The Game Design Forum

Part One Continued: Action Games and their Impact on Diablo 2

Diablo, the predecessor to Diablo 2, began its life as a fairly orthodox roguelike. While the project was still early in development, the main Blizzard team which oversaw the project told David Brevik that the game engine should work in real time. This decision changed the course of development in a big way, and introduced many important game design ideas that are not native to the roguelike subgenre. Because it started as a roguelike, Diablo’s relationship to its roguelike source material is fairly straightforward; it adapts roguelike concepts for a mass-market audience. The relationship between Diablo 2 and its action-game ancestors is more complicated and theoretical. One of Diablo 2’s greatest achievements is the way it recreates the fundamental structure of action games (Nishikado motion) through procedural means. This section will look at the action heritage of Diablo 2 and explain the way it reinterprets that heritage through an RPG lens.

One issue I want to tackle before looking at the history of the action RPG is how to define the level-up. Critics and fans have argued about the point at which an action game has enough RPG content that it becomes a true action RPG, instead of just being an action or action-adventure game. (David Brevik himself pondered this question in another interview [25] without coming to a conclusive answer.) For example, many question whether The Legend of Zelda games are really RPGs. I say that they are, because Link can gain more health, magic, strength, etc. over time. These gains are usually not connected to experience points, nor are they part of explicit “character levels,” but does that mean the game is not an RPG? For me, Zelda is still an RPG, because these things are still level-ups. A level-up is a permanent, periodic increase in the power of a player-character. When Link gains more heart containers, his maximum life is permanently increased, and that permanency makes it a level-up. When he gains the use of the Master Sword, it represents a significant increase in power over the starting sword, and he has it for the rest of the game. These gains are all connected to finding loot, but as we’ll see many more times in this book, loot is often the most important kind of level-up. Link makes one or more of these gains in every major dungeon. The period is simply measured in dungeons rather than being measured in EXP. Whether or not you agree, this definition is enormously important to the rest of this book.

The Birth of the Action RPG

Once RPGs started to appear on PC and home videogame consoles, it was inevitable that design ideas would leak between the booming world of digital action games and the older world of tabletop RPGs. For designers of RPGs, this was a boon, for it gave them another way to escape from the overshadowing influence of D&D. For all of its systemic completeness, D&D never implemented the kind of real-time dexterity challenges found in action games like Space Invaders, Pac-Man or Galaga. (It’s difficult to imagine exactly how it could do so, to be fair.) By implementing action-game challenges against a background of RPG systems, designers could combine their favorite parts of RPGs with any of the various action subgenres. The result would be a hybrid game that felt significantly different from either of its parents. I call this the combination strategy, which allows designers to disguise relatively old RPG tropes in the clothes of an action game.

The combination of RPG and action is not only for the action gamer’s benefit, however. Designers often found that action game skills could be developed right alongside RPG stats over the course of a game. Demanding higher levels of thumb-skills from the player is a great way to defeat one of the RPG’s greatest problems: power creep. From the very beginning of RPG history, clever players have found ways to exploit the vast, interlocking systems of RPGs to become unstoppably powerful. That exploitation can be mitigated, however, if the player cannot execute many of the action-game skills required to take advantage of it. Diablo 2 isn’t designed to demand a ton of dexterity from its players, but they absolutely need better command of the action skillset to survive their first trip into the final difficulty setting. That said, what Diablo 2 really gains from action games is the revolutionary challenge curve which they invented. At its most fundamental level, Diablo 2’s difficulty structure looks more like that of an action game than that of an RPG, even though it leans more heavily on the latter genre in terms of mechanics and goals.

Two Difficulty Curves

The first game that is still relevant to the history of action game design is Space Invaders (1978). The designer of that game, Tomohiro Nishikado, was also the engineer responsible for the construction of its arcade machinery. Because of an error in the way he configured the game’s chipset, the enemy aliens move faster when there are fewer of them on screen.

