Sunday, April 18, 2021

Injury and the Abstract Combat Round


Let me give a natural description of how combat works in D&D. Each person gets to make (at least one) to hit roll against an enemy per round to see if they can damage them. How high they have to roll to hit depends on how well-armored their enemy is, because being well-armored protects one from injury. If they do hit, they get to roll to see how effective their blow was by rolling damage. Different weapons do different damage, since some are more deadly than others. These weapons injure people more seriously in the way that getting hit by a sword (or a chainsaw) is likely to injure you worse than getting sliced by a switchblade. Inexperienced characters are easier to kill because they "can take" or "withstand" less damage than more experienced characters who have many more hit points and "can take" many more blows.  

Here are two common dissatisfactions with D&D style combat. The first is that there's something weird about how you can stab Lancelot with a knife 24 times with no chance of killing him. It suggests video game thinking, as if he had a life bar that could be "full" or "empty", a life bar that grew with each level, and shrunk eat time you got hacked with a sword. (Indeed, it may well have been the source of that video game thinking in part.)

One might object to this on grounds of realism. This might be a big deal for you. It might be hard for you to maintain suspension of disbelief with all these stabbings. Or, you may be into military simulations, and think this is missing out on the fun of simulating combat. Personally, I happen to not need that much realism in combat to sustain my immersion in the fiction of the game, and I'm not into military simulations. My objection to this picture is rather an aesthetic one, both a repulsion to the aesthetic of the life bar, as well as to the associated picture of people taking and dishing out endless beatings, like when the Hulk fights Thor. 

The second objection is to gameplay and the role it grants violence. Standard D&D makes fighting weirdly predictable, even with swingy D&D dice. As a seasoned player, I can weigh the odds, and engage with confidence that I have a big HP cushion to count on, sauntering into a fight with confidence that it'll be four or five or ten rounds before I need to reassess. 

We might want a game where the stakes to violence were potentially higher. Where death or actual injury with consequences were constantly hazarded by fighting. We might want this for a variety of reasons, for example, if we wished violence to play a different role in our game, or if we thought it would be more fun if things stayed on the knife's edge--at least when knives were involved. 

The Old School Fix

One old school reply to at least the first aesthetic objection is to emphasize the abstract combat round. In old editions of D&D (OD&D for example), a combat round lasted one minute. During that one minute all kinds of swashbuckling, ripostes, presses, and close shaves get abstracted down to a couple of rolls. Elaborating on this abstracted round, we might say that hit points and damage are equally abstract, and that they do not consist literal stabbings and such, but represent an abstract combination of luck, grit, pluck, fortitude, situational advantage, and skill that have to be overcome for a side to prevail in the course of a fight. Armor makes it harder to prevail against someone in this way, and being better at fighting makes it easier.

I think this is aesthetically much better than the picture on which every successful roll is a literal blow that lands on your opponent. But in conversation recently, Anne Hunter mentioned that although people say this all the time, no one playing retro-games really thinks about it this way at the table. This struck me as true. I got curious why. Here are some speculations.

(1) The names of things are misleading. You roll "to hit". That sure sounds like a sword swing connecting. Furthermore, only if your blow "hits" do you do "damage". That sure sounds like injuring someone. Spells that give you back hp have names like "cure light wounds" "cure moderate wounds", etc., and are described as "healing". That sure sounds like it's making literal wounds vanish! 

(2) There are mechanics that only make sense on the picture of actual individual swings and actual wounds. For example, for every single shot you fire with a bow you mark off a single arrow. So every shot is literally firing a single arrow. This suggests strongly that a hit really is getting hit by a specific arrow. Or again, in old editions of D&D, i.e. AD&D 1E, if you lose hp then you need to recover it slowly by resting for days (1-2 hp a day) in the way that suggests recovering from actual injuries. 

(3) But I think the biggest culprit here is how as players and DMs we describe combat and enter into it imaginatively. aPhilotomy is a proponent of the abstract combat round solution. Here is a nice quotation from Philotomy's OD&D Musings about the importance of giving the right descriptions of combat in order to support the abstract combat round approach to hp.  

I agree with Philotomy that we need to change our descriptions if we're to have an abstract combat round approach to hp. But I'm skeptical that merely sprinkling in DM narrated description now and again is going to get people out of the concrete headspace. 

