Many RPGs center around combat, so it's important to get it right. So I think my combat is going to blend ideas from Pathfinder 2e, D&D 5e, and Warhammer: Age of Sigmar Soulbound.
The main idea I'm borrowing from Pathfinder is the idea of actions. Every player gets three actions. Everything they can do uses those actions. Some things--movement, attacks, a few non-attack spells--take one action. Other things, like loading a crossbow, or casting most spells, can take two or three actions. Some things--for example, certain powerful spells--could take up to 6 actions, and have to be split between turns.
In addition, players get a reaction each round. The reaction works for opportunity attacks, and other things that only take a single action. This way you can ready an action--spend an action on your turn to set the conditions to trigger an action.
Pathfinder 2e uses penalties to make it difficult to make multiple attacks on your turn. You can use two actions to make two attacks, but the second attack has a -5 penalty. If you use a third action to make a third attack, you take a -10 penalty. I don't really care for this, and would probably use something like D&D 5e's Extra Attack feature. You can only make one attack during your turn unless you have an ability that lets you make an extra attack. And that includes spell attacks.
Damage by Degree
When you hit with your weapon, the damage will be a base amount (based on 0-5 for your weapon + Fortitude) times the number of degrees of success. The calibration will generally be that rolling a natural 10 will score 1 degree of success, possibly 2 against brutes--high hit points, low defense monsters. Getting an extra +1 degree of success is relatively easy (you'll get a >=15 or a Double 6 or Double 7 23% of the time), but while you could get +2 or even +3 on the roll, that will be extraordinarily rare (you'll roll a Double 8 or Double 9 only 2% of the time, and a Double 10 only 1% of the time). So most monsters will have HP balanced around how many hits they can take with 2 degrees of success from the PCs. At first level, the base damage will depend on the character, but a Warrior with a huge, two handed weapon should do about 10 base damage. So most hits will be 10-20 points of damage, and only if he's very lucky (1% chance) will he do 40 points of damage. I think at first level that 1% chance should probably take the enemy down.
At 20th level, most of the players shouldn't be getting more degrees of success fighting enemies of the equivalent level, but an optimized Warrior with a +3 magic weapon (which applies to both damage and attack bonus) may get one extra degree of success. So on average, the warrior will be doing 2-3 degrees of success. However, I expect that he will be doing considerably more base damage. Assuming magic weapon, a 20th level Warrior with maximum Fortitude and a weapon with a base value of 5, his base damage would be 20. He would also get two attacks most rounds (maybe three, but that's harder to achieve if you need to move between targets). So now he's doing 40-60 per attack, and attacking twice. I haven't taken into account how often he'll miss, but I imagine he'll be hitting most of the time.
Instead of a Warrior, let's consider the Expert. They don't do as much base damage. They do, however, have an ability that increases the degrees of success when they hit with a weapon. They'll probably tend to use weapons that allow them to add Agility to damage, which either have a long reload time (crossbow), or do low damage. So their base damage at level 1 is likely to be 6 at most, and at level 20 is unlikely to exceed 12 (even using a +3 weapon). They do, however, have an ability, Precise Strike, that, if their attack hits (scores at least 1 degree of success), they can increase the degrees of success. This means that at level 1, they should be doing 2-3 levels of success, potentially 12-18 damage. At level 20, they'll probably be doing 6-7 degrees of success on a hit. That's 72-84 damage. But they'll only get one attack per round.
Spellcasters rely on their spells, which should scale in a reasonable way, but I imagine they may apply conditions or spread damage around more often than they'll compete on single-target damage with the warriors.
Hybrid characters will get lesser versions of these features, possibly coming online at later levels and not scaling as quickly (only going as high as +3 damage or +3 degrees of success). They can be combined with each other, or with spells. For example, I intend to give the Holy Knight a 0th-level spell that allows his attacks to do 1 degree higher success against outsiders or undead, 2 degrees at 11th-level (and probably ignore 1 or 2 degrees of resistance as well).
One thing that I like is using theater of the mind for combat, something that is easier to do in some systems than others. I'd like to make my system friendly toward that, without requiring it. That brings me to the idea of zones, which is the way combat maps are handled in Soulbound. I'll refer to combat using zones as zone combat, and combat using miniatures on a grid as grid combat.
