An enduring feature of RPGs since the very beginning of the hobby is the attack roll. When a fight breaks out and you need to hit somebody, there’s a special kind of roll that’s different from the other rolls you make. Swinging on a rope? Roll once. Leaping to safety from a deadly trap? Roll once.
Hitting a kobold with your sword? That’s two rolls.
Why?
Lineage
Dungeons & Dragons inherited its combat systems from Chainmail, a wargame that evolved from battlefield combat to incorporate fantastic heroes with magic items. Its granular initiative system (ranged attackers go first in initiative order, then melee attacks, and then some spells go off if they weren’t interrupted) featured a process for attacking that still looks familiar. First the attacker rolls some d6s and adds in some math to see whether they hit the target’s AC. If the attack roll does not meet or exceed the target’s AC, they miss. If the attack hits, the attacker rolls some more dice to see how much damage they did.
In Chainmail, two rolls allow a whole bunch of factors to be taken into account: the “fighting capability” of the characters, the number of attackers, magic item bonuses, and the weapons used. Accordingly, there are a lot of tables to refer to and math to resolve in order to make combat run smoothly. The intention behind this system isn’t to make things fast, but rather to increase the fidelity to the events being ““simulated.””
Attack rolls in D&D, in all its iterations, work basically the same way. It was adapted from Chainmail, and because combat is the primary feature of the game, this makes sense. Unlike THAC0.
Two-Stage Rolls for Two-Stage Games
As RPGs grew distinct from their wargaming roots, they developed two primary phases: combat and everything else.
Every iteration of D&D has this dichotomy because they’re combat-centric games with a side dish of adventuring (and a little aperitif of roleplaying). When you roll dice outside of combat, you resolve the effect based on the number on the die. During combat, it’s a two-step process. This tonal shift between the two halves of the game occurs when the GM calls for initiative, sorts out the turn order, and then combat proceeds. It breaks immersion. Players make their roll and then lean back, await instructions from the DM, peruse their character sheet, stack their dice, and perhaps talk strategy with the other players.
I‘m taking about D&D here, but this applies to LOADS of other games. You have played a game that does this. You have experienced this weird lull in action while you move from one part of the game to the other, like the Scooby-Doo gang leaving the Mystery Machine and walking to the haunted house, but with the tension removed.
And my question is why? If anything, combat should be faster than all the stuff that happens outside of combat. Yes, “roll for initiative” is dramatic, but it’s also a signal that gameplay is about to slow to a crawl. It’s a move from the narrative part of the game to the rigidly arranged grid of combat.
Let’s look at the reasons for two separate attack and damage rolls in different games.
Roll to Attack!
Making a separate roll for an attack can create a push-and-pull between attacker and defender. In GURPS the attacker rolls to hit and then the defender rolls to block, and damage is dealt if the attack is successful and the defender failed to block. The damage is, you guessed it, a third roll. FATE works the same way, except damage is more abstract.
There is tension here, but it’s “looking at dice and doing math” tension instead of narrative tension.
Another interesting permutation is Vampire: The Masquerade first edition, where characters that are brawling both roll to attack, and damage is only dealt by the victor, which is the combatant with more successes. To me this concept recalls tabletop miniature wargaming. But at its core, it’s basically the same as “attack roll vs defend roll” with greater risk. This mechanic never appeared in later editions of Vampire, so I might call it an unwelcome risk.
To go a step farther, Troika! uses an attack roll and a damage roll, but the value on the die is not the damage you’re dealing. You check the weapon chart on the inside cover of the book to determine how much damage you do. It’s clunky, but it invokes the innumerable charts and tables of the weird 90s RPGs to which it pays homage.
One Roll, Multiple Outcomes
Another approach to combat mechanics is to use the die result for multiple purposes, which increases the variability of the outcome. Innumerable games use a “critical” rule that stacks bonus damage or imposes negative conditions when an attack roll is very high, or adds a complication if it’s very low. For instance, in Vampire: The Masquerade, you roll to attack and deal extra damage dice if you get extra successes beyond the ones you need for the attack to land. In Modiphius’ Fallout 2d20 system, attacks that go over a damage threshold apply hindrances to the target based on the hit location.