This causes each level in the game to become more difficult as the player progresses through it. Rather than correct this error, Nishikado kept it as a design feature. Indeed, he embellished it by making each successive level start at a slightly higher level of difficulty, thus reiterating his serendipitous difficulty structure at the macro level.

I call this up-and-down motion in videogame design Nishikado motion. Nishikado motion is the fundamental structure that underpins nearly all of mainstream videogame design. Even today in the era of extensive player psychology research and interest-curve diagrams, Nishikado motion remains the central pillar of mainstream videogame design.

Difficulty in traditional RPGs is structured very differently than it is in mainstream videogames. Games of all kinds teach the skills that players need to learn in order to overcome later challenges. In action games, players can often fail to learn, might only partially learn, or might even forget the dexterity skills which a game teaches. Players of action games can also become literally exhausted by the demands of a fast-paced action game, and their hand-eye coordination can suffer after levels as short as a few minutes. Nishikado motion accommodates this by routinely delivering lulls in the intense action, to let the player regain their stamina and catch up on skills they lack, or which they have forgotten. Traditional RPGs are different. In an RPG, the player’s character does most of the learning. It’s the character who learns how to pick locks, cast fireballs, parry sword-strikes, and speak with animals. The player just has to declare the action and roll dice. Thus, RPGs can afford a more linear difficulty curve. What’s more, because most things in a traditional RPG depend on statistics, it’s actually a lot easier for RPG designers to precisely tune their difficulty curves. This is not to say that all RPGs are perfectly linear in terms of difficulty. Even the best RPGs are inconsistent from quest to quest if the player hasn’t made certain choices, but the overall emphasis on undulating arcs of difficulty is less prevalent (and less necessary) in tabletop-style RPGs than it is in console videogames.

Although traditional RPGs are probably a little easier to balance than action games of equal scope, certain RPGs are harder to balance than others. Diablo 2, like most games that have some roguelike DNA, has occasional balance issues stemming from its widespread use of randomness and procedural generation. Diablo 2’s volatility is a little different from its roguelike counterparts in that there’s a much greater emphasis on positive variation. That is, in Rogue, Angband and Moria, lucky players will barely finish the hardest challenges. In Diablo 2, lucky players can quickly become unstoppable demigods. This is an intentional design decision, one that shaped that everything in Diablo 2, from the construction of monsters to the way that loot tables work. One of the most interesting things we’ll see about Diablo 2’s difficulty, however, is that it is very similar to the up-and-down rhythm of Nishikado motion. It’s not exactly the same, but it’s clearly analogous. I call this type of difficulty curve Schaefer variation, after the Schafer brothers, one of whom (Max) first explained the phenomenon to me, and who continue to use the structure in the Torchlight series. (I want to note here that Brevik and designer Stieg Hedlund also had a considerable role in shaping this design structure; it was clearly a group effort. Moreover, Brevik has toyed with the dynamic in his later games as well.)[26][27] Schaefer variation is simply Nishikado motion accomplished through procedural means. In part three of this book, we’ll take a look at the exact details of how this happens, but the idea of Schaefer variation is an important one throughout this book.

The Action Parts of Diablo 2

The influence of the action genre runs much deeper in Diablo 2 than merely theoretical ideas. Diablo 2 is definitely an action game, and its action mechanics shape the player’s experience in ways that separate it from other games in the same genre. No two action RPGs are the same in the way that they balance and mix their action and RPG components. For example, the most famous action RPG series of all time, The Legend of Zelda, depends heavily upon the non-combat applications of its action mechanics. Whether it's hitting switches with a boomerang, leaping chasms with a hookshot, or lighting a series of torches before the first one burns out, there are many tasks which depend on player dexterity and sense of timing but do not involve combat. During those moments, the game almost moves entirely into the action genre. Thus, we could say that Zelda games depend a lot more on their action components than Diablo 2 does. Diablo's action mechanics are almost entirely built around combat. In the moments when they’re not fighting, by contrast, players of Diablo 2 are doing things like crafting or managing gear, which is a part of the game that depends almost entirely on RPG systems. The only action mechanic that the player uses outside of combat is enhanced movement for exploring dungeons, like Teleport, Leap or Burst of Speed. Then again, all exploration in Diablo 2 is frequently interrupted by combat. So when we talk about the role of the action genre in Diablo 2, we're talking about combat.

