Stinger Magazine is a site fully dedicated to giving analyses and critiques on games in the action genre.

27 July 2018

Enemies and their Design

When talking about games, no matter the genre, the protagonist takes center stage. What he can do, his jumping skills, the way he cancels one attack into the next, how he upgrades his weapons - all the tools the player has at his disposal. But in truth the main character, be it Nero, Ryu, Mondo, Sieg or even Juliet Starling, are only half of what the gameplay offers... at best. The other half is often ignored, rarely discussed and left in the shadows. So today we shine a little light on this ‘other half’, namely: the bad guys. The enemies, the mooks, the villains, the dirtbags, the cutthroats, bastards, desperados, fiends, minions and hooligans. From Spider Clan Ninjas to Demons from Temen-ni-gru, this is your moment. Just because you are a bad guy, doesn’t mean you can be afford to be bad. You have to be excellent, lest the game falters.


In this article we’ll start with the basics of enemies and their design, then talk briefly about the artificial intelligence of the rascals, followed by some categories and examples of the ruffians themselves and ending with a short overview on how all these elements work together. Some bosses might also be covered, but this piece is generally about the small fries.
First off: the basics. Enemies can be big, small, fast, strong, weak, slow and everything in between. Their attributes and underlying stats tend to mirror this. It is all about how you want them to function. Give these vipers a lot of health and they risk becoming a nuisance, too little and they aren’t a threat. To build around this these knaves tend to be, broadly speaking, designed with five main characteristics in mind:

threat, mobility, health, resistance and volume

Let’s grab a generic brown ninja from Ninja Gaiden as an example. He deals minor damage, can harass the player with low damage shurikens, has a high mobility being able to jump off of walls, has low health, low resistance to stagger when hit and tends to come in pairs. These ninjas are a prime example of a good early game enemy.


Compare this to the Black Spider Ninjas from later in the game who deal more damage by combining explosive shurikens with various attack combos, retain the high mobility, have high health but still stagger when hit and come in large groups.
Several of the characteristics are upped to make them more threatening but they are still kept vulnerable by having them stagger after a single hit. If developer Team Ninja had for instance also made them immune to stagger or only take damage from one type of weapon it could’ve been a very tedious hoodlum to fight. An extreme example would be the Alchemists in Ninja Gaiden 3: Razor’s Edge. They are a big threat to Ryu with their magical ranged abilities, grabs and powerful aerial attacks, sport very good mobility being able to teleport and jump around, have almost boss-like health, do not stagger when hit and come in large numbers in the late game. While not necessarily a bad goon because of this, tedious brutes tend to lead to tedious gameplay, which leads to tedious strategies. Black Spider Ninjas are fun to fight because they leave room in their design to fight them creatively, while Alchemists almost force players to look for an optimal strategy just to get past them reliably.


All this is complemented by enemy’s actual movesets and how their fighting styles pans out. Do they attack off-screen, do their attacks have a visual or audio cue to alert you? Is the cue fast enough to react to, or do you need to know the pattern? What about punishing a recovery animation of those attacks, does it always play out the same way or can the goon randomly dodge away to keep a fight tense? One can divide enemy movesets in three simple styles: those that you combat with reactions, those that you combat with knowledge and a mixture of both. These factors can lead to either monotonous and exciting gameplay. As such:

In a sense, enemies should be designed with a tone of voice in mind from the outset.
What do you want your player to feel when facing them? Fear? Empowerment? Or something else?

A ‘tone of voice’ is a term that describes what kind of feeling something presents itself with. This can be casual, aggressive, frustrating and everything in between. So with a ruffian that has low health, medium mobility, zero armor and huge numbers you have a - at least temporary - empowering tone of voice, as seen in games like Dynasty Warriors. Likewise having a cutthroat be very dangerous but always come alone can give off a feeling almost akin to that of a rival battle. Nioh’s Revenants are nearly always fought one on one, while their other characteristics shine through the roof. This makes them rare, tense but also memorable engagements.

This tone of voice returns in the motivation and style of the villains themselves. While on the outset it seems like killing you is the goal, this is rarely really the case underneath the hood of the game. There are so many types of foes to fight, with so many differing goals that these evil-doers want to accomplish. Some exist merely to frustrate you, others to promote certain mechanics or even break those mechanics. While they all try to kill you in the end, the way they get there and the game they form is the element we’ll highlight here. As such we’ll look at a few examples:

Having gameplay-mechanics be there isn’t enough reason to use them. As discussed in our earlier article Depth versus Complexity some games have struggled with this key element for decades like God of War, while others found the solution on their first try like Devil May Cry with its style-meter. One way to motivate players to use available mechanics is by having foes be tough. In Yakuza 3 even regular thugs attack in packs, block frequently and try to grab you in hopes of draining your heat-meter. This forces players to actually use environmental items, spend meter smartly and look out for off-screen threats instead of mashing attack in what might appear to be a simplistic game. Undead enemies of a similar vein would be the Red Eye Knights and Black Knights from Demon’s- and Dark Souls respectively, both being exercises in what the player has learnt so far in terms of blocking, spacing, stamina management and dodging thanks to their durability, resistance to stagger, and high damage-output.