A zone is an area of combat, defined as much by the terrain as size. A smallish room (up to 30'x30') is a zone, so is the hallway, and also the other small room, while the large audience chamber might be two or three zones. A bridge over a river might have three zones, one side of the river, the bridge itself, and the other side of the river. The river itself could be another zone.
Zones affect range and movement and area of effect, but it's not hard to keep it pretty simple. It takes one action to move from one zone to another. It takes another action to move close enough to engage someone in the same zone as you in combat. Engagement is another word for being close enough to an enemy to attack them with a melee weapon, or for them to attack you. If you're using miniatures, you can just see which miniatures are adjacent, and not need to keep track of engagement specifically.
In zone combat, engagement is an important method to keep track of who you can attack with a melee weapon. Either combatant can initiate engagement, and if multiple enemies engage with you, you can be engaged with multiple enemies. If, however, you are engaged with one enemy and you want to engage with another one, you'll lose engagement with the first enemy. If you are engaged, you take an opportunity attack when moving to another zone or moving to engage with an enemy unless you use a step action to disengage. Engagement is not transitive--if you engage an enemy, and an ally engages them too, and another enemy engages your ally, you aren't engaged in combat with the new enemy. Your ally is engaged with both, but you'll need to engage with the new enemy to attack them, and doing so means you'll no longer be engaged with your current enemy--you'll either need to disengage or take an opportunity attack.
If you have a weapon with reach you can attack someone in the same zone without engaging. All other melee weapons require you to be engaged. With a ranged weapon with short range, you can attack anyone in the same zone or an adjacent zone, with medium range, you can attack anyone two zones over, and with long range, you can attack anyone on the battlefield assuming you have line of sight.
Usually, area of effect spells affect an entire zone. Some large area of effect spells could affect two adjacent zones or the entire battlefield, though area of effect could come into play.
The air generally counts as its own zone, so someone with flight can easily attack any zone (assuming the zone doesn't have a roof), but he can also be attacked by anyone with a short ranged weapon.
Now I do want rules that also work for miniatures and squares, so I'd want to present both. I don't think you need the engagement rule if you are using miniatures--you engage by moving adjacent to them, and you disengage by taking a step (moving one square with the miniature), if you don't want to take an opportunity attack. The engage action is really just a move action within your zone--you move to get close enough to attack an enemy who's in the same room.
I may also need an action to allow someone to block the way between two zones, assuming the GM agrees the passage is narrow enough. I think in that case, an enemy can't get past him unless they make a roll (probably a Fortitude contest of some kind), but the defender can't take the engage action. Enemies in both zones can engage with him, though. I probably need to make this an action that would work for both zone combat and miniature combat, so I'll need to think of how the rules would apply.
I think I want two levels of difficult terrain, difficult and very difficult (maybe think up a better name later). Difficult terrain halves your movement, and very difficult terrain cuts it to a third. (Remember the Elf feature that ignored a level of difficult terrain? This is where it applies.) Using zones with difficult terrain, I can say that if a zone is difficult terrain, it takes two actions to move, either to engage with someone or to move to a new zone, or three if it is very difficult terrain. Moving from one zone to another takes a number of actions equivalent to the highest difficulty of terrain of the two. If one zone is normal and the other is very difficult, it takes three actions to move to the very difficult terrain or away from it.
I haven't decided yet whether casting a spell grants an opportunity attack. I'm leaning towards no. However, I do want to allow an enemy to interrupt a spell. If they ready an action to attack you when you cast a spell, they can force you to make a concentration check to keep the spell. By the same token, if you cast a spell that takes more than one round, then if you are attacked between starting the spell and finishing it, you also need to make a concentration check or lose the spell.
Making a ranged attack may grant an opportunity attack, however, including a ranged spell attack. Perhaps both would require a concentration check to still attack, otherwise you lose the spell or the ammunition as the magic or arrow goes wide.