Diablo was originally conceived as a traditional, turn-based roguelike. When the game was altered to operate in real time, many of its fundamental RPG mechanics had to change as well. Traditional RPGs take place in discrete turns. Each player-character or enemy gets a chance to act in the order that their simulated skills and chance allow. During their turns, each character can move once and attack once, with some small variations. What happens to this time-tested procedure when real time is applied, and everyone is taking their "turn" at once? All those mechanics either need to be adapted or dropped. For the most part, Diablo 2 adapts them to real time in straightforward, intuitive ways that place most of the emphasis on simulated skills, but there are some action-specific adaptations as well. The following section will analyze each of the ways that Diablo 2 adapts RPG ideas into an action context, and the implications for the game as a whole. Some of this material was briefly covered in the character class section, but here we'll spend more time on the exact mechanical operation of action abilities with an emphasis on quantitative analysis and diagramming.

Movement

Most movement in Diablo 2 is uncomplicated. All characters move at a speed measured in yards per second. Characters run 50% faster than they walk, but while running, they have no effective defense rating. There are lots of items which grant increased movement speed, especially unique boots and magical charms. Many expert players seek to max out their movement speed through these items as an essential part of farming/leveling runs. Like many of the things that these players seek to max out, these bonuses suffer significant diminishing returns as the character nears maximum speed.

[28]

The gains in speed flatten out quickly. Diminishing returns on real-time effects is a theme we'll be revisiting often in this section. While many ultra-hardcore players still insist on having maximum speed, these diminishing returns are in place to encourage players to seek a more balanced approach to their gear. That said, speed bonuses granted by abilities (like Vigor or Burst of Speed) do not experience diminishing returns when combined with other speed bonuses—although the skills do grant lower bonuses per point invested in them. But I will discuss movement abilities in their own section, below.

MONSTER MOVEMENT

Monsters in Diablo 2 mostly follow the same rules as players. They have set run and walk speeds, although these can be increased by affixes and auras. Some monsters can leap and teleport, much like player characters can, although they appear to wait much longer between uses of these abilities than players do—probably because of a special timer built into their movement behavior, or else every battle would be nearly as difficult as player-vs-player combat. There are several monster affixes which enhance their speed as well. Champions and Fanatic enemies move faster than normal, and any unique enemy with the “extra fast” affix will move at double speed. Because these speed bonuses are based on abilities rather than gear (which enemies don’t wear), the enemy speed bonus does not experience diminishing returns, and so some of the fastest enemies in the game can become terrifyingly fast if they happen to fall within a movement speed buff of a procedurally spawned enemy. This is one of the ways which we’ll see (in part three) how the difficulty of the game can suddenly spike thanks to unfortunate procedural generation.

There are two modes of enemy movement that the player doesn't have: flying and phasing. The Scavenger class of enemies can fly over the battlefield, immune to all attacks, before landing to strike.

This is the only family of enemies which moves like this, and it gives the enemy minimal advantage, as they cannot attack while in flight. Similarly, Black Souls (bane of Baal runners) can phase out of reality, becoming become hard to see (and therefore hit) while moving.

The Black Soul can be exceedingly dangerous because of how quickly it can phase in, take a shot at the player, and then phase out again. Although many movement abilities (like the Cave Leaper’s jump) are annoying to deal with, this is the only one that makes an enemy significantly more dangerous without a procedurally added affix.

