Warning: this is a very technical design post hashing out an experimental mechanic in great depth. If you're the kind of person who enjoys seeing the full thought process behind mechanics, this is for you. I'd especially appreciate input from folks like that. Everyone else... this might be dry, even for me.
A brief overview of some rules
This entire post is about exploring an idea I have for a specific mechanic to be used for a game engine I'm currently tinkering with. You don't actually have to care about that, though. It's an unfinished mess, anyway. I have tried to articulate most of my rationale in a system-agnostic way, but it may be helpful to talk about some of the traits of this system as it currently exists.
- All d6s. No dice pools, no 2d6 bell curves, no gradations of success, no noting margins of success or failure. The core mechanic is "GM says a difficulty from 2-6. Player rolls 1d6. If they meet or beat the difficulty number, they succeed." This one you can largely ignore, I think. The math can be converted to whatever dice system you use.
- All rolls are player-facing. If you’re looking for a thief, you roll to notice. If a guard is looking out for you, you roll to sneak. This means that there is built-in asymmetry between PCs and NPCs. NPCs do not follow the same rules.
- There are no character stats. However, character creation does involve picking traits. You can ask for advantage on any roll that you can convince the GM involves the talents you picked ("sneaking and hiding," "running and jumping," "thinking and calculating," "lying and distracting," etc.), but you'll also apply disadvantage on rolls where your flaws are relevant ("clumsy," "rude," "cowardly," etc.). This one is a bit tricky to reconcile with other systems but the key takeaway is "you don't get an attack bonus from your stats."
- The baseline lethality level is high. This doesn't always mean literal death. In some genres, characters can only be "defeated" or "taken out," which is equivalent to death for combat purposes. But either way, the amount of health an average PC has and the amount of damage they stand to suffer in a single average attack is close enough that it's really risky to ever get hit.
Like I said, this mostly doesn't matter. If you want to instead imagine a basic D&D norm for everything that follows, the general ideas should make sense.
Fleshing this out
There are many variables and situations which need to be resolved. D&D combat gets complicated fast. Having good answers to these questions is the key to ironing out this mechanic into something usable.
As a general rule of thumb, I don't want to ever allow you to have a baseline modifier to a clash roll. CN is determined by the NPC you are fighting (it's, like, their only stat), and for you to adjust it must come from stunts and clever thinking and using the environment. The core decision loop driving combat is a cycle of trying to think tactically to gain an advantage and then capitalizing on it by making a safer-than-baseline clash roll. If a PC could just be talented in "fighting" or even "swords" or "guns" as a core trait they pick at character creation, then that would imply the player can always have advantage on clash rolls. Multiple sources of advantage never stack, so you could probably get away with freely making clash rolls without fear. And that takes away the fun part of the challenge. Therefore, whatever the list of talents offered in any given game using this system, those talents can never include a straightforward "good at fighting" option. That's the one source of advantage you should have to always work for.
But that creates at least one complication. Someone is going to want their thing to be "good at fighting." The "fighting-man" archetype is essential to fantasy adventure gaming. What do we offer to the player who wants to be the designated tough guy of the party? D&D's traditional method was giving the fighting-man class a scaling bonus to attack rolls. But again, we don't want baseline modifiers to clash rolls. So what do we do? Well, there's actually tons of ways in which you can mechanically give someone a leg up in combat. All we have to do is offer an option that would make a PC the most effective at violence. Something that makes everyone in the party agree "this is the guy we want to roll the clash test."
I think the answer lies in damage, and therefore, weaponry. The squishy wizard doesn't have a weapon that does a lot of damage, the rogue has a weapon that does pretty good damage, and the warrior has a weapon that does lots of damage. And when the stakes of a single risky clash roll is so high, that's why the party agrees to send the warrior up to be the one who rolls the clash test. Not because they have a higher chance of success (remember, CN is set by the enemy and is the same for everyone), but because they're the most likely to finish the job in the fewest number of attacks. And of course, I know all of you love equipment-based advancement.
While we're talking about weapons, we can talk about armor. I think this one is fairly easy. In D&D, armor affects the chance of getting hit. That would interfere with the CN thing. So we go with the most common alternative: armor is damage reduction. This also reinforces why the party would want to send the warrior up to actually attempt the clash roll. Because even though they have the same chance of failure as everyone else, they are the most resilient to the consequences because of their armor.
The next big snag to think about is difficulty and balance. I mentioned that I see the baseline chance of a PC succeeding a clash roll being <50% most of the time. For my little 1d6 system, that means that most enemies would have a CN of 4-6. This is because I like when combat is difficult and players have to tip the scales in their favor. Monsters are scary, adventurers are frail, and the most satisfying combat is where the underdog won because of clever thinking. Bosses are fun because everyone's working together to beat an overwhelmingly powerful foe.