Finally, I'd like to borrow from Pathfinder 2e's way of sustaining spells. You don't cast one spell and concentrate on it. Instead, you need to use one of your actions to sustain a spell. You can sustain more than one spell, but you quickly run out of actions that way. Pathfinder 2e doesn't let you end concentration on a spell by just doing damage to the caster--there are no concentration checks for that. Instead, you stop someone's sustained spells by causing a condition that prevents them from using an action to sustain the spell. I do like having to concentrate to maintain a sustained spell when you're hurt, so I may keep that aspect of concentration. Your degrees of failure determine how many sustained spells you lose if you're sustaining more than one spell. I may let the player decide which ones they lose.
Here's another place I borrow from Soulbound. You can attack with two weapons with one action, either attacking the same enemy or two different ones you're engaged with. However, in order to do so, both weapons need to have Agility for both attack and damage (currently, there are only two: daggers and rapiers, but I could see adding unarmed attacks for martial artists, perhaps some paired weapons purely for dual wielding, maybe even bucklers will be treated as an offhand weapon). However, for both attack and damage, you divide your Agility in two and apply half to each, rounding up for your main hand (whichever is holding the larger weapon), and rounding down for your off hand. This is similar to what you do in Soulbound, where you divide the number of dice into dice for one weapon and dice for another. I may add talents for Two Weapon fighting. If you have an Agility of 4 or more, you can take a Talent that will let you add +1 to each weapon's attack and damage, and if you have an Agility of 6 or more, a Talent with the first one as the prerequisite will let you add +2 to each weapon's attack and damage. I may also add +1 to Defense when you are wielding two weapons but only attack with one, but I haven't decided whether that's automatic, or part of these or a different Talent.
Another thing I'm mulling over is resistances. Some creatures are just hard to hurt with certain elements. For example, casting a fire spell at a fire elemental should do a lot less damage than against a plant creature. However, I don't like the idea of complete immunity. There should be at least some way to harm enemies with fire spells, even if they're resistant.
My initial idea is to have resistance that builds off the idea of degrees. When you hit a fire resistant monster with fire and achieve any degrees of success, the number of degrees of success is reduced by one. More resistant enemies can have more degrees of resistance. A fire-breathing dragon has two, a fire elemental has three. However, when you have an enemy with three degrees of resistance, that can turn a degree of success into two degrees of failure. If I have fumble rules (something that I'm still considering), then I'm not sure that's fair. However, if I start setting limits: a resistant creature can't reduce the success of the roll below a near miss, for example, then I run into an order of operations problem.
Let's say I introduce a talent that says when you score a success against an enemy with fire, you deal an extra degree of success, or two degrees of success if they're resistant.
Let's get back to my fire elemental with three degrees of resistance. I score two degrees of success, they apply their resistance to turn that into a near miss, and then I apply my talent--only I can't, because the talent says I can't apply it unless I have a success first. Okay, let's say I solve that issue, and I say that I can apply the resistance first, then the talent--then it goes to two degrees of success again. But if I go the other way, and add my two degrees of success first, bringing it to four, then deducting three degrees of success, I only have one degree of success. By applying the resistance first, I actually did more damage than I would have if I applied the talent first, because I ran into the limit.
I think the easiest way to do this is to not set limits to resistance, and allow it to turn a hit into degrees of failure. It's really no different than attacking someone with a very high defense. Failing to do damage is more a matter of the heavy armor than a fumbled roll. You can just open yourself up if your attacks, no matter how skillful, are ineffective.
This does still leave the problem that I still have some degree effects with conditions. I don't want to change the Precise Strike rules, where I get more degrees of success when I hit, since I don't want to make my roguish characters more likely to hit than anyone else. I think, instead, I'll apply the conditional degree effects only if the condition still holds after the other degree effects. It seems like I can, however, make resistance unconditional, and adjust the elemental damage to something like "Your fire attacks ignore two degrees of fire resistance. In addition, any successful fire attack gains one degree of success."