Building on this are the type of enemies that cause you to experiment with the gameplay mechanics. The Minotaurs and Gorgons from God of War 1 boast a large amount of hitpoints, making a direct offensive a boring event of blocking and minor punishes, lasting for minutes. But this frustration and boredom can lead to the player trying new things, thinking “what if I launch the Gorgon and air-grab her”? Suddenly a boring slog is over in less than six seconds and you feel invigorated. Difficulty can lead to experimentation, though it can be dangerous with some players not experimenting due to their own personal nature and just playing in the most boring way possible, ruining the experience. Other examples are Phantoms from Prey (2017) which are very powerful in the early game, making head-on encounters suicide. However by using the environment, like luring them into flammable liquids surrounded by turrets, you can turn the tables, introducing the player to the other mechanics of the game.

But some dastardly outlaws aren’t meant to be fought straight up. The Hysterics from The Evil Within 2 are so fast and resistant to stagger that they’ll always get a few hits in before you can kill them. As such these undead require preparation: instead of rushing in you have to think on how you will engage the fight, adding a tactical layer.

Sometimes instead of only motivating them, developers try to force players into using their mechanics. For instance, foes that are of a specific color and can only be damaged by weapons of that color, like in DmC: Devil May Cry, or the shielded Heavy Soldiers in Ninja Gaiden 3 which first need to be removed by heavy attacks.
While interesting to novice players, urging them to dive further into the game, it serves only as a restriction for veteran players by limiting options and thus creativity. This limitation on combat can take many other other forms as well, like God of War (2018)’s later Traveler enemies being immune to the game’s prevalent stun mechanic. A different way of incentivizing play variation was done in Devil May Cry 3 where each weapon had a specific element which some enemies were resistant or weak against, leading to slightly higher damage output. While not outright forcing players to switch things up it does reward experimentation.


Sometimes foes just frustrate the player. For instance: the name “Mastiff” is infamous on the Metal Gear Rising: Revengeance forums. They have a lot of health, jump around wreaking havoc on the camera, but their biggest offence are their grabs. They’re very quick to activate and force you to wiggle the analogue stick at a rapid pace to reduce the damage taken. On higher settings the amount of wiggle needed to escape can lead to very physically tiring fights. Other foes like the shield-bearing enemies in The Cursed Crusade are very hard to kill because they have little openings, turning the fights against them into long waiting games. In a nutshell a frustrating foe is hard to kill, come in groups, has long attack animations and / or instigate sections in combat where the player can’t really do anything, all leading to players pulling their hair out.

Other times they are designed to keep you on your toes. A gameplay mechanic can be such a big driving force behind the game that stopping the player in his tracks is a bigger priority than killing him. In 忍Shinobi each ninja killed has you grow stronger until the final enemy of the encounter is killed or 5 seconds pass without bloodshed.
As such their main focus is stopping your kill-chain. They have an heavy emphasis on blocking, dodging and other random moves to keep you occupied while your chain runs dry. Foes like the spider-like Tsuchigumo slow you with their webs while Tounin- and Sennin Kunoichi stun you in place with their kunai. One mistimed dodge with these foes spells doom not because they kill you but because they put you in a bad spot.

On the flip side an evil-doer’s design can also lead to an unintended playstyle. A foe that was meant to be menacing ends up being dealt with in the most boring way possible. The Tactical Engineers from Devil’s Third are slow, extremely powerful, have more hit-points than some bosses and are immune to any form of stagger: these beasts will hunt you down. But because of their danger and by often pairing with other enemies players tend to cheese them, hiding behind edges of cover or keeping them at a distance where they can’t harm you. Instead of a fun and tense engagement these fights just make the player sigh and say “not another one”. Instead you’d want a foe that is strong in such a way that they essentially require you to become a better player. Fighting the Bogeys on God Hard difficulty in Vanquish is one such encounter, where just being good at the game isn’t enough. Because of their high aggression, mobility, team work and no real way to ‘cheese’ the fight you need to know the advanced tactics, the level’s layout, the animations, sound cues, item locations, the works.

