed the partner was AI-controlled.

(2) Morello talks more about the problem here.

Case Study: Design Flaws In Infinity Blade 3

Now, I don’t do straight game reviews. I’m not usually the sort of person to say that something is “better” or “worse” than something else.

But Infinity Blade 3… I have to say it kind of sucks.

And all the reasons why I think it sucks are practically laid out right here on this blog. Every essay I’ve published here was about some kind of problem I saw in game design. It honestly feels like Infinity Blade 3 intentionally tried to hit every single flaw I’ve written about.

So I guess this essay is part game review, part rant, and part recap episode. Now we can see all the game design theory I’ve written about in action.

The Mana System Paradox: Potions

The Potionmaster at his cauldron.

It’s strange for me to say “mana system paradox” because honestly, I’ve always liked Infinity Blade‘s magic system. To cast magic, you have to draw a little shape on the screen: a zig-zag will shock your enemy with lightning, or a wave pattern will invoke water magic. That’s cool, and I have no problems with it. But when I say “mana system paradox,” I mean “mana” as in consumable resources in general.

IB3 introduced a new element to the series: potions. These potions have a variety of effects, usually boosting your combat ability in the next battle. Some of them will give you additional elemental damage, some of them will heal you every time you successfully block, and so on and so forth. However, they’re only temporary, so they’re more like a one-time burst of strength that you’ll want to save for a boss battle or other tough fight.

Previous Infinity Blade games balanced combat around the character’s level and equipment, but IB3‘s combat has to be balanced around potions as well. They could either make combat engaging without potions and ridiculously easy with potions, or they could make combat impossible without potions and doable with potions (or any point along that scale). However the game gets balanced, it needs to accomodate for potions somehow, and balancing around a consumable resource brings the mana system paradox back into question.

When I wrote the original mana system paradox essay, I was focusing on gameplay and engagement, but potions in IB3 aren’t even at that level. Potions in IB3 aren’t fun to use in any way, but they’re necessary to keep up with the game’s balance levels. Chair designed the game around the assumption that players would be using potions, but if that was their core assumption, why even give players the option to not use potions in the first place?

I can’t think of a positive reason why they introduced potions into the game. Maybe they wanted to get in on MMO-style crafting (potions are either found, or crafted with materials you find). But what would MMO-style crafting even do for Infinity Blade? Maybe I’m just being too cynical, but I honestly have no idea why they added potions.

The Nature of Grinding: Skills

The skill points menu.

The Infinity Blade series has always had a strange relationship between player skill and character skill. On one hand, any battle can be beaten regardless of how strong your enemy is, because as long as you dodge and parry everything you will never die, so player skill is an important factor. On the other hand, battles like that would take hours, so character skill is needed to keep each battle in a reasonable time frame.

Infinity Blade always handled its character skill aspect (part of it, anyway) through a level-up system. After fighting a bunch of battles, your character levels up, and you get to increase his health, damage, shield durability, or magic power. Equipping weapons or armor would increase these same four stats: weapons give you more damage, more expensive shields have more durability, armor boosts your maximum health.

However, the term “skills” took on a new meaning in Infinity Blade 3. Each of the four base stats has five perks that are obtained by investing points into that stat. For example, “Fast Magic,” which charges the magic meter faster, is unlocked by putting ten points into Magic. A lot of these skills are just basic utility boosts that do silly things like make your blacksmith forge weapons faster (I’m not even going to talk about the forging system).

But the problem is that some of the most basic abilities, the abilities that were considered givens in the previous two games, are now unlocked through this skill system. A staple of the Infinity Blade series was the Super Attack, which charges each time you deal or receive damage and can be activated at full charge to stun an enemy. Now, before you can use Super Attacks, you need to put four points into your shield stat. Same thing happened with combo attacks: in previous games attacking in a certain pattern executed a stronger combo attack, but now you simply can’t do them unless you’ve unlocked the skill for it.

Now, the player’s skill level is limited by their character’s skill level. You’re disincentivized from distributing points to suit your own play style, because after a certain threshold they won’t give you perks anymore. I almost never used magic in IB2, so I just never put points into magic and focused on the things I used more often. Now, I need to increase magic because I can only unlock certain kinds of treasure chests if I’ve unlocked the Lockpick skill, which requires points in Magic.

Maybe they did this so that skills would be more dynamic, because leveling up was honestly a little bit boring when it was just “add a few points and that’s it.” And I can see that. But there’s probably an even more dynamic way to rework the system than to have blacksmiths upgrade items faster when you have fifty points in health.

Narratology/Ludology: Awakenings

IB3 has “Awakenings” to replace Rebirths in IB2 or Bloodlines in IB1.

Infinity Blade has as much story as you give it. If you skip through every cutscene, you’re just playing a game where you fight a bunch of things. The narrative and the gameplay are set so far apart that the few moments where they seem related require that you read the Infinity Blade books to understand the context.

