How RPG Combat Can Help the Story

A classical painting of two women fighting a duel.
This post is 1 in the series: The Challenge of RPG Combat

If your RPG upbringing was anything like mine, you were taught from a young age that combat and story are eternally opposed to each other. You were raised to believe that players are either obsessed with combat to the point of ignoring the narrative or carefree roleplayers who would never deign to roll for initiative. “Combat” didn’t even have to mean physical violence. It referred to any extended conflict system, from Spycraft’s Chase rules to Burning Wheel’s Duel of Wits. It was all a power gamer’s dream, anathema to a good story.

But it doesn’t have to work that way; combat and narrative can coexist in peaceful harmony! In fact, combat can be a serious boon to a campaign’s story, provided it’s used the right way.

Combat Makes Problems Matter

A dice roll’s most basic function is to provide a mechanical resolution to a narrative problem. Once dice are involved, the possibility of failure makes success far more satisfying. Players are much more likely to remember accomplishments if they have to roll for them.

But this simple dice mechanic has a problem: it’s difficult to make any roll stand out from the others. Narratively, the final battle has more weight than problems early in the story, but without mechanical backing, that contrast starts to fade. GMs can try to address this by giving more important rolls higher difficulties, but difficulty can only go so high before it becomes untenable. Likewise, the GM can call for multiple rolls to resolve important problems, but that becomes redundant fast.

Combat provides a powerful solution. When a problem is expanded from a signal roll to more involved combat rules, it sticks in the players’ minds. Combat allows for the back and forth that any good climax should have. It’s no simple matter to ratify the peace treaty in parliament; the PCs have to trade barbs with the war hawks and fight for every concession. To be sure, a really skilled GM can accomplish the same thing with description, but not all of us are so gifted.

This use of combat depends on contrast. If most problems are resolved with a single roll, then a full combat here and there really sticks out. If combat is used for everything, that contrast will fade. Fortunately, this isn’t a huge risk in most cases, since even the best combat systems take enough time that using them for everything is impractical.*

Combat Gives Players Agency

Narrative agency is critical to creating a strong story. A good GM listens to what players want and says yes whenever possible, incorporating those actions into a greater whole. But even with a skilled game master, players are far more limited when it comes to mechanical agency. They might be able to spend meta currency or get a bonus from their fellow PCs, but when it comes to a single roll, the player usually just has to pick up the dice and hope for the best.

Combat gives players a greater level of mechanical control. They have an objective and multiple ways to accomplish it. They can choose to play aggressively or take the cautious route, and they can choose which skills to bring to bear. In some systems, they even change what their objective is midway through the combat.

Players with agency are more likely to invest in the story because they have a hand in shaping it. The story isn’t just something they’re handed by the GM; it’s a group creation. What’s more, combat offers a safe and convenient arena for those choices to play out. Since most combat systems are self-contained, players can fully take over the driver’s seat without the risk of crashing the whole campaign into a mountain. At the same time, the different choices have all been created ahead of time, so the GM doesn’t have to spend precious energy thinking up options in the moment.

Once players are invested in the combat’s outcome, they’ll care more about the story it’s part of. This kind of energy is priceless for GMs because it keeps everyone active and roleplaying, even through occasional dull patches.

Combat Is Mechanically Engaging

Beyond granting extra agency, well-designed combat systems are simply fun. There are no obvious choices for victory, so the players have to really think about their options and then decide what the best choices are. You might recognize this as the same reason people play board games: it’s fun to strategize and win!

However, there’s a reason to use purpose-built RPG combat systems rather than pausing your campaign to play Settlers of Catan for a while. In an RPG, the players finally get to make full use of all the gear and abilities they’ve accumulated since character creation. Did they pick up a jeweled fan that makes courtiers stop and stare? Time to use that in their oratory contest. Maybe they just learned a sword trick that lets them easily disarm an opponent. Now they get to strut their stuff.

This is where the divide between narrative and combat most often rears its ugly head. How can players who are concerned with using their gear and abilities to win a combat possibly become interested in the story? The key is that they’re using mechanical aspects of their characters, and their characters are part of the story. Since it’s certainly story-relevant when Luke Skywalker uses his piloting abilities to destroy the Death Star, then it can also be relevant in an RPG where the PCs drive their sandskimmer to victory in the Wasteland Derby.