The problem is that when I, the DM, ask you at the start of a round, as the rules require, "What are you doing this round?", it's utterly natural for you to say, "I'm going to stab him with my sword." And then you roll to see if your proposed action succeeds, and then if it does succeed, you roll further to see how effective it is. So if you hit, you naturally imagine yourself as in fact stabbing someone, and you imagine them as taking more or less damage from being cut by your weapon. 

Ron Edwards call this sequence of description and dice rolls, "fortune at the end". The idea is that you appeal to fortune after you have fully described what you are doing. The fortune shows you if the described action succeeds, and if it does, how effective that action is. He contrasts this with "fortune in the middle", where you describe a general goal, but then roll and narrate what actions you perform to fit the outcome of the die roll. I'll come back to "fortune in the middle" in a little bit.

The point for now is that the "fortune at the end" kind of sequential ordering, when combined with the names for the rolls and the variable weapon damage, really pushes players to imagine that their character is trying to stab their opponent all right, but the opponent turns aside the sword (if the player misses), or that the player's character does in fact stab them (if the player hits). I mean, that's what you were trying to do, and you either succeeded or failed with your binary to hit roll followed by a damage roll based on how lethal your weapon is, right? So naturally, that's how you'll be imagining the situation. There's a constant pull away from abstraction if we enter into the situation imaginatively, as we are constantly invited to do by the rules.

The point is that these three things, the language, nuances of the rules, and the structure suggested by the sequences of declaring and then executing actions, and the way this invites us into the shared imaginative space, all work against the abstract combat round type of approach to hit points.

The New School Fix

Suppose we eschew the abstracted combat round and stick with the surface logic of striking blows in D&D. What can we do? John Bell in conversation pointed out to me that Justin Alexander interprets the strange ability of heroes to absorb so many more stabbings in a different way. Individual blows really are individual blows and damage really is damage. But your hit point total represents your general fighting capacity, and damage is indexed to that. It's true a sword is (potentially) twice as effective as a dagger against Conan, but given his honed instincts, training, and the luck he makes for himself, the best you're going to do with the first eight sword blows is scratch him. He's just that good. By contrast, for a 1st level recruit, a sword will still be twice as effective as a dagger, but a single blow from a sword might very well kill them. Here's a nice quote: 

I think this makes a kind of sense, and it's probably about as good as you're going to get while sticking with the logic of individual blows and effects suggested by a straight reading of D&D's mechanics. It's true that it has a couple of problems, like why can the cleric heal a grievous wound for the peasant, but not more than few scratches for Conan? Better fighters need bigger miracles? 

But the real issue for me is that it doesn't address my objections. This is not to say it doesn't work--it basically works, it's just if you want what I want from a game it's not going to satisfy. It leaves the endless pounding aesthetic in place, and leaves combat (starting at mid levels) in that weirdly predictably not-very-scary space. Again, I'm not saying you can't die with this kind of hp inflation, you can, but combat's got a weird "we've got this" type logic most of the time. 

By the way, I facetiously called this a New School fix, but the truth is that this is, arguably, to a certain extent how even OD&D is handling things from Supplement I forward, basically as soon as variable weapon damage gets introduced. Certainly there's a ton of this kind of thinking in AD&D 1E. So it's an especially clear statement of a rationale for something that's not very new after all.

Making it Abstract FOR REAL

My proposal is to fix the problems I mentioned so that we might take the abstract combat round more seriously. I'm going in the "old school fix" direction, although it takes me is to place where we use certain narrative practices that are not very well entrenched in retro play. I'd also like to try to make combat more a space of real injury and also stretch the peril characteristic of low levels to cover a lot more of the progression curve. I'd also like to stay away from the superheroes trading blows upon blows aesthetic. 


Let's start with language. Let's make it more abstract. Instead of calling it "to hit roll" let's call it a "combat roll". Instead of a "hit", let's call it a "success". So you "succeed" or "fail" at your "combat roll". Instead of "rolling damage", let's call it "rolling effectiveness", or "rolling effect". If you succeed at your combat roll, then you roll to see how effective that was success was. 


Here's a  bit of language that bleeds into more significant rules. Instead of calling them "hit points", let's follow Logan Knight (at one stage of his development) and call the counter ticking towards death "grit". The basic idea is that grit will track some abstract combination of honed instinct, resolve, situational upper hand, minor wounds, and the like. Think about grit as a tracker leading towards actual wounds. (If you are familiar with Apocalypse Worlds and Blades in the dark, think of it as a "clock".) When the grit tracker is depleted then actual wounds and death ensue (knives in bellies). If you haven't read Logan Knight's original post on Flesh and Grit, as well as his more evolved rules here on this I recommend you do so, as my whole approach takes inspiration these, and mechanically speaking, just ring a couple of changes on them. In fact an even better version is in Ava Islam's Errant here and a post where she unpacks her reasoning here.