The beginning of the game presents a refreshing view: you actually play as the God King Raidriar, who was your antagonist in the first two games. In terms of gameplay, being Raidriar really doesn’t change anything at all. You’re sent on a scripted boss fight against the Worker of Secrets (IB3‘s antagonist) where you lose and you pick up some datapod thing and do a magic trick with it or something, I don’t know. That’s the whole problem: you never know why you do anything in the game.

At one point, you have to go rescue a blacksmith, but the closest thing you get to a reason is that you need his skills for your cause. You can go through the whole game without using the blacksmith’s services if you want. There’s one little sequence where he’s told to forge some thingy, but that scene and the scene where that thingy is used are both cinematic cutscenes completely separated from the actual gameplay. And I still don’t even know what that thingy was ever supposed to be.

It’s sad, because the original Infinity Blade was one of the few games that ever had a narrative connection to having multiple lives. Think about it: when games give you the option to retry after dying, how many of them actually explain what’s going on? There was no narrative reason for Mario to have three lives, it’s not like he ever found some magic trinket that lets him live three times. But Infinity Blade tied it in to the narrative: every time you die as a character, you take control of that character’s son decades later on a quest to enact vengeance for his father. The narrative was poignant, it was simple, and it connected to the gameplay. But each successive game just twisted that narrative more and more.

Gender Roles in Mechanics: Isa

Isa firing her crossbow.

One of the interesting new things about Infinity Blade 3 is the ability to play as Isa as well as the previous games’s protagonist Siris. Isa is characterized as being a speedy, swift assasin kind of person who’s always climbing up walls or doing other sneaky things. Her trademark weapon is her crossbow: she initiates every fight with a few shots from it, instantly taking off a bit of health from her enemies before she engages.

That’s not all there is, however. Remember the skills each character could unlock by adding points to their four attributes? Isa and Siris level up separately and have different skills. Surprise, surprise: Isa gets most of the support skills. Some of their specific skills will work for both characters once it’s unlocked: once Siris unlocks the ability to execute Mega Hits, Isa can too. Likewise, once Isa unlocks the ability to remove gems from items, or the ability to open special treasure chests, Siris can too. Now, which one sounds like more fun: performing Mega Hits, or removing gems from items?

Chair was trying too hard to fulfill the image of a female character: fast, sneaky, supportive, lightweight, magical, etc. Normally, that would be a sign of bad game design, but I actually don’t think it’s that much of a problem in IB3. Hooray, something went right?

Well, no: it’s not a problem in IB3 because it’s not even at the level where it can be a problem. Isa and Siris play exactly the same way. There isn’t enough gameplay difference between the two characters for the gender biased mechanics to make a difference. Everything that makes Isa and Siris different takes place outside of combat, but when it comes down to an actual fight the difference is really just cosmetic.

Both Siris and Isa can choose one of three types of weapons: light weapons, heavy weapons, and dual weapons. What if Isa’s three weapon types were completely different from Siris’s? Maybe she could use her crossbow in combat, and instead of looking for openings to unleash long combos she would whittle down opponents slowly but surely with consistent damage. Or maybe Isa can use Tifa Lockhard-style gloves for unarmed combat, and instead of parrying attacks she can execute grab-style counterattacks at the cost of being unable to block or dodge. There could even be a weapon that doesn’t attack, but can channel to fill Isa’s magic quicker. They had a huge design space and they didn’t go anywhere with it. Although if they had, they probably would have run into problems with gender-biased mechanics anyway.

The Relationship Between Weapon and Wielder: Weapon-Specific Bonus Combos

A bunch of swords with their weapon-specific bonus combos listed next to them.

I actually wrote about Infinity Blade and Infinity Blade 2 in that specific essay, but Infinity Blade 3 screwed up so badly that it needs to come back again. Now, they tried, they certainly tried: one of the things they added was weapon-specific bonus combos. Each weapon can execute a special combo that you can only use when you’re wielding that weapon. Siris’s sword Xwiss is left slash, left slash, right slash, and Isa’s Double Mace staff is downward slash, upward slash, downward slash. Stronger weapons may have longer inputs to memorize as well.

The combo doesn’t actually do much except deal more damage than you would by wildly swinging the blade around, but it does help. Instead of just switching weapons every two seconds without any thought, now you can memorize its bonus combo as well to maximize your damage. It’s certainly a neat idea, and it tries to make the player feel like an omnipotent sword god who understands every weapon he’s ever touched. That’s a cool fantasy to fulfill, and I like it.