So long as the combat is relevant to the story, the players’ tactics will be too. This is true even when the players are 100% engrossed in the mechanics and couldn’t give a toss about the campaign’s plot. By giving them combat goals that move the plot forward, the GM can turn all that tactical energy toward narrative ends.

Combat Creates Unforeseen Twists

Perhaps the greatest draw of RPGs as a storytelling medium is that no one knows for certain what is going to happen. The GM usually has a general plan, but that plan must be kept vague to respond to whatever the PCs throw at it. Even so, there’s a limit to human creativity, and few GMs can produce twists at the rate their players might prefer. Some systems try to force the GM to be more creative, like Edge of the Empire with its unusual dice mechanic, but in most cases, this just stretches the GM’s already-limited creativity.

Combat is a major boon in this department. The rules have codified effects for unusual die results, allowing the GM to relax for a moment as the burden of creating unpredictable twists is temporarily taken off their shoulders. No one was planning for PC Tanya to drop their sword while dueling the villain, but that’s what the dice say, so now Tanya has to deal with it. And if you thought that was surprising, just wait until PC Alex’s sleep spell misfires and summons a sheep-demon.

These twists have a special zing to them because they weren’t decided on by any human mind, but by impartial rules. They make for great anecdotes to share with friends, but more importantly, they make the game feel alive, like anything could happen. That kind of suspense keeps players on the edge of their seats, paying close attention and wondering what might be around the next corner.

Of course, most games don’t want too much unpredictability because that makes it impossible to tell a coherent story. But in moderation, these twists can liven up a session that might otherwise have gone by completely unremarked.

So if combat is so useful to RPG storytelling, why does it have such a bad reputation? Have roleplayers simply been wrong all this time, unjustly laying problems at combat’s feet? No, no, they have not. The sad truth is that despite all of combat’s potential, the vast majority of combat systems don’t work properly and are a drain on any story. I’ll explain why next week in part two of this series: Why RPG Combat Is Broken.

Treat your friends to an evening of dark ritual murder. In a fictional game scenario, of course. Uncover your lost memories and save the day in our stand-alone game, The Voyage.



  1. Michael Campbell

    I’ld like to point out that “combat” and “task resolution” are two different things, even task resolution involving opposed rolls.
    Combat involves risk of injury (and frequently risk of death).

    • Jonny Wilson

      What about something like escaping a burning building? That carries a risk of injury and death but is certainly not combat, unless you’re having to fight your way out or something similar.

      • Michael Campbell

        It’s amazing how often the healing rules don’t differentiate between combat injuries and non-combat injuries.

        Combat is a subset of task resolution. Falling damage taken when one leaps from a burning building’s window but fails the landing roll is injurious task resolution but not actual combat.

        The first rule of cinema is:- sell the stakes.
        So too, dialing up the stakes by making the spectrum range from, jump and take no injury to, jump and take some injury through to burn to death because you didn’t even try.
        You’re doing a good job out lining the stakes.

  2. Michael Campbell

    Also combat often leads directly to narrative.
    After defeating the goblins, you find a map showing the goblin plan to attack to local village. But it also shows the location of the goblin lair.
    So successful combat yielded a story option. And unsuccessful combat wouldn’t.

    Seems to me, there’s more to combat than simple grognard appeasement.

  3. Michael Campbell

    Plus combat places a clock on the party.
    Do we have enough health points to stay down here all night?
    Does that question in and of itself; generate tension???

  4. Grady Elliott

    I just wanted to say thank you for writing this, because I’ve been inundated for *decades* by the opionions RPG-ers who insist that combat and story are mutually incompatible. It’s nice to hear from someone else understands that isn’t the case.

    • Michael Campbell

      It’s an extension of an ancient debate. Is role playing about problem solving & combat resolution or is it about acting out your character & imagining the character’s sensory experiences.

      \i{And the simple answer.
      Why pick one, when you can have the mix of both that works best for your group!}

    • Oren Ashkenazi

      Hey Grady, you’re very much welcome, and I’m glad you enjoyed the article. I firmly believe combat has a useful place in RPG storytelling. Unfortunately, the state of RPG combat is often so bad that it’s understandable why so many people think that it can’t be useful to telling a story, which I’ll cover more in the second and third installment of this series.