To make this work, we need to distinguish mechanically bonafide wounds from the kind of thing grit tracks. We can have real recovery rules for wounds, but a different recovery mechanic (more like long rests in later editions of D&D) for grit. I'll present a system below drawing on Knight and Islam's approach. 

Keep Grit Dice Low

We'll need to keep grit low throughout the duration of a campaign if things are to stay on the knife's edge. In Jorune: Evolutions, you can only gain a Grit Die (GD) through what I used to call "Big Ticket Sandbox Advancement", and now call Signature Achievement Advancement (diegetic accomplishment of some task that makes you a badass). Since signature achievements are hard to accomplish, characters will likely remain at 1 GD for a while. Also, the system maxes out at 4 GD as the most a player can ever get. So we'll never get into superhero range, although 4 HD could arguably represent a Conan figure pretty well.

Ditching Variable Weapon Damage

Let's also decouple "effectiveness" from how big a blade a weapon has. In other words, let's ditch variable weapon damage and go back to OD&D pre-supplement 1's flat 1d6 for damage. This will help a lot with abstraction and is exactly what Jorune: Evolutions does. Variable weapon damage probably does more than anything to push us to imagining combat rolls as concrete blows that injure our foes to a greater or lesser degree. (There are other, more interesting, ways to differentiate weapons that I discuss here.) 

Sundry Rules Modifications

Perhaps we can handle the ammo question the way Gus L does (and many others) with a usage die for ammunition. This will drive us towards abstraction and away from arrow counting. With different healing rules for grit and real injury this will smooth over the "cure light wounds" difficulty. 

Describing Combat

Now for the harder bit: how to describe and imaginatively enter into combat. The most interesting change will be in how we describe and think about combat narratively by replacing "fortune at the end" with "fortune in the middle". 

Players will still be asked what they want to do at the start of the round. They still say what their character is going to be doing in combat that round, like trying to kill someone or attacking with a sword. But combat as it actually unfolds will be described differently. I think we need to bring the players into combat imaginatively in a different way to make this work by asking them the right kinds of questions as DMs. 

I'm thinking of the way even in retro gamist heavy play, DMs will often ask a player to describe how they kill someone, giving narrative control over the moment of victory to the player, asking "What does it look when you take them out?" Here I'm envisioning something similar but re-centered on an abstract understanding of grit. 

Suppose the player succeeds at a combat roll and rolls a 6 (max) on the effectiveness die. The DM might say, "Wow, that was pretty effective. Tell us how you're gaining the upper hand." Or, supposing the grit counter has ticked low for the enemy as a result of the roll, we could go with the more melodramatic: "Tell us how we know that the hour of your enemy's doom is approaching." 

This will encourage players to enter into a space where they think of the struggle of combat as leading up to grievous injury, rather than consisting of a mechanic series of blows. It may also add a bit of (welcome) flavor to what is famously kinda dry combat in OSR games. 

Now, to make this work, the rule will have to be that the player may not describe the opponent as suffering a (real or serious) wound when all we're talking about is the loss of grit, since that hasn't happened yet. Furthermore--and this is the really tricky bit--there needs to be an understanding that what is described doesn't constitute fictional positioning with mechanical benefits. There already are mechanics for combat in place.  

So, for example, the player can't say, "I split his breastplate and give him a gash from shoulder to rib." That would be a wound, so the DM might say, "Well you didn't wound him yet, so let's say you actually dent his breastplate, and you hear him groan and gasp in breath as his ribs bruise." Players will pick up on this soon enough and it will effect how they imaginatively enter the combat space, "I beat him back and the best he can do is frantically parry my blows".

A harder case: what if they say, "I strike him hard, and he tries to parry, but the blow knocks the sword across the room"? The problem is that some weapons and unarmed combat have rules for disarming people, and so this is already covered by the rules elsewhere. They can't get that condition "for free" through narrative control. You don't want players "double dipping", both gaining the benefits of an effective blow that significantly advances the tracker towards wounds and death, and also gives them further mechanical benefits. Trust me, if players have this power in a gamist space where they're required to "try to win" it's not going to work. 

As a DM, I would redirect this description in a "yes and" way, "Amazing. But he's not disarmed per the rules, so he dives for his weapon and retrieves it, sweat forming on his brow."   