PLAYER CHARACTERS AND THEIR MOVEMENT OPTIONS

With one (glaring) exception, melee classes have the best special movement options. Like basic movement, these abilities are not especially complicated. The rules are as follows:

  1. The player-character must have line of sight to the target location
  2. The ending location of the ability must be navigable (not a wall or a pit)
  3. There must be no impassable obstructions between the character and the ending destination

These rules are fairly straightforward. The first two apply to all movement skills equally; the last rule varies in its application from skill to skill. Ground-based movement abilities are subject to enemy attacks, and can be stopped as such; a leaping barbarian does not have the same problem. Ease of movement isn't the only factor in the utility of a movement skill, however. Other things like movement speed, movement distance, mana and time costs, and impact on enemies all affect the usefulness of an ability. The paladin and barbarian movement skills have reasonably strong attacks built into them but aren't terribly fast or far-reaching. The assassin's Burst of Speed ability is easy to sustain indefinitely and includes an attack buff, but is easily stopped. The sorceress’s Teleport is twice as expensive a other movement skills, and has no effect on enemies, and it leaves the caster somewhat vulnerable. And yet, Teleport can move the player across the landscape faster than any other skill in the game, and doesn’t collide with terrain objects except map edges. Many of the game's most popular items (like the coveted Enigma runeword) offer the Teleport skill to non-sorceresses as their primary benefit. The skill is simply too desirable to pass up.

There are three great object lessons about game design that we can learn from the way that movement abilities in Diablo 2 are configured. The first lesson is that, if you're designing an action RPG, you can bet that players will find ways to employ action solutions to your RPG problems. One theme common to action games of all genres (whether platformers, survival horror or anything in the open-world format) is that players eventually learn to avoid most enemies. Many enemies offer either no reward, or only a trivial one. In a game with powerful running and jumping abilities, it is easy (and sensible) for the player to avoid all of these encounters. (There is an entire genre of YouTube video where players do this for long stretches in Dark/Demon's Souls games.) And sure enough, that's what experienced players in Diablo 2 do. Whether it's rushing to Baal or Mephisto, or running around Lower Kurast looking for the infamous super-chests located there, players use their character's movement skills to exploit all manner of shortcut and AI patterns in order not to fight enemies. This is yet another area in which the Teleport skill is overpowered; it deals no damage and has no stun or knockback effects. It doesn't need them, though, if the player is simply avoiding combat.

That last point makes for a good segue into the second lesson, which is that qualitative differences are not the same as tactical balances. Both of those things are necessary parts of a good class system. In the previous three Reverse Designs, I made the case that qualitative differences in videogame design are greatly underappreciated. I maintain that stance here, with one important caveat: qualitative differences only matter if the quantitative differences don't make them obsolete. If there are two paths through a game, and one of them is much harder than the other without any commensurate reward, it doesn’t matter if they’re qualitatively different. Players are going to choose based on difficulty. The same thing applies to RPG abilities. It doesn't matter how great the barbarian and paladin movement skills feel—though they do make you feel like a badass when you use them right—if they lag way behind Teleport in speed and range. It's tempting to counter this with "but sorceresses are fragile!" That's correct, but remember that we're talking about mobility and avoiding enemies here. Fragility hardly matters if the player is deftly zooming past anything dangerous. In order for players to be able to choose freely among qualitatively different options, those options must be roughly equal in effectiveness.

The third lesson we can learn from movement skills in Diablo 2 is that there are always lots of creative solutions to game design problems, but those solutions need to make sense in the context of the game they apply to. The problem with movement in Diablo 2 is that Teleport is too powerful, and the designers knew it. But also, Teleport is far too resonant with the sorceress class—and really, the game in general—for the designers to remove or nerf it too harshly. Their solution was to offer Teleport to any class on a variety of items like amulets, staves, or even armor. This is the quintessential Diablo 2 strategy. The design philosophy of Diablo 2 includes the tenet that any character should be able to become an overpowered demigod. Teleportation is a part of that strategy, and inasmuch as any class can gain Teleport, it makes sense in the context of Diablo 2. So rather than nerf one class and its abilities, the designers simply make the other classes more powerful by addition.