But there’s got to be a better way to go about doing it, than to have players memorize random three-hit sequences every time they switch weapons (which is quite often). There are so many ways to personalize weapons, and a random sequence really doesn’t create a meaningful connection because of how quickly your short-term memory fades out. What if a weapon could parry against attacks that are normally unparry-able? And another weapon is stronger when attacking from a certain direction? Even better: take all those potion effects mentioned before, distribute one to each weapon, and voila, you’ve both fixed the potion problem and created meaningful weapons! They probably would have had to have fewer weapons if they used that design philosophy, and I don’t know what design philosophy they were going for, but if they wanted to connect the wielder with his weapons then they should have used something more engaging than random sequences.

Why Layers: Worker of Secrets

Siris killing the Worker of Secrets (right, in golden armor).

So the game was about Siris and Isa trying to defeat a dude called the Worker of Secrets.

1. Why were they trying to defeat the Worker of Secrets? Because Siris was trying to atone for some abstractly-described horrible crime he did in the past.

2. Why did Siris commit that horrible crime in the past? Well, don’t ask me, I don’t even know what he did in the first place!

The game fails a why layer analysis very quickly. This happens for every other game event: you have no idea why you’re doing anything in the game. Why is Isa still hanging out with Siris? Why did Siris lose his memory? Why is there a strange alchemist dude in your hideout who crafts potions for you? Why does everyone hate Raidriar’s guts? Why do Deathless exist? None of these questions get answered in the game.

There are two books in the Infinity Blade story, and that’s two book’s worth of content that should have been expressed through gameplay. If all the essential information is in a book, then people who haven’t read the book don’t get to experience the game the way it should have been experienced. The ending of the game made no sense whatsoever because I had no idea why I was doing any of the things I did.


I just want to emphasize that this isn’t a game review at all. That’s why this is a case study, and not a game review. For all purposes, Infinity Blade 3 is a pretty decent game. This essay isn’t here to persuade anyone against buying IB3.

But I’m still kind of sad. I wish Infinity Blade had really taken advantage of all the design space it had. Instead of focusing on the dynamics of combat, the developers instead forced emphasis into the metagame and stat-based growth. Now, even if someone creates a game with Infinity Blade style combat that has innovative design elements, it’ll just be called a ripoff clone, and the game’s original developers locked themselves out of doing their own system justice because IB3 was the last game in the series.

So I guess this is more like a eulogy than anything else.

Farewell, Infinity Blade. Once, I had loved you, but now I must move on.

Case Study: Randomization In The Binding Of Issac

The Binding of Isaac is about the young boy Isaac, who lived peacefully with his mother in a small house on a hill. One day as his mother was watching Christian broadcasts, she heard a voice from above: “Your son has become corrupted by sin. He needs to be saved.” Quickly, Isaac’s mother removed all the evil toys and games he played with, before they could stain his soul. But the voice came again: “Isaac’s soul is still corrupt. He needs to be cut off from all that is evil in this world and confess his sins.” Following this advice, Isaac’s mother locked her son in his room to protect him from outside influences. Still, it was not enough: one last time the voice called out to her.

“You’ve done as I’ve asked, but I still question your devotion to me. To prove your faith, I will ask one more thing of you. To prove your love and devotion, I require a sacrifice. Your son, Isaac, will be this sacrifice. Go into his room and end his life as an offering to me to prove you love me above all else.”

Isaac, watching through a crack in his door, trembled in fear as he saw his mother coming with a butcher’s knife in hand. As he desperately searched for a way to escape his fate, he found a trapdoor to the basement hidden under a rug. His mother burst through the door: he had no other option but to leap into the unknown depths below that trapdoor.

The Binding of Isaac isn’t a game about religion: it’s a game about how religion is experienced by those who aren’t a part of its world. Isaac is one of those people, terrified at the thought of being sacrificed for a concept he never understood. As he explores the strange psychological depths of his house’s basement, he spirals deeper and deeper into confusion, never managing to find an answer. The basement represents Isaac’s psyche, using the game’s depth as a metaphor for Isaac’s futile attempts to understand God’s ways.

Initial Attempts As Cumulative Knowledge

This is the death screen. You see it quite a lot.

When first beginning to play the game, a player will inevitably die many deaths in the game’s early and middle stages. The game does not save sessions, so the player’s ability to progress relies on his knowledge of the game’s mechanics. Eventually, the player develops an understanding of how the basement works (or rather, as good an understanding as mere humans can develop) and starts penetrating the dungeon’s depths deeper with each playthrough.

A key aspect of the game is the use of various items. As Isaac progresses through the basement, he finds many different kinds of collectibles. Some of these are pills that have effects ranging from positive to negative, unknown until revealed through retrospect. In addition, there are trinkets that can be held to grant a passive effect: sometimes the effect is obvious, whereas other times it may seem completely useless. Other items like tarot cards have a one-time effect that is again unknown until tested.