Whether as DM you allow the cinematic narration to bleed into tactical advantage via fictional positioning will be delicate. Probably you can't stop it altogether, and you'll just have to use your judgment and not let it get too far. For example, take a still more subtle case. Suppose a player describes bashing against someone and driving them back towards a canyon behind them. It's a judgment call whether you redirect that in a "yes and" sorta way. 

If you allow it, it will have tactical consequences, for example, about the possibility of someone trying to grapple the foe and toss them over the edge in later rounds. But the flip side is that opponents can wriggle out, or gain some fictional positioning, when they score effective attacks too. So it's a two-way street and given the right dynamics at the table, this might be pleasing.  

What I would say is that it's a judgment call, and something to be worked out in practice. Generally speaking, my advice would be that the DM should use the abstraction and narrative flavoring to try to keep separate directly mechanical effects from narrative descriptions via gentle "yes and" redescriptions where necessary. 

This approach will work best with theater of the mind play, where tactical maps are used (if at all) only roughly to indicate where people are at in a fluid situation. This approach requires a degree of abstractness that is a poor fit with five foot squares and the like. The mantra is to make combat mechanically abstract, as a counter towards wounds and death, and then allow narrative descriptions to shape the space of our shared imagination by asking questions that lead away from the concrete narration of trading blows. 

The System 

Here's a first pass at implementing the approach I've been outlining in this post for a particular game, Jorune: Evolutions. A classless OD&D inspired sword and planet game. 


  • To attack in Jorune: Evolutions make a combat roll, which is 1d20 + Modifier (Strength for Melee, Aim for Missile Weapons) against the target's AC. 
  • If you tie or beat the AC this is a success
  • On a success, you make an effectiveness roll which is always 1d6. 
  • Subtract the effectiveness roll from the opponent's grit, which is a tracker towards wounds and death. 

Wounds and Death

Grit never falls below 0. When it reaches 0, the target makes a Stamina check, which is 2d6 + their stamina modifier. (These are like deathblows in the video game The Darkest Dungeon if that helps you, except that the blow that takes you to 0 also induces one.) Note that starting stamina modifiers range from -1 to +1 and never go higher than +2.

6- Target slain.

7-9 Target wounded.

10+ Only a scratch. 


The wounded target receives a -1 for the duration of combat to all rolls (including future stamina checks, effectiveness rolls, etc). 

Consult the effectiveness roll of the attack that wounded the target. 

(1)-(3) The target gets to describe the wound received.

(4)-(6) The attacker gets to describe the wound dealt.  

Such descriptions can have consequences for healing, and also fictional positioning, although for mechanically speaking, the total effect is -1 to physical rolls. 

Here's a chart:


All GD can be recovered with a full night's sleep. Whatever your grit score upon falling asleep, roll all your GD afresh as you would normally upon wakening.  

Wounds on the other hand heal more slowly and perilously. The penalty to rolls persists until proper healing can take place over downtime. 

During the Adventure

If a party member is wounded, reduce overland travel for the wounded by 1 hex per day for each wound they received. Having to camp in the wilderness with wounded party members is not a good situation. When the party makes camp for the night, the person with the highest medical score must check to treat their wounds:

6-    The wounded must check vs. stamina or acquire an infection
7+   No infection

Infection: Open a three step infection tracker and give it one step. On three steps the character dies. Each night a further stamina check must be made. (There are limilates and the like that can help with recovering from infection.)

6-    Add a step to the tracker
7-9  Remove a step from the tracker
10+ Remove two steps from the tracker

During Downtimes

Resolving Infection

If the party has not yet camped, when the party returns to the village, you should first test for infection by having the member of the cohort with the highest medical skill make a check as above. (The party may also recruit an NPC to treat the wound if someone has a relationship with the NPC.) Resolve the rolls for recovering from the infection all at once using the above rules with one exception. 


If the character reaches three steps on the infections tracker while at the village, the player can opt to undergo surgery. The person with the highest medical skill in the cohort then makes a medical check. 

6- Death
7+ Recovery with permanent injury

The player may decide what the permanent injury is. It must have a mechanical effect of some kind, and needs to fit the description of the wound. 

Rest and Recuperation

Once it is determined that the character will live and whether they have suffered permanent injury, the character must rest. They may not go on missions or perform downtime actions for 1 week for each wound received. If they underwent surgery, add 2 additional weeks to their recovery. During this time, the player of this character can play with their alternate character during recovery time. (Here modest troupe play makes these injury rules workable and perhaps even fun as a change of pace).