Crowd Control Skills

Only a few classes have skills which greatly enhance their movement. All classes, however, have the reverse, skills which make enemies slower or stop them entirely. We've already covered some of this in the section on character classes, but here we're going to pay more attention to the particular action dynamics of the various crowd control abilities in the game. One of the most common problems in Diablo 2, especially for inexperienced players, is becoming surrounded by enemies.

In large numbers, these enemies can even kill player-characters whom they ordinarily could hardly touch. What can a player do if they lack the escape skills like Teleport or Leap—or can't use them because the crowd of enemies is too big? The answer is, of course, crowd control skills, which either restrict or redirect the movement of that enemy mob.

The first thing to understand when it comes to crowd control dynamics in Diablo 2 is how the enemy AI controls those crowds. Brevik and Hedlund both explained that, except for bosses, AI in the game is very rudimentary.[29][30] Most normal enemies can idle, move or attack. Some enemies also have a simple evasion script, like the Dark Archer, who will flee the player to perform a ranged attack, or the Cave Leaper, who only comes into melee range to make brief hit-and-run strikes. Aside from these occasional, small variations, common enemies operate in the traditional videogame herd. A herd is any group of enemies whose primary movement pattern is to simply collapse on the player. In the diagram below, the red circles are enemies, and the blue one is the player.

No matter what shape a herd begins in, the enemies in that herd will collapse to form an amorphous blob around their target. The player can take advantage of this in a few ways, with or without crowd control abilities. With a few exceptions (Ghosts, flying enemies), enemies cannot move through one another. Because their AI is so rudimentary, large groups of enemies of roughly similar speed will slowly get stretched out in a path moving toward the player. The herd and its three basic shapes (initial placement, amorphous blob, and stretched-out) are an important to the way that crowd control works, as we’ll see below.

CATEGORIES OF CROWD CONTROL SKILLS

There are three categories of crowd control skills: slows, repels and distractions. All of these are useful, but they have different effects on the real-time crowds that the game hurls at the player over and over. Slows reduce the movement of enemies in a group. (Note that I include all stuns in the slow category because the crowd-control impact of stuns is essentially the same as those of slow effects.) The most prevalent example of the slow-type debuff is the chill effect from cold damage.

Most cold damage slows an enemy by about half, although this can vary based on the source and the target in question. Because all cold attacks apply this debuff, cold damage is weaker than other elemental damage of the same level. For example, the level-35 affix for cold damage on magical weapons adds an average of 25 cold damage to an attack, while the fire-based suffix of the same level, adds 56 damage.[31] There are also a few cold attacks, like Frost Nova, which trades away 75% of its damage (compared to same-level spells) for three times the freeze duration. This is merely an extension of the same principle which underlies all cold damage. Similarly, some cold attacks can freeze enemies solid. Like all stuns, this is just a more extreme slow. But it also trades away even more damage to achieve the highest level of crowd control.

Slow effects can also be administered through a necromancer curse, or through a weapon affix which appears on unique and set items. Obviously, only the necromancer can use curses, but the slow applied by items is also limited. As with most slows, the classes which benefit from the weapon-based slows are ranged. By slowing a herd of enemies, the player gains separation from them, and can subsequently pummel them from a distance. So it's a little strange that of the unique weapons which have a slow effect, 10 are melee weapons and only two are ranged. (Ranged weapons have yet another small drawback.) Nevertheless, curses and weapon-based slows stack additively with cold and with each other, and can greatly reduce enemy movement speed. There is a cap to the amount of slowing that an enemy can receive (an 85% reduction in speed or one yard per second, whichever is higher), although nothing in the game indicates this.

Regardless of the type or degree of slow inflicted, there are specific action-oriented effects of using this type of crowd control. In the context of small groups of enemies, slowing effects are simple and have predictable impacts on the herd.