Progressing through all of these mysterious items can be a harrowing, mind-bending experience. The tarot card “II The High Priestess” sounds like it would have some kind of beneficial effect, but when used causes a giant foot to crush Isaac and immediately kill him. A pill which damages Isaac when taken can inexplicably transform into a pill which fully heals him when he’s low on health. Items which teleport Isaac into a random room can sometimes teleport him into a glitch room filled with unobtainable treasures, as if the game’s own teleportation mechanic somehow broke itself. Sometimes an item will do exactly what Isaac needs most, but other times it will just outright kill him. Trying to understand all of these effects is difficult, and trying to use them in a practical context is even more so. Is the basement trying to kill Isaac, or somehow protect him? Or is the basement controlled by some whimsical higher power who enjoys watching Isaac scramble helplessly in confusion?

At the start of the game, the player has no idea what any of these items do. But with each foray into the basement, the player deepens his understanding of the game’s various items through trial and error. When a player discovers that “II The High Priestess” can be aimed at enemies, it suddenly becomes a valuable offensive option that enables him to defeat a boss more quickly and thus conserve more health for the adventure. The basement is like a well of knowledge, and the player’s attempts to better understand it parallel Isaac’s attempts to better understand the circumstances that led to his situation.

However, it’s not a positive journey. The sheer depth of the basement’s content means it’s impossible to ever fully understand it. All content is placed randomly in the basement, so the player is constantly assaulted with strange new items despite the knowledge he accumulated over previous playthroughs. Not only that, but the interactions between items change drastically: that glitch room might never appear again. In the end, the basement is never truly able to be “understood” because of how much content there is to explore, and how it’s presented in a randomized order.

Symbolism of The Basement

The “reward” that you find at the “chest” at the “end” of the “basement.”

Even putting aside the mechanics as metaphor, the basement itself is still a mystery. It’s obviously not a real basement: real basements don’t have strange spiders wearing the face of a bleeding child in them. So where exactly did Isaac end up when he leapt through the trapdoor? “Where” might be an inappropriate word to use: it becomes very clear that the basement is not a mere physical space. If Isaac travels far enough, he’ll enter areas like his mother’s womb or Sheol, the land of spirits. Discovering the paths to these various floors is another trial-and-error business just like the game’s item system: sometimes you must hold a specific item to enter an area, or sometimes an area will unlock after clearing the game a number of times, or sometimes an area will just not appear at all.

Different endings will be shown depending on which area Isaac was in. Of course, none of these endings make any sense: the basement is not something that can be understood by simply completing the game a few times. However, a recurring item throughout nearly all of the endings is a giant golden treasure chest. Sometimes, Isaac will open the chest and find an item inside, which is then added to the list of things that will appear in the basement, continuing the theme of building off past experience. Sometimes, Isaac will crawl inside the chest and shut the lid (and sometimes in subsequent playthroughs, Isaac will find the past version of himself who had been locked inside). Sometimes, Isaac will just stare sadly at the treasure chest. But the recurring nature of the treasure chest raises an important, if unanswerable question: just what exactly is inside it?

If a player wrestles with the basement enough, he may discover the last area of the game: the inside of the golden treasure chest. Clearing this area triggers the final ending, which reveals that the chest contained a series of family photographs. Although these photographs tell a disjointed, confusing story, they still show scenes from happier times before Isaac’s mom began trying to kill him. Despite all of the confusion and terror and randomness of the game, the treasure chest’s contents reveal that Isaac was only ever driven by the desire to have a loving family and a mother who wouldn’t try to kill him. That fact is quite possibly the only fact in the game.

After reaching the photograph ending, the game displays a giant “THE END” image on its title screen, bringing the game to as much of an ending as it can possibly reach. But the ending isn’t just about narrative: Isaac himself attains a small form of closure as a result of his journey. If the player managed to fulfill a difficult and unknown condition, from then on Isaac starts his adventure holding an item, “The Dice.” Using the dice rerolls all items on the screen: if Isaac finds an item he can’t use or doesn’t want, he can reroll it and hope the basement gives him something better. It’s not a lot, but in the end Isaac managed to gain a tiny, tiny degree of control over the basement’s wild randomized insanity. Did he come to terms with his family situation? Did he receive a revelation from the photographs? Did he accept his gruesome fate as a sacrifice? There still aren’t any definite answers (and there probably never will be), but by the end of the game both Isaac and the player managed to develop at least some degree of understanding over that which is impossible to understand: the will of God as channeled through Isaac’s basement.

In theology, the story of Abraham and Isaac has been fairly thoroughly studied. However, all of those studies focused on Abraham: he was being tested by God, or he was being punished for sinning, or all sorts of other interpretations. But how did Isaac feel about the whole ordeal? The Binding of Isaac not only answers that question, but forces the player to experience the same emotions Isaac felt under the knife: terror, confusion, and hesitation. The player’s journey through the basement is a journey through Isaac’s tumultuous psyche in the moment between life and sacrificial death.