Well, my first thought is no, that's perfectly fine. After all, the party's clever thinking for overcoming a difficult enemy should start well before the fight itself. If a monster has a high CN, the PCs should avoid them until they've come up with a good plan of attack.
But then, if you were fighting enemies with a medium or low CN then wouldn't you skip all the stunts and just go right for direct clashing? Well... yeah. And I think that's actually also perfectly fine sometimes. Not all combats need to contain the same core loop. Teaming up to take on Thanos is a different kind of challenge from gleefully hack-n-slashing your way through goblins. But the GM just needs to remember that when the odds are flipped, so too are the strategies. Monsters with a low CN should probably be doing lots of stunts against the PCs, turning the tables and using clever thinking to harass and outwit a stronger foe.
My second thought is that, if playtesting shows the difficulty is still too high, then maybe only PCs benefit from armor (damage reduction). It would be one mechanic that's biased in their favor, unlike all of the rest of this. And besides, I like having one less thing to worry about as the GM. These are the benefits of asymmetrical systems. If an NPC or monster is meant to have armor or a tough hide or whatever, that can just be accounted for with a high CN.
Up to now we've been assuming a traditional HP + damage system, where the amount of harm you suffer on a failed clash test is both variable and granular. When you win the test, roll damage. When you fail the test, GM rolls damage. But if, say, you're clashing against a basilisk, then do you just substitute the damage roll with instant petrification? Jeez, at least with dice-based damage, you have a chance of surviving a hit.
I usually feel better about these kinds of attacks when they're split up into steps. As in, Step 1) frog monster grapples PC with their tongue, Step 2) PC is swallowed whole. Or like, Step 1) PC begins stiffening from petrification, movement is halved, Step 2) PC has disadvantage on all rolls, Step 3) PC is fully petrified. Step 1) the mind flayer wraps your face in tentacles, Step 2) they dig their lamprey mouth in, Step 3) they schlrrrrrp your brain out. It takes multiple failed clash rolls for a PC to actually suffer the consequences of the attack, or at least the full consequences. And because armor is damage reduction, it wouldn't help at all. I think that makes a lot of sense.
I also think this could be an awesome foundation for combat magic. Some old schoolers dislike "damage spells" because they lack the creativity that magic should inspire. Eldritch blast can never make for as cool a story as, say, luring an enemy over their own trap door and then casting knock to make it open. But magic that is intended almost purely for combat purposes could be made more appealing if they're always broken up into multi-stage effects. Turning someone into a frog is an enchantment that takes 3 successful clashes to complete. After 1 clash, they're a little froggy. After 2 clashes, they have frog legs and they croak. After 3 clashes, they are fully a frog. Then again, there are tons of classic spells that would be difficult to translate into this format. Lightning bolts may need to be converted to a more Emperor Palpatine-like sustained lightning effect, and fireballs replaced with causing targets to spontaneously combust in escalating levels.
This made me consider perhaps if I should adopt static damage. I think sometimes people forget that you don't have to determine damage with dice. Especially since I only want to use one die size in the whole game, it might be easiest to differentiate weapons just by simple numbers. Daggers do 1 damage, longswords do 2, greatswords do 3, that sort of thing. I haven't done a lot of research into the design of games that use static damage and what considerations they have to make, but I imagine it's a big decision. Still, what I find appealing is that all attack types could then be basically described as "complete X steps to defeat enemy" nice and concretely. At the very least, adopting static damage for all NPCs and monsters would be yet another effort-saving measure for the GM.
Next we have to address, in my opinion, kind of the elephant in the room. What about attacks that clearly are not mutual clashes? This framing has been assuming, y'know, two people locked in swordplay. It also works for two people shooting at each other. One guy flinging spells and another flinging arrows. Swinging at an owlbear with an axe while it swings back with claws. As mentioned before, it's simple enough to emulate a one-sided clash, like me shooting at you from a distance while you have nothing to shoot back with. A character who lacks the means to deal damage just doesn't get to deal damage, even when they win the roll. Which is a very desirable strategy, because that's a pretty low-risk clash for you.
But like, imagine a giant. It's like 20 ft. tall. And it tries to stomp on you. Is that a clash? If you succeed the roll, do you get to deal damage straightforwardly? It's hard to picture you parrying and countering that sort of thing. Or, for that matter, the big stomp being the giant's standard counter when you miss a strike that you initiate. Same with a dragon breathing fire or a rhinoceros charging towards you.
