[Game Unnamed Yet] My Combat Model is Running Out Of Control

[Amphetryon]

This is a fantasy RPG I'm working on.  Link to the combat rules.  Specific issues I'm grappling with to date:

• This system will need additional tweaks to handle combat outside the presented 1-on-1 model, as it would just get ugly, quickly, in its current state.
• Rules for ranged weapons, both fired and thrown, are sorely lacking because I'm unsure how to model them without changing some of the established parameters significantly.
• As the designer, my own perception of whether the combat is balanced, or heavily weighted toward either one-hit kills or painfully prolonged fights, is suspect.
  

I'm particularly looking for feedback on the above points, though I'd obviously appreciate folks pointing out glaring issues not addressed above.  Design influences for the combat model are primarily Warhammer Fantasy Roleplaying Game, 2e, and Wu Xing: The Ninja Crusade, if the references are helpful.

[stefoid]

Admission: I only scanned your doco.  but multi-character conflicts ARE hard.  I can offer the solution I used, maybe it can give you some ideas.

I categorised all possible actions as being either:  'acting neutrally',  'acting directly against another character', 'retaliation against another character', and resisting (defending only)

The arbitrary rule I added, purely to simplify the kind of out of control complexity you are worried about is 'a character may only have one type of action/retaliation per round, but as many instances of resist as required"

ORDER OF PLAY:
Often the order of play occurs without having to think about it much, but when complex, multi-character conflict occurs, the following is the suggested order:
1.   GM asks "Who is acting neutrally ?"   Every character that wants to do something that isn't a direct attempt to overcome or interfere with another character must state their intention now.
2.   GM asks "Who is acting offensively?"  Every character that wants to overcome or interfere with another character must state their intention, and their target now.
3.   GM asks "Who is reacting or resisting?"  Now every targeted character must state their intended reaction or resistance now.  This includes characters who were intending to act neutrally – they have a chance to change their intentions in light of being targeted.
4.   Any character that hasn't got an intention by now does nothing of consequence.  At this point, either the GM or the character's player can narrate why that is the case.

Your additional complication is that every type of action is timed to 0.5s granularity.  Perhaps that is the first thing you can look at abstracting away, in order to reduce complexity?

[Chris_Chinn]

Hi Amphetryon,

Welcome to the Forge!

Have you had a chance to playtest or try any of this in play?

Actions

There's a lot of games that have really detailed combat rules that are ideal at 1 vs. 1 combat, and can scale up to 5 vs. 5, provided the GM is organized, so if you're looking to stay within that scale, this seems manageable.   If you're looking to go larger than that, that's when things start to get complicated.

One thing that might be worth considering is this: is it necessary to track 20-segments of actions or would it be easier to simply track 4-5 "possible actions"?  In a way, this is effectively what D&D has done since 3rd edition with "Standard/Move/Full" being basically a 2 section combat action - except, of course, with 4-5 you actually have a lot of tactical options to use.

Ranged Combat

It seems like ranged combat would fit into your system just fine.  The two big questions are:

1.  Do ranged attacks work at the same "speed" as melee?  That is, would just making an "attack action" cover it, or would you want to make ranged attackers spend a "Draw/Load" action and/or an "Aim" action before they could attack?

2.  Does ranged defense depend primarily on an opposed roll, the target's distance, movement, and cover, or what?

Judgment

So, a lot of games range from one-hit fights to long battles- where do you want your game to land?  Do you want to see one or both ends of the scale occasionally, and if so, how often?  It's your game, you set what you want it to be, and you can playtest and adjust until you have something you're happy with.

Pretty much the big question everyone gets here at the Forge is "What do you want your game to do?" - because when you tell everyone what your goal is, then we can start giving good examples of games or mechanics that do similar things, or bring up ideas that might meet that.

Chris

[Amphetryon]

Thanks for the input so far, guys.

@Stefoid:
I will consider further the "aggressive/neutral/passive" model you've posited, but right now I'm not sure how much further tweaking its inclusion would require.

@Chris_Chinn
All I've had the chance to do so far is brainstorm combats with some other helpful critique-providers.