      • Michael Campbell

        To a certain degree some people prefer to see role-playing games as a means of generating a series of tactical combat encounters.
        It gives their minis an outing and challenges their minds in manner similar to miniature war-gaming but with a smaller commitment (30 minutes battles rather than 3 hours).

        As a society (role-players), we bemoan the grognards.
        But what we should say is…it takes all sorts to make a community.
        Live and let live.

  5. Recursive Rabbit

    It’s easy to mess up and create the divide, and video games are very much prone to the same mistakes.

    Some points about combat I think are generally worth discussing:

    Don’t make every fight “kill or be killed.” The Angry GM wrote an article I may need to dig up with a useful question to ask when a possible combat comes up: “What is the dramatic question?” He also brings up motivations behind the actors.

    How is violence perceived in the setting: In a D&D game, it’s usually accepted that a bunch of adventurers can raid the local chaotic evil civilization and take their stuff. But, what if you’re in a modern setting? Violence in the human world is going to have a lot more consequences in my Changeling game, like police investigation.

    • Michael Campbell

      This is one area where pass-fail systems beat, “Pass. And… / Fail. But…” systems.

      What is the magnitude of failure?
      In an attempt to get a discount on a purchase through using one’s fast-talk skill. The magnitude might be being told to leave the store by the proprietor but it’ll probably be just paying full retail price.
      Jumping over a bottomless pit in order to pick up a few copper pieces on the other side? Suddenly the magnitude of fail is certain death.
      Giving Game Masters an understanding of the verisimilitude attached to failure, is better than attaching a penalty within the rules because the spectrum of ramifications of failure can change radically.

      A simple failure isn’t boring. Not if the Game Master knows what a good set of consequences should be.

      Pass/Fail is like an “Or” gate in programming…or is it Boolean algebra???
      Much like choosing; the left door or the right door:- changes the story. So too, pass/fail, changes the story.

  6. Leon

    This is only tangentially on topic. If you squint. And turn your head. And it’s starting to get dark out, so that turning the lights on actually makes the house feel darker.

    In a dice pool system(d6) with a fixed tn of 6, pool size averaging about 8-10 plus bonus and bogie dice. Does it make sense to replace up to 5 bogie dice with a single d6(with a tn), anything above the tn cancels one hit?

    Fewer dice on the table. Makes a faster game.
    Opposed rolls (or rolls where many targets are defending themselves) would be easier to manage.

    Would it just feel wrong for some reason?
    Could players feel cheated when a second and third die are introduced against a single target or task?

    If you don’t know (I didn’t), statistically d6 with a variable tn up to 5 and a dice pool (tn6) up to 5 produce very similar results – except for extreme failures with the dice pool but there is a limit to how killed you can get in any scenario.

    • Oren Ashkenazi

      I’m afraid I’m not quite sure what you’re describing here. By “bogie” dice, do you mean penalty dice that subtract successes? If so, can you give me more detail on what you’d be replacing them with? Maybe an example?

      • Leon

        Yes, sorry, penalty dice.
        I think the main advantage would be in situations where PC face multiple defenders. Especially in a science fiction setting where you could have cyborgs with autonomous arms, or partitioned brains so they can command multiple weapons platforms simultaneously.
        The concern is that having the dice pool, and then looking for a target number on the penalty dice could just feel plain weird.

        • Michael Campbell

          You might try using a crayon to fill in the pips of the one different dice, leaving the other die as having black pips.
          Call that dice the control dice.
          And then let the colour of the pips in that dice dictated the target number, so 4, 5 or 6 are red and have a target number of 6, 2 & 3 are yellow pips and have a target number of 5+ and 1 is green and has a target number of 4+.

          Indeed if you kept the control die in a pill dispenser, you could have varying difficulties associated with the the dice selected for the individual throw.
          Easy: 1, 2 & 3 Green. 4 & 5 Yellow
          Medium: 1 & 2 Green, 3 & 4 Yellow
          Hard: 1 Green, 2 & 3 Yellow.

      • Leon

        the penalty die is just the same as the other dice, just a different colour.
        dice pool is 10d6 tn6
        penalty die is 1d6 tn-x

        it works almost exactly the same as penalty dice pool nd6.

        • Leon

          For clarity;
          the penalty die is just the same as the other dice, just a different colour.
          dice pool is #d6 tn6
          penalty die is 1d6 tn#

          it works almost exactly the same as penalty dice pool #d6.