Players memorizing patterns and becoming better does have its problems however. As enemies tend to have specific attack patterns it can lead to fights being over once you know them. Once you know the pattern to Gwyn from Dark Souls it doesn’t matter what gear or setup you have, you’ll never lose to him again, the threat is gone. This is a problem that plagues Souls games and others of its ilk heavily. Some games try to avoid this by having a random element that forces players to stay on their feet and adapt. Alma from Ninja Gaiden randomly blocks and auto-dodges your attacks so that every time you attack you have to keep in mind it might miss, forcing you to think on your feet. Ninja Gaiden 2’s Incendiary Shuriken Black Spider Ninjas can be extremely straightforward to fight until their shurikens hit you. Sticking to your body they will explode in a few seconds meaning you have to look for iframes very quickly or suffer for it: their unrelenting pressure forcing you to think on the fly.

Even with all the above, there are many more variables regarding enemy challenge, for instance enemy combinations and how different foes work together. Or how a single foe doesn’t have to be hard if the stage lacks checkpoints or that an enemy’s goal doesn’t have to be killing you, but weakening you for the fight that comes after him. Or how elemental resistances and actual stats factor into it? What about RPG mechanics, level differences? Creating a bad guy is one thing, but using him correctly is a topic for another day.

One element we will touch upon shortly is enemy AI, meaning their ‘brains’ and thus how they function in combat. This can go from extremely binary, i.e. Ninja Gaiden 2’s obsession with rushing you down to Max Payne 3’s enemies trying to flank you. These also have to fit the tone of the game but it is important to note that one common tone always seems to be set: the AI tends to be really dumb. As they seem to be operating with only a single command, to kill the player, they feel very restricted. When fighting a selection of foes in Devil May Cry 1 they can lock you down and corner you, covering each other’s openings and recovery with attacks from their allies, yet they never do this. Why not?
Making a difficult AI isn’t hard as they have access to your inputs and can play in on these as seen in fighting games like Street Fighter 2’s infamous Arcade Mode where enemies punish you with frame-perfect reaction. Making good AI is about making one that fits the tone of the game and gives the player a sense of feeling that they can beat this. Technically it is the art of making the foes feel threatening while under the hood they are actually holding back, without having you realize this. Some games tie this to difficulty like Metal Gear Rising: Revengeance and God Hand where enemy aggression increases per level. In God Hand on LVL1 foes rarely attack and cannot attack off-screen while on LvLDIE all restrictions are removed and you really need to keep up the pressure lest they lock you down completely.

In a blogpost, one anonymous developer shortly talks about how he one time made the AI of a shooter he was developing too difficult. He states:

In essence, we had built an AI that really did outsmart most players and players really don’t respond well to being outsmarted, so they had turned the tables by repeatedly dragging the AI into situations where it was dumbest, and then winning through attrition. And that was a problem, because it wasn’t fun. But it sure was effective.

He/she continues by stating that:

In hindsight, I believe that the way we approached the problem was fundamentally flawed. We had erroneously chosen the role the AI took, and that had been our downfall. We made the assumption that the AI’s primary goal was to be given the same tools as the player, and to use those tools to defeat the player in an intelligent way. The problem with this approach is that it assumed the fun was inherent, when there was no real way to prove that.

What fun is, is of course a very hard topic to cover and not one we’ll touch upon here in depth. Fun is, like most things in life, dependant on the person, his needs and also his expectations. It is important in tests as noted above to not only note the fun, but also note if it is fun for the audience that a game’s tone of voice appeals to. Because in the end that’s the most important thing when fighting a foe. It has to be fun, engaging and exciting for your target audience. When the music starts and the enemy walks into sight their hands should steady and their eyesight narrow, thinking “yeah, let’s do this”! Not “ugh, not this again”. That, in a nutshell, is a good enemy.


★ postscript notes ★
----------------------------------------------------------
  • The idea for this article came from playing Yakuza 3 on EX-Hard, realizing that while mechanically it was one of the less complex, deep and interesting games in the series, it held my biggest interest since fights were hard and threatening;
  • I never got why Zangief was in Wreck it Ralph’s talkgroup for evildoers. While sure as hell looking like a bad guy he was never a villain nor a member of Shadaloo;
  • I had originally planned this article for June, but God of War(2018) took all of my attention;
  • There were, originally, a lot more examples present in this article but I decided to scrap them in favor of making it more easier to read. Some examples that were removed were some obscure enemies from Lollipop Chainsaw, a nod to Killer is Dead, Chaos Legion (sorry Birdman), some annoying foes from Devil May Cry 3, how enemies designed around Nero from Devil May Cry 4 could feel weird to fight as other characters and many other obscure action titles;
  • To all those that hunted me down to send me death threats for my take on God of War(2018), your tears sustain me and this site; 
  • Contrary to popular belief this is the first article where I use the thesaurus for alternative names for badguy. Dirtbag was the most creative one I could come up with on my own.

★ sources ★
----------------------------------------------------------

No comments :

Post a Comment

Post a Comment