Skip to content

Action Roguelite Enemy Design: Aggretsuko

You are here:
Estimated reading time: 8 min

Case Study

This article transposes a Slack conversation that illuminates action roguelite enemy design principles. It is lengthy, but contains potent insights into the genre, player expectations, game feel, and fun.

The conversation has been preserved (for the most part) in its original form, to provide the necessary context for all remarks.

Enemy Concept Proposals

  1. IT Tech
    1. Attack 1
      1. creates an error/cancel/no wifi icon on the ground, with a radius around it
      2. suppresses the player’s abilities when they are standing on it
    2. Attack 2
      1. Firewall/Pasword Protected
      2. creates a wall of force that the player cannot move through
      3. The idea is that he would be a debuff enemy that appears with other mobs
  2. Corded Telephone
    1. Swings the handset around like a flail in two cycles
    2. For 4 seconds, in swings the handset in a circle with a short radius near it
    3. Then for 2 seconds it swings the handset in a wider radius
    4. Could also have the handset and cord destroy the player’s projectiles, to make it more deadly
  3. Mail Room Cart
    1. Tanky and slow
    2. Launches brown cardboard package bombs/grenades
    3. When it takes damage, it sprays out letters/papers. These could be for aesthetics or they could also deal damage
    4. Could also leave a damaging trail of letters/papers/garbage behind it
  4. Packing Tape Gun
    1. Leaves streaks on the ground that makes you slower or decreases your dodge range when you come in contact with it


Thanks for the ideas cats, some of these are great ideas and some of them I have strong hesitations about. I’ll explain everything with my reasoning and a way of thinking about player debuffs for figuring out alternatives. Seeing as we have a lot of enemies to design, I’ll prefer to over-explain my thoughts here so don’t be worried by the huge amount of text.

IT Tech:

Firewall is a great idea. I like this way of restricting the player in an indirect way. We want to a have a limit in place for how many of them there can be, but that should be fine to implement.

Re: ability suppression, this is a creative idea, but in an ARPG suppressing the player’s abilities will frustrate the player and in the least-worst scenario, contribute to a feel problem with the game. The simple explanation of the main issue is that the expected average player’s experience with it is:

Imagine in the heat of battle as they are strafing and dodging around rapidly, they’ll run across it either not noticing it at all (due to craziness going in) or thinking it’s another attack telegraph (like other ground highlights in the game), and then get confused/frustrated for a moment as to why they can’t attack. They’ll immediately move out of it because you’re constantly moving in the game anyway, and assume that what happened was maybe a bug or controller issue. Details like symbolic communication and the player making the connection in the middle of combat as to why they can’t attack is definitely more than we can expect from most players, especially if we’re aiming for a first-roguelike design.

Because this affects the player’s perceived responsiveness of the game (e.g. an input did not result in an expected outcome when it did 100% of the time before), this is also contributes to a feel issue. There is broader-scoped discussion related to how things like this add up and affect player’s perception of feel, which I’ll explain later in this message.

Corded Telephone:

I love the idea of the attack destroying player projectiles. I love this sort of mechanic as a way to impede the player’s power which doesn’t affect feel.

Packing Tape: 

Regarding it slowing the player, slowing effects on players are not a good fit for the game/genre unless in very specific circumstances, because it damages the overall average player feel and is another frustrating mechanic when already under the pressure of combat. While it functions as another challenge to present to the player, we should use and find alternative mechanics to challenge the player which don’t contribute to hurting the game feel.

I want to clarify that the two problematic cases on their own aren’t a massive problem – they would be frustrating, but not game-killing. The deeper concern is the trajectory of adding abilities and effects which under some circumstances cause a mismatch between player input and expected output (e.g. slower movement, less-far reaching dodge, can’t use attacks, etc).

As we add more enemies, elite effects, player items (which certain elite enemies can use), etc. the greater the % of the time the player in-game is experiencing being slowed, disabled, or otherwise CC’d. This affects the overall perception of game feel, especially as people remember the negative moments where they were slowed or lost control much more heavily than the positive moments.

They will remember the time their dodge went less far than normal so they failed to get out of a telegraph and died, or the time they were about to kill an elite telegraphing an attack at them, but their weapon stopped working as they were strafing around, so the enemy lived long enough to kill the player with the attack. These are the types of deaths that players on average would label as BS whether they knew why it happened or not, and from my gut & anecdotal experiences, most new players won’t understand why the input -> output mismatch occurred, and this is a “quit” moment for them.

Additional Design Discussion

I’ll share some thoughts and open-ended discussion here regarding how some debuffs or penalties on the player can hurt feel and cause frustration. These problem cases are specific to very fast-paced realtime combat. The faster and more frantic the combat, the more exacerbated the problems they cause.

Examples of things which hurt player feel significantly, and are extremely frustrating:

  • Things that prevent player movement, such as
    • Stuns
    • Slows that are strong enough they are almost practically movement-halting
    • Root effects
  • Things that prevent player actions from being usable when the player expects to be able to use them. Example forms of ability disabling directly or indirectly
    • Silence
    • Cooldown duration increase (e.g. player expected it to be ready after 3 seconds, but they pressed it and nothing happened because it’s now 12 seconds from a debuff)
    • Stuns as well
    • Removing player MP/Mana/other resource, such that abilities become unusable during the realtime combat, if this is a game where the player is otherwise not required to watch the resource.