      • Oren Ashkenazi

        I’m still having some trouble here, can you describe an example of what rolling these dice would look like?

        • Leon

          (c)d6(fixed tn6) – (1 to 5)d6(fixed tn6)
          Is replaced by
          (c)d6(fixed tn6) – 1d6 (tn 1 to 5)
          The probibility curve is almost identical.

          Engaging multiple defenders you could streamline things by using the same result from (c)d6 against each defender (each defender represented by their own penalty die.)
          Defenders working together effectively could stack their successes.

          Sade’s12d6 = 2 = S
          S – Jake’s 2 of d6tn3 = S – 1 = 1
          S – Paul’s d6tn2 & Dons d6tn4 = S – 2 = 0
          S – Bruno’s d6tn5 = S – 0 = 2
          Sade scored a marginal success against Jake, total failure aganst Paul & Don, and convincing success against Bruno.

          I think this could work because with a fixed tn of 6 the average roll is going to be 1, 2 if your lucky, 3 if your very lucky. And your never gaurenteed success.
          Extra tough opponents or dangerous tasks could have extra dice, but i think in those cases it would be simpler to revert back to the standard penalty dice, especially since such tasks deserve the extra focus and balancing the target numbers for tiny variable dice pools would only be distracting.

        • Leon

          And sorry it took so long to get back, my daughter woke up early.

        • Leon

          (poke) I hope your well.

        • Oren Ashkenazi

          I appreciate you putting this example together but I’m afraid I just can’t follow the math notation. This might be easier if I weren’t an English Major, but all the parenthesis and variables are losing me.

          I suspect you would have that problem with players as well, so I think the first thing to do is figure out a notation system that’s easier for non-math types to follow. Perhaps find an existing RPG that’s similar and phrase it in those terms?

          • Leon

            ok, try this on.

            unless stated otherwise the target number is 6 (though i like 1 because it’s quicker to spot the 1s)

            Sade a cyborg is shooting at a space pirate. Her player rolls 12 dice representing her stats, skills & augmentations*. But there are difficulties, she is still a bit drunk from last night, the target is standing behind a vehicle, and the rising sun is shining in her eyes. These difficulties add up to 4 penalty dice.
            Sade’s player sighs, ’16 dice is just too many dice.’
            the GM says, ‘Instead of rolling 4 penalty dice, you can roll one penalty die with a target number of 4. The probability curve is almost identical.’
            They do that.

            Later Sade is engaging ambushed by a space pirate, a ninja, and a robot.
            The experienced pirate puts himself in a position to give Sade 4 difficulties.
            The cocky ninja comes at her with his katana (what else), he does not know how deadly Sade is but his speed gives her 2 difficulties.
            the robot is a bit depressed, he is struggling to hold his big head up and he’s not sure why he should care. He only gives Sade 1 difficulty.
            To streamline combat, Sade’s player will roll stats+skills+augmentations only once, and play that score against the difficulty factor each enemy individually. ‘Rolling 4 separate dice pools is far too many dice.’ Sade’s player says.
            ‘Wait.’ the GM says ‘you only have to roll your own dice pool, the pirate, the ninja, and the robot each roll only one die, with their individual target number.’

            bang, bang, bang.

            I hope the tone came out right

            *augmentations include weapons, cybernetics, drug effects, and even cosmetics and have varied effects depending on the situation, like conditions in mouse guard.

          • Oren Ashkenazi

            Hmm, well that does sound doable, although I think you could save some time by having static difficulties rather than penalty dice.

          • Michael Campbell

            @ Leon
            Actually I think you might be looking at a problem that actually doesn’t exist.Too many dice. Just stop using those 16 mm dice and start using those 12 mm ones.

            I think you can take a leaf out of long division.

            Have two different die colours and then just use long division to shrink the number of die being cast.
            So a target number of six just uses the number of die developed for the character.
            A target number 5+ uses half the number of die with the possibility of a remainder die if the developed number of die was odd. With a target number of 6 for any and all remainder die.
            A target number of 4+ uses one third the number of developed die. With the possibility of; none, one or two, remainder die.
            A target number of 3+ uses one quarter the number of developed die, with between zero and three remainder die.
            And a target number of 2+ uses one fifth the number of developed die, with between zero and four remainder die being used.