What I'd call a design goal of the 20 segment system and the 'buffet table' of actions is to make a combat round both well-defined and fluid.  Well-defined means that we know how long it takes for everyone to get a turn.  Fluid means that actions that don't fit neatly into those time packets feel less artificially constrained than the 3e D&D model you mentioned, where, for example, a good enough Jump check causes you to take 2 rounds to land.  Let me know if that doesn't make sense, please.

re: Ranged Combat... I would ideally like to model a 'Draw/Quickfire' option as well as a 'Draw/Aim/Fire' option that makes Ranged Combat marginally less efficient than normal melee attacks, as a price paid for being generally less vulnerable to counterattack.  Ideally, most of the actions of the game - in and out of combat - will depend primarily on opposed roll, and secondarily on skill, with range and cover and such being tertiary concerns.

As far as combat lethality is concerned, the intent is to make a combat system that feels inherently dangerous while only being truly deadly in exceptional circumstances; I would like "running away" to seem a viable, intelligent option without making it the default.  (On that note, the movement in combat rules probably warrant additional scrutiny, as situations where it's effectively impossible to engage in combat may be too likely at present).  To refer back to the 3e D&D model, I do not want injuries to work on the binary system of 'full capacity or dead' model they used, which is why I wanted an Injury Threshold system.  Combat of 3 rounds or more will, ideally, be a dangerous thing, but dying before you get a chance to respond should be all but impossible.  I'd like combat against multiple foes to get exponentially more difficult for every instance in which you're outnumbered two to one.

[btrc]

I may be the biggest rule, mechanic and realism geek here, so I'll chime in on the rules so far:

Generic first things for anyone trying to model something, either "realisitically" or "genre-realistically":

1) Know your limits. Example: If your ability to model something is not at a 1% level, don't use percentile dice as a resolution system or a 1% scale of skill or attribute improvement. It would just add complexity without adding any realism.

2) People do not want to leaf through rules in the middle of a fight. Make as much of the situation as intuitive as possible, and failing that, make it as structured and straightforward as possible (i.e. step 1, step 2, step 3), with as much consistency as you can manage.

Attributes:
I can envision situations in which balance and coordination are entirely separate attributes, but most of the time this would be due to a physical defect rather than an inherent bodily nature. Very few Stradivarius players are fumble-footed (low balance), and I imagine most Olympic gymnasts have no trouble tying their shoes (low coordination). In a melee-heavy system, the correspondence between the two attributes should be very high, so is it necessary to have both attributes?

Round structure:
Half-second seems a bit fine for a melee-heavy system. Granted, this allows for a lot of neat, specific actions, but a melee combat is going to resolve in aggregate based on the overall skill, armor and weapon differential. If the final outcome is the most important result, then the fewest steps that get there with the most drama is more desireable than a lot of mechanical steps that get there with the most number-crunching.

Combat:
I can add, subtract, multiply, divide and do cube roots in my head, so personally, I have to strongly curb the urge to put complex math into mechanics. Anything in multiplication that involves rounding is probably not a great idea in a combat step.

I concur with others that your system can bog in other than 1-on-1. Since your attributes are already reflected in your chance to hit (I presume), a separate comparison step with each opponent does not seem to be necessary prior to the die roll. Just have your attribute affect your die roll and their attribute affects theirs.

Muscle modifier: Is this always a positive value? If not, then you get unusual damage results because of your weapon use limits.

Dice rolls: Need to do what they are supposed to do, but with as few dice rolls as possible. Sometimes it is possible to combine rolls to do multiple game mechanic functions, sometimes it is not. For instance, you could have the chance that a gun jams be part of an attack roll, or link damage to the quality of success on an attack roll. I'm specifically looking at your criticals, which tack an extra roll onto things, but which you might be able to do without. Similarly, you have a 1 to X spread on weapon damages and a 1 to X combat roll. You might be able to get away with something like:

weapon damage is the amount your roll beats your opponent's roll:

+2 if you are using a light weapon
+4 if you are using a medium weapon
+6 if you are using a heavy weapon

+0 if making a "light" attack
+muscle mod if making a normal attack
+muscle mod + 2 if making a heavy attack

At first glance, this might give you most of the variability you desire in damage effects, but with several fewer rolls and no multiplying. I'm not sure if it gives too much variability on light weapons and not enough on heavies, but I'm just throwing it out as an example of streamlining.