Here are some anecdotes for games I know personally well and seen plenty of community feedback around:

  • General player sentiment in ARPGs like PoE and Diablo is that CCs (crowd control effects) which affect the player’s ability to control their character are frustrating and annoying. I’ve heard it countless times in podcasts, player feedback threads to devs, endless posts on reddit, and so on. In PoE there’s a variety of these effects and people will often only build characters that are immune to all of them, and there’s intense community outrage whenever they add a new one to the game.
  • In RoR2, a similarly fast-paced dodging-oriented game, enemies can randomly obtain & use player items with a huge variety of effects, but the developers specifically blacklisted the enemies from obtaining the few items which have effects in these categories, such as Stuns.
    • To demonstrate how punishing these effects are, in the final content update to the game, they added a rare & late-game enemy whose major feature is if it hits you, it stops you from moving for 2 seconds. This thing is a very deadly run killer despite the seemingly innocent effect. I just want to emphasize that they were deliberately trying to make a severely dangerous enemy you’d rarely encounter, and the resulting design was one which stops you from moving for 2 seconds.
  • The faster and more aggressive the gameplay, the more that enemies which apply slows, stuns, ability disablers, etc. end up being the most deadly and hated enemies and effects in the game, because they often result or contribute to getting dogpiled by other enemies, which can happen in as little as under < 1 second in pretty much all of these games, but sometimes contribute to losing a battle of attrition amidst the chaos.
    • Anecdotal example: in PoE, all caster characters must get “Silence” immunity or accept dying almost every time you encounter a silencing enemy. The scenario would often occur in which you’d walk into a room, get silenced from an enemy near the door as you enter, and end up being barraged to death by all of the enemies attacking you unimpeded, as your “go in guns blazing” strategy hinged on your attacks working.

People just.. hate CCs in these games, myself included. Some people don’t mind a little bit, but my general feel from being in these communities and playing them is that forms of CC in this genre are generally regarded as very frustrating at best, and game-ruining at worst

Regarding alternatives for challenging/weakening the player without as much frustration or hurting game feel, we have a lot of options thankfully. Some of the designs mentioned like destroying player projectiles or adding obstructions to the environment are great alternatives. Those can both be as deadly as CCs, but it’s not slowing the player, obstructing input->output expectation from the player, etc. which would evoke frustration

Examples of penalizing the player that I believe are not as frustrating:

  • Forms of reducing player damage output numerically
    • Enemies blocking/parrying attacks
    • Damage reduction buff on enemies (e.g. firewall, antivirus themes would work well here)
    • Damage dealt debuff on player
  • Forms of reducing player damage output mechanically
    • Reducing player item proc-rate
    • Conditional enemy avoidance/immunities to effects applied by the player
  • Forms of increasing damage received by the player (be careful that this doesn’t result in one-shots against any enemy, people really don’t like those)
  • Messing with the player projectiles:
    • Destroying them
    • Altering their flight path to go off-course
  • Add environment obstructions.
    • Great if these can be destroyed via enough attacks
  • Add environment hazards
    • Damaging environment hazards also make navigation/position choices more complex in battle, but don’t block movement
  • Reducing player max HP
    • This only functions as damage if at (or close enough to) your max HP, and adds the additional challenge of being a recovery cap.

My broad opinion on the best recipe for a player’s first-roguelike is to rely less on numeric & mechanical stuff, and instead make the enemy as fun as possible at their face-value through things like interesting movement & attack behaviors, bullet hell patterns to dodge and so on. Players love dodging things and learning what enemies do. Enemy movements and attacks are self-evident and don’t need to be communicated and parsed through effect symbol popups, icons/descriptions in the UI, etc.

Because the game can be made plenty challenging and engaging with just the fundamentals in place, and we avoid a lot of clarity & frustration problems that way, more of the design focus should be placed on their physical behaviors, rather than the more nuanced mechanical interactions similar to what a player would expect in an RPG. (edited)

The problem we’re presented with is that it’s certainly a lot harder to qualify and describe on paper what an interesting behavior for an enemy is, such as the way they move, than a mechanical interaction on what X effect does. And how an enemy behavior plays out in someone’s head can feel different in reality, in the context of gameplay. This emphasizes the importance of designing them less on paper and more in-engine through prototyping. The prototypes get evaluated on the merit of things like “is this enemy fun to fight when they’re alone?” “are they fun to fight in a group of these other enemies in the area?”, depending on what the goal of the enemy design is

I heard that with [REDACTED], the design direction with enemies was in the general way of “each enemy is simple (does one or two clearly understandable things), but still manages to be interesting and holds meaning in the moment-to-moment combat”, and that this can be pulled off without mechanics other than the very physical behavior of the enemy. That’s exactly the same sort of idea with Aggretsuko too

Almost forgot to add – an important detail to specify in enemy designs, given the above is to try to describe their gait.

Examples from the current game:

  • Mop: Slowly lurches forward with a pause between each lurch.
  • Binder bat: Flaps around at an even pacing, but when they get in range of the player, perform a long swoop towards the player.

And to clarify, this isn’t just about the animation visuals, but the actual physical movement the animation provides. Because we use “root motion” for enemy movement, this means their physical displacement is tied to the animation.

This is what allows the enemies to have the inconsistent hobbling, lurching, slinking, etc. that they do. But the sky’s the limit as to what’s possible with this. The ideal for me would be if each enemy had a distinctive gait, and it’s even better if it’s a meaningful part of their combat identity somehow. Differing gaits improve the baseline intrigue and personality each enemy contributes to the game.

Was this article helpful?
Dislike 0
Views: 10
Back To Top