Slow effects delay the threat from one or more of these converging hordes. In the small groups which populate many narrower areas, this works great. But not all crowds of enemies are so small or move in toward the player piecemeal. Some herds number in the dozens and come in an indistinct blob. These groups have different dynamics when it comes to slow effects. With a few exceptions (ghosts, teleporting enemies, flying creatures), monsters cannot move through one another. Thus, enemies under the effects of a slowing debuff become a natural barrier, while giving a new shape to a large collapsing herd.

Shooting a cold ability or other slow into a crowd can actually cause secondary problems if the radius of the slowing effect is not large. The enemies will split up to come around their slowed comrades, and in doing so, spread out. For a ranged caster who's kiting a herd (leading a herd while firing at a distance), this creates more diffuse targets, requiring more spells or more kiting. This is not a huge problem, but it's one that could be solved by the right shape of attack. The "Cone of Cold" is an attack dating back into the tabletop years, and cone-shaped attacks are plentiful in the action game heritage. The lack of such an attack in Diablo 2 is a little strange. Diablo 2 does use a few cone-shaped and linear attacks, but none of them are cold-elemental. It's not a game-breaking deficiency by any means, but it is a curious omission.

REPELLING ATTACKS

The second type of crowd control attack is the repelling type, which does not (usually) slow or stop enemies, but which forces them to move away from the source of the attack. There are four characteristics which define a repelling skill: shape, source, distance and duration. Some repelling skills, mostly those which are connected to melee combat, are radial in shape.

The barbarian's Howl skill is the best example of this, but certain weapons and armor (the unique helm Howltusk, for example) also repel enemies within a small radius of the character who equips them. Weapons only cause one monster to flee at a time, but the AoE abilities that use this kind of repelling attack don’t allow players to be selective in which enemies they target. All nearby enemies will run away. This seems like it would hurt melee attackers who need to get in close to at least one enemy, but this problem is somewhat mitigated by the fact that radial repelling skills don't have a huge area of effect, nor a long period of effect. Duration and distance are an important part of repelling skills. The small window of time and space afforded by radial skills like Howl give the player just enough time to target the most dangerous enemy of the group with an engage skill like Leap Attack. This is useful for sorting through crowds more than it is for getting out of them entirely.

The other major shape for repelling crowd control abilities is the linear shape. This type of crowd control includes most knockback skills, which target either one or a small group of enemies which are knocked away from the attacking character in a straight line. Because of the low number of targets on most of these skills, this type doesn't control "crowds" quite as much as it controls individual members of that crowd. Knockback skills tend to be cheap and high-damage relative to other crowd control skills. Knockback is also usually the property of an attack rather than a spell. So, on the one hand, it can miss its target. On the other hand, it is one of the few types of crowd control that can easily be combined with other types. Even low-level gear and abilities can grant a clever player the means to both knock back and chill his or her enemies. Indeed, this is a good way to increase the effective duration and distance of knockback. Those two dimensions are an important part of all crowd control skills, but knockback offers little separation and very little duration. Thus, those combinatoric possibilities are key to the knockback's long-term utility.

OUTSOURCING THE SLOW AND REPEL

Slow and repel attacks have different dynamics if the source of the skill is not the player-character's body. The clearest example of this is necromancer curses, which can inflict a variety of crowd-controlling debuffs, like slow. This spell is in the typical radial shape, but it's centered on the player’s mouse click, rather than on the character’s body. This affords the player a lot more discretion in changing the shape of a collapsing herd, and also allows players to use the ability without having to wait until enemies are on top of their characters. Similarly, the barbarian can use a radial repel attack, Grim Ward, which can be placed on the battlefield. Although its radius isn't very large, the ability to place it away from the barbarian gives the player an advantageous angle on enemy movements, and the totem can even obstruct passageways in dungeon settings. The necromancer's Bone Wall is similar in effect, but linear in shape.