Greg Porter

[Amphetryon]

Thanks for your thoughts, btrc.

Re: Attributes
I made the conscious choice to separate Coordination and Balance because the ability to maneuver in space is not the same thing, to my mind and experience, as fine motor skills and eye-hand coordination.  A tight-rope walker uses a very different skills-set than an archer, for example.  For another example, I was always reasonably good at dodgeball, but can't hit a baseball well at all )and there are many professional examples of good fielders with poor hitting skills, and vice-versa).  Also, the attributes modifiers are listed at the top of the document linked; yes it's always a positive modifier or no modifier at all, as indicated.

Re: Combat
The attributes that contribute to the chance to hit are specifically listed: the attacker's Coordination and the defender's Balance, relative to each other.  Your inclusion of the parenthetical "I presume" indicates there's some confusion about this, is that correct?  Without a comparison of die rolls, we're left - as I understand your suggestion - with rolls against a static DC, something I'm specifically trying to avoid as much as possible in the system; it's included only as a passive defense at present.  I would prefer not to mimic Shadowrun's "counting 6's" variety of combat, because it does not model the sort of oppositional combat structure I'd like to create.

Re: Dice rolls
I have yet to see a system where "the chance that a gun jams" did anything more in the long run than punish players to a significantly greater degree than it punished NPCs, so I do not want to include any sort of fumble/catastrophe chance beyond simply missing.  That doesn't mean such a system doesn't exist or is impossible, simply that I've never seen one and am not a fan of that additional complication.  Damage is already linked to the 'quality of success on the attack roll,' as the differential between your Attack roll and your opponent's Defense roll is added to the damage.

I think I'm missing part of the rationale behind your weapon damage suggestion, because it reads from here as an extra step rather than a streamlining.  If it's not an extra step, then I think it is intended to replace multiple aspects of my current model, and I am unsure which aspects you propose to replace.

[btrc]

Regarding attributes, stuff like this is always a designer decision. I was just trying to see the rationale behind it in the interest of streamlining.

For combat, you have your attacker's Coordination and defender's Balance, but isn't there also a separate bonus if one exceeds the other? This is something where you have to make a separate check against each opponent, which would slow down a melee a little. I wasn't suggesting going against a fixed number, just that a separate comparison step for each of multiple defenders could add some slog to things.

On dice, the "gun jams" was just an example of making one roll do multiple things. In your case, you have a "critical", but that critical requires an additional roll, rather than being an effect generated by already existing parameters. It would be like a "gun jam" roll that then forces me to roll again to see the nature of the jam.

On the weapon damage suggestion: You have a 1d10 roll to hit, then potentially another roll for a critical, and a third roll for the damage of the weapon. All of these rolls are in the same ballpark in terms of the dice used. Specifically, you have a success margin already existing in the roll between you and the defender. If you get a really good hit (like a critical), then odds are that it is going to be a more damaging hit than an attack that is barely successful. However, the way you have it, it is possible to get a marginal hit that rolls well for damage and a critical hit that rolls minimum damage. Since you have damage as a random roll and "to hit" as a random roll, why not just factor damage into the success of the random roll and knock off an entire step?

So, let me give an example:

Yours:
1) Roll to hit
2) if get a critical, roll again
3) Roll for damage

Mine:
1) Roll to hit. If you hit, take the amount you hit by and add it to a fixed damage for the weapon type.

So if I know that my damage in combat because of my weapon, attack, muscles, etc. is (success margin + 4), the combat roll instantly tells me the damage my opponent has to deal with.

You see, if you hit your opponent by a lot, you automatically get the benefit of increased damage on them without needing a separate critical roll. A "critical" is just a really good roll to strike your opponent.

As a side effect, this actually gives you the potential for more variation in weapon types. Instead of d4, d6 and d10 weapons you can have additions of every amount from +1 to +10 as a fixed value for a particular weapon. So, instead of all "light" weapons being d4, you could have +1, +2 and +3 additions to the success of your combat roll.

