Designing enemies: top-down vs. bottom-up (conceptual) approach

Here's something I'd like to discuss and hear about how other people in the RPG Maker community work, and if they feel a particular way is better or more effective than the other. When you design enemies for your project, do you decide on the enemy concepts that would best fit your dungeon purely in your imagination, and THEN choose battlers that match them as best as possible? OR...do you take stock of your library of battlers, and THEN conceptualize enemy types based on the monster artwork that you have available?

Basically, in your design process, is the artwork driven by the enemy concepts, or are the enemy concepts driven by the artwork? Have you tried both ways and found one way works better for you personally? (Note: I'm obviously talking about most amateur / single-person developers who typically lack the budget to commission 100% custom artwork for their game)

For me it's a bit of both. Usually I have an idea of what I want in my dungeon, and then go looking to see if I have anything that matches. If not, then I have to re-think my enemies. But sometimes when looking (or downloading) I see something which sparks off an idea which I then run with.

While I've used both methods, I tend to lean towards designing my world and then determining what kind of stuff the player will encounter there since I find it allows me to be a bit more creative in my designs. I enjoy the process of asking myself what kinds of monsters might be native to a given area and since I have a bad habit of developing a setting much more than I develop the characters in said setting, designing the monsters is a big factor.

Unless if the encounter is REALLY specific, I usually tend toward looking at the stock asset and create / visualize something based on it. The encounters on my game are quite trivia that it's really separated from the game world itself. In short, it's there for gameplay purpose.

Thing is, if I visualize the enemy first then look at the stock asset, I usually didn't find anything I want. Sure I can try to create that on my own, but it's unnecessary because I can just try the reverse approach. So, it's the artwork that drive me to create the enemy concept rather than vice versa.

For example

I kind of design whatever comes to mind at the time, I have no specific theme I really follow. If it pops up in my head, and i'm in the right mood, it's going to be made!

I use Kes's method. My art skills are lacking so sometimes you have to make do and sometimes the lack of something forces you to be creative.

I feel like enemies need to be designed around the battlers. Each battle is a challenge to be addressed by the player using the tools at their disposal, in this case, the actors and their skillset.

Each battle should be designed with the question in mind "how can I challenge the player's knowledge of the game's mechanics this time?"

When i design enemies I look to create mechanics first then i add names and pictures second.

Example... i want an enemy to function as a dps race and force the player into acheiving the best burst damage possible over a single turn..

So.. an attack that drains hp and restores most or all of the monsters hp faster than the player can recover (assuming the monster has a total hp pool around the maximum possible damage we expect the party to be capable of at the given range..

Seems like something a vampire would do.. so i use a vampire sprite and call the attack drink.. but looking mechanically first and artistically second gives me better control over the challenge of the game.

If you care more about the story than the challenge you might do the reverse. Its all about priorities

When designing enemies my first step is to work out their mechanics and what role they play in combat. It's important to communicate to the player what the enemy does, so the enemy should be designed to express it's function.

If an enemy is supposed to be a fast attacker that will hit hard but go down in one or two hits then it doesn't make sense to have the enemy look like a big, lumbering oaf in heavy plate armor. A fast enemy should "look" fast. Maybe tall and lean in a pose that suggests it's about to pounce. A damage-sponge enemy could be a guy in thick armor standing in a defensive position, like maybe holding a shield forward to suggest he can block the player's attacks. Likewise enemies that center around gimmicks like status effects should probably try to reflect their specialty in their design. Especially for annoying status effects like poison; designing them with purple miasma clouds or poisonous mushrooms helps convey "kill this guy fast or he'll poison you". Telegraphing tricks like that can make them less annoying (a little anyway) since the player can see it coming or will realize it was pretty obvious after it happened.

But this pretty much hinges on actually having said art assets available or the skill to make them yourself, which not everyone has. Sometimes picking the most appropriate design from a stock image folder is all you can do but the underlying principle is the same.