Knockback is perpendicular to the player while the linear shape of Bone Wall runs, when placed, parallel to the player. Although Bone Wall blocks more than it repels, it nevertheless has largely the same bottleneck-creating effect as Grim Ward when used in the right spot. Putting it in the right spot is fairly easy, though, because the source isn't the player.

ENEMY INFIGHTING

Certain abilities can cause enemies to target members of their own group with attacks. The notable examples of this type of crowd control are the assassin skill Mind Blast and the Necromancer curse Attract. If you don’t consider monster stats, these crowd control abilities are balanced sensibly. Both Attract and Mind Blast have small effect radii compared to other debuffs. Attract does not work on boss mobs; Mind Blast only has a low chance of controlling enemy monsters. When used, they still distract enemies from attacking the player, and they still create the wedge around which unaffected monsters have to walk, although the size of the effect is noticeably smaller. All of this seems like a fair tradeoff if the debuff is going to cause monsters to deal damage to one another, taking some of the burden off the player.

The problem with the balance of enemy infighting abilities is that enemies deal very little damage compared to the amount of HP they have. In fact, you can barely see the HP if you put it on a graph with damage.

[32]

This dynamic exists because Diablo 2 puts much more emphasis on killing power than it does on character survivability. The most HP a player character could reasonably accumulate is around six thousand (without having to sacrifice every other stat). A damage-focused character with the same amount of time invested might be able to output four or five times that much damage per second. Thus, enemies have tons of HP on the harder difficulty settings, but deal relatively little damage. Any skill which causes a monster to damage itself or another monster just isn't going to be significant. These skills do a reasonable job of keeping enemies occupied, but Mind Blast doesn’t last very long, and Attract is overwritten by any curse which might actually help the necromancer deal damage to those targets. Other crowd control skills are simply more useful.

Key Lessons from Crowd Control

Crowd control skills in Diablo 2 teach a clear and widely applicable lesson about action skills in an RPG. Although all the crowd controls are meant to be different—meant for different characters and contexts—certain characteristics that they share can make them more or less powerful. For example, the shape of a crowd control skill doesn't matter very much. As long as a linear crowd control can affect more than one enemy (as Bone Wall does), and the player can aim it well, it can be as effective as a radial crowd control (like Psychic Hammer). On the other hand, the source of a crowd control skill affects its utility immensely. Howl is a useful skill that sends enemies running away from the barbarian, but Grim Ward and Terror can do the same thing at essentially any point on the screen, offering the player the same effect with more versatility. Finally, longer durations are a huge asset to crowd control, but are also the first target of balancing efforts. The most common crowd control effect, chill from cold spells, is also the one to which all enemies gain resistance on higher difficulties. The chill effect in hell difficulty lasts for only a fraction of the length it does on normal. Less common sources of crowd control effects don't suffer from diminished effect on higher difficulties, which is a pretty typical balance dynamic for an RPG. Rarer skills and items tend to be better than common ones.

The UI Problem

For all the nuance and complexity built into the action abilities in Diablo 2, using those abilities is occasionally difficult because of the antiquated UI. I want to say up front that in its own time, Diablo 2 offered a slick player experience. The part of the UI which matters the most, targeting enemies with skills, works quite well. Right clicking to target with a spell is natural and has the precision of mouse movement. Moreover, hitboxes in Diablo 2 are on the generous side for most monsters.

That said, sometimes certain terrain doodads also have these larger hitboxes and can block shots you wouldn’t expect. Moreover, in large crowds, it’s hard to target a single enemy because of the overlapping hitboxes. Some of this is probably intentional; it should be hard to hit enemies in large crowds. And in any case, it’s not the kind of game where it matters which enemy the player hits first—at least not often.