Greg

[Chris_Chinn]

Hi Amphetryon,

I guess something that would be useful for helping guide the discussion is to let us know how set in stone what you already have is, vs. how fast you want things to run for how many-ish combatants?

I mean, there's a lot of games that use maps and run well with lots of combatants (Savage Worlds), but they're usually highly simplified, and lots of games that have detailed combat choices and run well with up to 5 vs. 5 numbers (Burning Wheel, Riddle of Steel) and a whole lot between.

Knowing what you're aiming for on the spectrum is going to be necessary to pick out what kind of advice you're really looking for.

Chris

[Amphetryon]

Re: "separate bonus"
It's a single, static check of attacker's Coordination vs defender's Balance, which should only 'slow down' combat in the smallest amount and only in the first round, since either the game-master already has access to the relevant stats to compare or can get and share the necessary information in a single exchange ("What are your Coordination and Balance?" '53 and 37' "Okay, you apply +1 to your attack rolls, then").  Subsequent rounds, and combats, the GM would not need to ask this for comparison unless notes went missing.

Re: criticals.
Criticals would be exceedingly rare in this system, since an intelligent defender would not pair 1s against 1s  if possible.  They are included to emulate a lucky shot, which may still only graze the temple or throat and thereby do less than spectacular damage.  They're also intended to help shape the system overall in a way that doesn't favor rolling high or rolling low especially.

Re: what's set in stone.

• I want to keep a declarative initiative round; it need not be 20 segments of 1/2 a second each. 
  
• I want the choice of going first or reacting to combat to be a viable one, rather than having going first as the only good choice. 
  
  • I want to keep actions measured in segments to allow for a a variety of actions within a combat round, and to allow for actions to spill over into subsequent rounds such that combat order does not feel artificial.
    
  • I want to model combat so that 1-to-1 combat is the most common model, 2-to-1 combat gives a reasonable edge to the group vs the solo, and so that more than 4 combatants against 1 means the 1 should run away.  At the same time, I want the system to actually be able to model 4 on 1 (and up) combats, which I'm not convinced is something I've achieved as of yet.
  • I want to figure out if the combat model, as presented, is too lethal, or not dangerous enough to keep players engaged and excited, given the '3 rounds of combat should be dangerous' comment, made before, as the baseline.
    
  • I want to figure out of the current combat model can handle reinforcements arriving after combat starts, without everything going all wahooni-shaped.
    
    • I want to assign reasonable segment costs to ranged attacks in combat, both mechanical weapons(bows) and thrown weapons, with a "draw/quickfire" as well as a "draw/aim/fire" option available, with the latter marginally less efficient and damaging in combat than 'normal' melee attacks, and the former less efficient and damaging than that.
      
    • I want to figure out a way to model movement in combat, without radically rethinking the entire paradigm, that doesn't result in situations where combat can never be joined if the first to move simply runs away.
      

[Chris_Chinn]

Hi Amphetryon,

That's a great set of clear goals!  It seems like a lot of those are detailed on the specifics of numbers and probably are best found out with a bit of playtesting.

The issue of folks staying always out of reach is something a lot of games work with, differently:

- Old school D&D has pretty swingy odds on initiative- basically the only way to stay constantly out of reach is to win initiative and run, run, run (provided your movement is equal/greater than the opponent's).

- Games like Burning Wheel use an initial "positioning test" sort of thing- winner gets to take an advantageous position.  For your game, maybe it gives them a number of free segments to take movement actions?   Or, if they're holding position, it lets them dictate at what distance the opponent can get to before they get to start firing ranged weapons/charging/whatever. 

- A lot of minis war games do stuff like split up movement and make melee "sticky" - that is, if you're armed with melee weapons and get within reach of an opponent, they can't just flee.  In some cases you get free attacks, in some cases, they just can't flee at all until they've managed to injure/stun/whatever to get a second to run.

I'd probably look around at some war games for movement ideas and do some playtesting to see about the numbers, and perhaps get some inspiration about what might be a better approach to avoid the movement problem you mention?

Chris