I said this earlier, but I want PCs to recover hit points after every battle. I haven't decided yet whether I want this to be a short rest thing, in that they need to rest for an hour after a battle and recover all their hit points. If so, I may also tie it to recovery of spell points. Alternatively, I may just make it so that if you have five minutes to catch your breath after a battle, you recover all your hit points. (In this case, I'd keep the spell recharge rate by hour and not require rest to do so.) The key to understanding this is that hit points are not health. They measure your endurance, your ability to roll with the punches, and otherwise avoid serious injury. Only when your hit points go down to 0 are you hurt enough to take you out of the fight. And any significant injury is sufficient to take most people out of a fight.
When you recover from being taken down to 0, you return with an penalty. Some games call these Wounds. They can be tied to levels of exhaustion in D&D (especially the One D&D version which has ten levels instead of five). In earlier posts, I called it Fatigue. But I think I'll borrow from other games and call this Stress. Stress can be physical--injury, bruises, sore muscles--or mental--being worn out, discouraged, depressed. The total amount of Stress you can endure is Fortitude (minimum 1) + Will (minimum 1), for a minimum of 2. Most players will want more than this. For each level of stress, you get a -1 to all rolls. Once you exceed your maximum, you are broken. This doesn't necessarily mean you are dead, but it does mean your character is no longer an active participant in the adventure, and cannot make any rolls. You recover one point of Stress per full night of rest, or two points if you rest in a comfortable and safe location (in town, or your team base, or some other safe area).
Going down to 0 is one way you can gain Stress, but not the only way. Traveling in hostile terrain (arctic or desert conditions, for example), 2 degrees of failure on some checks, the price to turn an important Near Miss into a success.
So, if you regain hit points each time you finish combat, that means you need less HP, right? Well, not so fast. In most groups, players are going to be close to full health before most battles. Also, damage can stack up quick if the enemy rolls a dreaded Double 10. So you don't want to make the HP too much lower. Here's what I'm thinking--you start with 10 + Will (minimum 1) + Fortitude (minimum 1), and add a fixed amount per level in each class. This amount can range from 2 through 5, depending on the class. Initially I'm thinking 5 for Warriors, 4 for Skirmishers and Champions, 3 for Experts and Adepts, and 2 for Magi. The HP will start higher, but scale slower. Pretty much everyone will have a decent Fortitude or Will, though both may be rare.
So if you don't die when you go down to 0 hit points, and you don't die from hitting your maximum Stress, can you die in this game? Well, two facts come to mind: One, death usually sucks. Two, fear of death is a great motivating factor.
In general, going down to 0 does not mean you die. But an enemy can use three actions to coup de grace a downed character. If the entire party goes down, then it's understood that the surviving enemies could very well kill them. Not being able to fight back as they put you out of your misery will definitely kill a character. A coup de grace will definitely be part of the rules. It will take three actions, so most enemies won't do it while in the middle of combat, but a GM can use it to put some fear into his players.
The GM can also declare that some actions are so risky that failure may mean death, such as trying to jump over a gorge. He should inform the players before they attempt to do so, though.
Another way to die is to go down in a Blaze of Glory. This is a rule I'm borrowing from Soulbound. If you're at or are knocked down to 0 hp, you can declare that you're going down in a Blaze of Glory. You immediately stand up with 1 hp and you take your turn, your spell points are fully recharged, you have three actions, you are immune to all damage and all conditions, all rolls you make are Double 10s and ignore resistance, and all rolls to save against your rampage subtract two degrees of success. At the end of your turn, you immediately die and cannot be brought back (I'm not sure whether there will be resurrection in the game at all). Then play returns to where it left off in the turn order before you took your turn.
Can this be abused? Definitely. It's up to the GM to make sure it's not. For example, the first time it happens, let the player make a new character at the same level. If he does it again, the next character can be a level lower, and then another level lower. Even going out in a Blaze of Glory doesn't make a 1st level PC stand out among a group of 7th level PCs.
At this point, I think I'm done with this series of posts. I've described the system and how it works, thought through some numbers, and done about as much as I can do without playtesting. In some ways, coming up with the basic rules are easy. The hard part is in the details: working out the features and progression of each class, coming up with all the spells and spell schools, creating monsters, writing it all out in a book, and playtesting it all while tweaking the numbers. Overall, it's a lot of work, and I'm not likely to do it unless I have a group that's interested in playing this game. We'll see.