The real problem with Diablo 2's UI is how difficult it is to switch between skills. We just saw two whole sections about the many action applications of movement and crowd control abilities. It's not unusual at all for a good player to switch between a movement ability, a primary attack and a crowd control skill, or even two of each. But how should the player switch between them? The big limiting factor is that the player's primary hand cannot leave the mouse. For people who don't have a deluxe gaming mouse, that leaves five fingers on the other hand that the player can use. Four of those fingers are dedicated to potions most of the time. The thumb can handle the item drops or the “hold still” command (with some button remapping). But the designers found themselves without a convenient solution for switching between a variety of skills. They went with the obvious (but somewhat inconvenient) solution of using scroll wheels or the function keys.

The lower part of the QWERTY keyboard is standardized, so the designers can design and playtest for it without having to do extra research. Everything above the standard QWERTY layout and letters can vary significantly across different models and eras. The space between the number row and the function keys (which can be assigned to abilities) can vary between different models of keyboards. The alignment of the function keys can vary. The size and texture of the function keys can vary. The average human hand can't comfortably reach more than two function keys without having to move the entire arm a little bit. The muscle movements of the arm (especially the non-dominant arm) are not nearly as precise for most people as those of the hand, and can cost precious milliseconds while the player is in combat. What's more, the player cannot look away from the screen because the game is so fast-paced.

There are quite a few problems with the configuration of Diablo 2's keyboard controls, but none of them are actually fatal to the game. With some practice, players can either reprogram their keyboard or learn to use the first few function keys reasonably well. The player can also use the scroll wheel on the mouse to cycle through abilities, but that isn’t as precise as analogous controls in other action games. Because Diablo 2's systems are so fun to use, players will adapt, but they shouldn't have to. One of the few things that Diablo 3 unequivocally improved over its ancestor was the design of the hotbar. In that game, the player has easy access to six abilities and a potion by only moving their fingers. I'm ambivalent about the change in potion consumption rates (and health systems in general) in Diablo 3, but some kind of change had to happen in order to facilitate easier ability use.

I realize that it's somewhat unfair to criticize Diablo 2 for not inventing keyboard setups which didn't exist at the time. The design of action keyboard maps and quick UIs were largely driven by the FPS boom of the early 2000s. Millions of dollars of playtesting went into the creation of those games, and each one learned from the errors of the games that came before it. I still point out the problem, however, because it's important to the history of action RPG design. Perhaps the exact changes which Diablo 3 used to allow players to mix abilities weren't perfect, but it should be clear that they were necessary, and why.

Looking Back at Action Mechanics in Diablo 2

Although Diablo began its life as a turn-based roguelike, by the time of Diablo 2 the designers had put a lot of thought into the implications of including action mechanics in their game. Special abilities for crowd control and character movement show that the design team didn't just import traditional role playing ideas into the action space without thoughtful adaptation. Not only do the movement abilities of each class feel different, but they also match the design philosophies of the characters that use them. Whether it's chilling enemies, stunning them, knocking them back, scaring them away, putting up walls or even causing enemies to fight amongst themselves, there are many ways for the characters to deal with dangerous swarms. None of these abilities are merely token efforts, either. If used correctly, each class's abilities are useful and distinct from what the other classes have. This brings up another important aspect of the influence of the action genre on Diablo 2. When a game mixes genres well (what I have elsewhere called a composite game), it is possible for the player to use the mechanics of one genre to solve the problems of the other. Good action mechanics in Diablo 2 can absolutely make up for statistical shortfalls, especially in the early game. Good use of kiting and crowd control can allow fledgling casters to over-perform their early damage and mana shortages. Movement skills, meanwhile, can help melee classes target the most dangerous members of enemy packs and eliminate the unique and champion enemies who present a real threat, early in the game. Whereas a good command of the action mechanics is nice to have early, it becomes essential in the late game. Bad action mechanics can cost even well-built characters some deaths if the player strays carelessly into situations no character can reasonably survive. Action mechanics could easily have been a place where the designers simply added real time without any depth, but they took the time and care to make sure that the action part of the Diablo formula shines through.

Character Classes - Back | Next - Randomness in Diablo 2

Site questions? webmaster@thegamedesignforum.com

All material copyright by The Game Design Forum 2017