I'm not against the idea that some monsters just defy clashing. Your skills of fencing and dueling and wrestling and whatnot aren't applicable when you're fighting something whose form of attack doesn't resemble those sorts of things at all. I have a handful of ideas to make these work.
- Every monster should have at least one minor attack they can do whenever a PC fails a clash roll, even if they have a special attack they prefer to use their own action for. In fact, if their special attack is really awesome, then their default minor attack can be less threatening. I like effects that impose disadvantages back, throw players around, or stun them for a whole turn. And these kinds of minor attacks can help a solo monster carry the same weight as a group of adversaries, able to "be their own henchmen."
- The big special attacks themselves aren't resolved with a clash roll at all. Rather, they are always telegraphed one turn in advance. The giant rears up its foot on one turn and stomps on the next. The dragon spends a turn inhaling and then exhales fire on the next. If the PCs fail to defend themselves adequately in that time, they will either certainly or almost certainly take lots of damage.
- Big monsters could instead be modeled as a series of attackable points. You cannot simply attack the whole creature. You have to call your shot. Going for the legs is separate from going for the eyes, which is separate from disabling the big flower on its back or the external organs containing its fuel. Corresponding with idea 1, there should be a standard counterattack the monster employs whenever you fail a clash roll against one of these pieces. It's high time someone actually make good on Ty Mindstorm's greatest mind storm.
This is where we're most clearly reminded of the advantages of D&D's combat system. The traditional attack roll vs AC system is a very elegantly "good for all purposes" mechanic that can model all kinds of combat abilities. But I can tolerate mechanics that are a lot more situational, especially when the system is otherwise very lite. There's something intriguing about monsters and enemies that defy a perfect template. Where each one needs to be designed as an almost unique problem to solve. A stat block that's just filling in the same set of blanks with numbers kinda sucks. I am eager to create stat blocks that instead say, "how does this monster challenge the players?" and "how do players go about defeating this monster?"
This next part is probably a bad idea and will likely not fit into the finished version of this mechanic. But it's a separate sort of mechanic that I've long been interested in exploring and maybe they can fit together.
Most dueling mechanics I've seen are 1) like a minigame within a greater combat ruleset, and 2) involve hidden choice followed by reveal.
In Ava Islam's Errant, each player takes 3 playing cards (a king, a queen, and a jack) and lays them face down in front of them. Then, they take turns guessing each other's cards. Guess correctly and you get to take an action (with a bonus based on the card), guess incorrectly and the other player gets to take an action (with a bonus based on the card).
That Prismatic Wasteland guy has one where players bet a secret number of d6s, then reveal. High number bet wins, but then both have to rolls the d6s and suffer that much damage.
White Hack has its auction system that comes down to a d20 blackjack roll. The hidden element is a d6 you roll and add to your normal target number, increasing your upper bound by a secret amount. Then everyone bets a number that's their lower limit. The higher the number you bet, the harder it'll be to roll a d20 between your lower limit and upper limit. But everyone's attacks resolve in the order of highest bet to lowest bet, and getting your attack in first could be a difference of life and death.
Probably my favorite, however, is Emmy Allen's Feint/Parry/Push dueling mechanic. At the start of each round, each combatant secretly picks a tactic. They simultaneously reveal and compare, with the results modifying rolls for that round. It's sort of a Rock-Paper-Scissors, but the specific effects of each tactic are asymmetrical and there's an actual result for when both sides pick the same tactic. Whereas Rock-Paper-Scissors is basically arbitrary, this at least gives you a reason to choose one option over another.
My instinctive way of implementing this is for a combatant to have the option of turning a clash into a duel if they're skilled at the form of attack they're using. By doing so, they can make additional maneuvers (perhaps specifically dueling maneuvers) with each subsequent clash, as long as they remain entangled with that foe. It could be a simple thing to implement the Feint/Parry/Push as the option you pick for how you want to modify the results of the following clash test, while still preserving the basic mechanic in its normal form.
I said before that I don't want you to be able to choose a talent at character creation that gives you a baseline modifier to clash tests. But with this, there could be an option for "skilled at lightsaber dueling" so that any clash you get into while wielding a lightsaber at least gives you an opportunity to do a dueling maneuver alongside the roll. Even so, these sorts of skills may be potent enough to offer them only as an advancement benefit, a skill that you train into as you level up.
I'm sure that the more time I spend with this, the more situations I'll need to account for. Mounted combat, warband brawls, dual-wielding, sneak attacks, and plenty more. I think I have a good idea how to bolt onto this a basic system for item quality and breakage. But I suspect I'll need help bringing this idea to fruition, if it has any legs at all.