            In this way, if someone is throwing some cumbersome number of die like 17 of `em.
            You pick a divisor of 4, take on target numbers of 3+ and roll four regular die and one remainder dice.
            The result should be 2.83′ successes on average (2.66′ regular + 0.16′ remainder) which is what 17 dice with a target number of six are going for (17 x 0.16′).

          • Leon

            @ Oren

            Do you mean to keep the target number the same, and add dice to the initial stat+skill+augmentation depending on the bonuses and “difficulties”?

            If so, how do you deal with multiple meaningful target’s – the special agents, troublesome weapons, monsters & super mooks?
            I suppose it’s like I said earlier, If they matter they deserve the attention and drama of being rolled against as an individual entity, and if they don’t, they don’t.

            @ Michael

            Got 7mm dice. It’s more about the way they scatter on the table – with a single cluster of dice we could just throw them into the backgammon board and know at a glance who’s frowning, who’s smiling and who’s complaining.

            The whole reason for penalty dice/die’s is that people don’t like having their dice taken away from them.
            Also, no stat+skill+augmentation combination should ever get much higher than 14, they are each rolled on a single d6, and the characters don’t level up, They’re already as smart and fit and strong as they’re going to get (though, they can change their focus), they gain skills, and conditions.
            Injury, disgrace and despair is where dice are taken away

          • Leon

            Sorry, by target number I meant the required number of hits/passes/successes.

          • Michael Campbell

            @ Leon
            Yeah, I had been tinkering around with a system of variable target numbers.
            “Add the pips & subtract the numerals.”
            So long as your total of pips outnumbers your numerals, you’ve succeeded.

            Making hits happen & hurt will speed the most time consuming part of play (combat). The old regime of:- Swing, miss, swing, miss, swing, miss, swing, hit, kill. Has taught a bad habit to RPG designers.
            Better to have a more entertaining combat system:- Swing, miss, overstep, counter-strike, hurt, concussion penalty.
            Suddenly choosing wisely becomes a priority.

          • Oren Ashkenazi

            I can’t really go into more detail at this point, but what I mean is that rather than using penalty dice, in most cases it’s easier to say that a PC needs a certain number of passes in order to succeed at a task.

            For example, in Mouse Guard, every d6 that rolls 4-6 is a pass. An easy roll usually needs 1 or 2 passes to succeed. A moderate roll needs 3, a difficult roll needs 4.

            That model just cuts down on the number of dice you need to roll. Now if the PC is up against an important enemy who you think deserves their own dice, you can roll that enemy’s skill+attribute against the PC’s and see who gets more passes.

          • Michael Campbell

            @ Oren

            The trouble is the granularity of the spectrum.
            If I want, Pathetic / Lousy / Dull / Average / Good / Better / Best; as the attribute spectrum, I’m really looking at a spectrum of 1-7 die (maybe 0-6). I kick in a skill spectrum of 0-7 and suddenly there are rolls of 14 die at the high end.
            And with averages, it’s still a lot of dice, 5 (four for the attribute and one for a basic level of training)

            If I go for a spectrum of Pathetic / Average / Best, I can get the skills down to 0-4. And while that means maxing out at 7 die. It also means any bonus dice thrown into the mix has a massive influence. After all the average die roll will be three die (two for the attribute and then one for the basic level of training).

            Now while a 20% increase from a bonus die is a strong modifier, a 33% increase is probably going to be game breaking.

          • Michael Campbell

            @ Oren
            Also double the number of die cast if you’re using random target numbers. I forgot to mention that.

          • Leon

            Thanks Oren,
            I really appreciate the input.

  7. Lizard with Hat

    This is pure gold for me, Oren. Thanks a bunch
    I’m working on homebrew system for my groups and combat was giving my trouble.
    My players a very adaptive so the go along with the rules as I explain theme (they are just great but I would like to give them more story while they fight – and still have enough options left. Most Combatsystems (even the fun ones) are still really slow.

    To put i simple: I can’t wait for the next entire of this series.

    On a somewhat related note: Thanks for the stuff you put here on Mythcreants, helped me a lot with my GM-ing

    • Oren Ashkenazi

      Awesome, glad you enjoyed it! Length is one of the biggest problems combat systems have, in my experience. Even a good combat system flags when you’ve been at it for two hours.

Leave a Comment

Please see our comments policy (updated 03/28/20) and our privacy policy for details on how we moderate comments and who receives your information.