:!WARNING!: this is a long post where I ramble cause I like talking about my projects. Good luck.
Combat is something of a sticking point in tabletop design for me because if the system is going to have combat be a significant part of the gameplay experience then it NEEDS to make sure the combat system is pulling its weight but at the same time not taking up more space than it needs to. It's a fine balance and that balance changes based on what the game is trying to achieve which throws yet another complication into the machinery because the ideal system for a highly tactical combat simulator like PF2 is going to look completely different than the ideal system for a comfy storytelling game like ryuutama. When I was designing to battle mechanics for Celestial Solstice I already had a model to base the design on since I was attempting to directly port the battle system from ogre battle 64 into paperspace, but Delvers has proven to be a different beast because it's systems are untethered to any existing property that might grant some guidance concerning what the expected end result should look and play like.
SO.
Here's some thoughts I've been mulling over concerning the design and how I want to use the design to accomplish some specific goals.
The game master doesn't roll for the things that NPCs do. The game master already knows how they want NPCs to interact with the world and whether or not they can do those things, so rolling is pointless UNLESS what the NPC is doing is opposed by a PC, in which case the player is the one doing the rolling, not the GM. For example, if an enemy attacks a player, the player makes a roll to defend themselves rather than the gm rolling for the enemy to attack. As another example, if a player casts a spell that needs to overcome the mental defenses of their target, the player will test their mental might statistic instead of the NPC rolling to resist the effect.
Having GMed for a long time using a lot of systems one of the few things I think is a constant is that it doesn't hurt to make things easier on the GM. The other is that players like self-determination and they LOVE to roll dice. These things kinda fit together nicely because I've had a lot of success in the past with off-loading dice rolling to players and I want to try to translate that into delvers. The idea is that, if dice are getting rolled, that players are most likely the ones doing it. Rolling dice is fun, but as a player you're rolling dice for one character, maybe two, MAYBE some henchmen, but as the GM you're rolling dice for EVERY SINCE MOTHER FUCKING that isn't a player and that can start to become a major distraction. If the GM is going to be rolling things, it should be for how the world is behaving, not whether a guard notices a goblin creeping just beyond the torchlight. When I say "how the world behaves" I mean for things like weather, and random encounters or, more specific, how the world behaves in response to the PCs actions, like NPC reactions to players or morale during a fight. These are significant rolls that have a major impact on the gameplay and can define a moment: they're rolls that a worth making.
This is all a very long-winded way for me to segue into my next point, which is that this will SIGNIFICANTLY simplify what constitutes an NPC. So what exactly will creating an enemy look like?
Most importantly, enemies won't have any ability scores or saving throws or anything like that since players are rolling all that stuff. This is the most important part to be because this is what enables an essentially infinite amount of variety in enemy design with VERY little effort.
The enemy will have a RM, a roll modifier, which is a % that is either added to or subtracted from rolls that target the enemy. Rolls that players make to do things in delvers are done with a d% in an attempt to roll under their statistic, and that roll is what gets modified so if they player rolls to hit an enemy with a +15% roll modifier, the player's roll result is increased by 15 making it harder to succeed. The modifier can also be negative if the enemy is supposed to be particularly easy, or it can have no modifier if it's an even challenge for the PCs.
Determine how many hits the enemy can take. You say "I want this guy to be able to withstand 10 hits" and do that for each enemy and then add together all the hits in the encounter into one pool. As hits get removed from the pool, enemies will have to make moral checks which they may fail and flee or surrender, or they'll be removed from the board as hit thresholds are reached. The standard rate for player damage is 1 hit for small weapons, 2 for medium, and 3 for large. Simple.
If you don't need anything complicated then, well, that's it. If you need five tough orcs you can say "here are five RM10 combatants with a 25 hit pool" and you've got everything you need for the encounter. The same encounter with five easy goblins might be RM-10 and 10 hits instead. Super simple, but you can add some seasoning if that's feeling a little bland.
You can choose to set the enemy apart by giving it a role. This could be a striker, supporter, leader, whatever. It'll give that enemy some static bonus that makes it a little better in some way than another version of that enemy. Using those five orcs, one might be a striker (against which players have penalties to defense rolls), two defenders (against which players have penalties to attack rolls), one a supporter (has access to some simple magic effects), and one is a leader (increases the RM of enemies in the encounter).
If you want to keep going, you can start adding special features. These can be anything, but some examples will be included such as actions that inflict statuses, or that change damage types, or give special movement types.
The final result is that enemies can be as simple or as complex as the GM wants with the goal being that the gameplay experience is smoothed because there is no longer any need to look up stats for enemies.
"What about the battlefield?" You might ask, obliging as you are. Well, what I'm NOT doing is making a tactical strategy game. As much fun as that can be, that kind of combat eats up a LOT of design space and I don't want combat to be a central focus of the game, I just want it to be another part of it. Instead I'm experimenting with something a bit more elementary...SHAPES!
I love prototyping physreps, and these shapes represent engagement zones. An engagement zone is any area in which creatures can reasonably engage in battle with each other, such as a room, or a hallway. Larger spaces, such as a dining hall or outdoor area, or engagement areas that include multiple spaces, like a hallway that leads to a room, will use multiple engagement zones to represent those areas. He's a little set-up I threw together to demonstrate.
1.) #1 is unengaged. There is an enemy in their zone, but that enemy is already engaged so it cannot engage them. A character cannot attack in melee without engaging with their target first. Each turn a character can use one action, of which attacking is an example.
2.) #2 shows two foes engaged with one another. Unless otherwise stated a character cannot engage multiple characters at once so unless one of them breaks off neither can engage another enemy.
3.) #3 shows two enemies engaged with a single enemy. While a single character cannot engage multiple characters, multiple characters can engage a single character. For each additional enemy engaging a character beyond the first, all rolls against that character get a % bonus.
4.) #4 is unengaged but in danger from...
5.) #5, who wields a ranged weapon. Ranged weapons can target foes in another zone but suffer a % penalty when firing across multiple zones or at an engaged target.
6.) #6 is not in the same zone as an enemy and doesn't have a ranged weapon, so they can't interact with the combat unless they move. Each turn a character can use their move to move to an adjacent zone or to engage an enemy in the zone they currently occupy.
There are no spaces within a zone so positioning within a zone is largely unimportant. My hope is that players will be less obsessive about specific positioning like this and can focus on moving the game along instead of agonizing over the huge variety of options they have available to them.
This is all still very much in the theory crafting stage and just represents my first crack at attempting to assemble my initial thoughts into a cohesive system. Like I said at the outset of this post, this is largely just an excuse for me to get to talk about one of my projects. My players don't care about the design process and trying to bounce ideas off them is like throwing water into a lake, I just get nothing back lol.
Anyway, thanks for reading this post that WILL remain UNPROOFREAD cause its LATE and I'm LAZY. Questions and comments are more than welcome, but until then happy gaming! - Forge