Using the same approach with damage gives us exploding dice mechanics, as in Savage Worlds, Deadlands, and Shadowrun. Rolling the highest number on a die allows you to roll it again, and keep rolling if you keep rolling high! This creates a great deal of excitement around the table at the cost of being unpredictable and more difficult to balance around.
Roll to Damage!
When we look at systems that simplify their combat, we find that the attack roll (the most exciting bit) overwhelmingly persists, but damage is determined in different ways. Attacks in Numenera are treated like any other Task which makes it very natural at the table. When attacks land, they deal flat damage according to the weapon or special ability used, which is reduced by the target’s armor rating.
Removing the randomness from damage means less time spent rolling dice, which in turn makes for faster combat. Predictability enables more direct and tactical thinking when players can more easily calculate how much damage they can do. This is a good example of how combat is not its sole focus of the Cypher system because it’s built for simplicity and adventure.
Unknown Armies has a d100 roll-under system where the result of an attack roll is used to determine damage when they are successful: deadly weapons deal the result (a roll of 35 deals 35 damage), whereas weaker ones deal damage equal to the sum of the dice (a roll of 35 deals 3 + 5 = 8 damage). That’s a huge difference! Using your fists versus shooting somebody appropriately feels like night and day.
Boot Black, my action combat RPG, takes this approach. Your weapon determines which dice you roll and how many, e.g. D8s for rifles, D10s for shotguns. The number of successes you get equals the amount of damage you deal. It’s very fast when you’re in the swing of play.
Simplifying
My design sensibility is to make things simple and fast. I also want to remove the immersion-breaking rift between adventure and combat. If you want to smack a bad guy or blast them with a magic spell, I want to roll just once and get on with it.
Into The Odd eschews the attack roll and instead has players throwing a single d6 to deal damage. The result is modified by their weapon bonus and the target’s armor. If (for some strange reason) you compare this to Chainmail, there is some complexity sacrificed in the name of making it super easy and very fast.
What do you lose?
The trade-off is that simplifying mechanics is like taking arms off a coat rack—you end up with fewer places to hang coats. By “coats” I mean things like the combat modifiers and initiative quirks of Chainmail. There is an appeal to having lots of options and features to your character, but the other option is to build simple mechanics and allow all of that flavor to come from narration. I want my coat rack to have space enough for only the most useful coats. It’s cool to have a big list of abilities and character options, but even cooler to let the player create whatever they want!
Simplicity can also mean (as in the case of OSR) rules ambiguity. If you don’t have an improvised weapons table, how do you deal with a player trying to hit a troll with a chair? If there’s no place to hang your coat, you need to improvise. For what it’s worth, I would argue that edge cases have obvious resolutions in systems with thoughtfully designed core mechanics.
Designing For Speed
The “one roll” idea was front of mind as I developed Brightknife. When combat begins, I want players think “how do we solve this problem?” instead of “how can we most efficiently win by attrition?” To do this I endeavored to remove the friction between narrative play and combat to keep everyone in the narrative mindset.
I created a system where attacks simply hit unless you critically fail the roll, and the result tells you how much damage you’ll deal. In practice, this system has some concessions. First, I made the decision to make attack Rolls slightly different from ordinary Rolls: there is no “success” necessary to land an attack. A very small difference that nonetheless trips up some players at first.
In a perfect world, the number rolled for the attack would also be your damage. In Brightknife, players are rolling anywhere from 2 to 6 dice, though, so this would feel like a big disparity; it also conflicts with my other design ethos of keeping numbers low. The actual damage is usually the lowest or highest number rolled, while magic effects make you add them up and divide.
You don’t need a chart like in Troika! It’s not as simple as Into The Odd, but unlike Odd you aren’t changing dice to make an attack. The call for initiative (if it even happens) is a d6 roll that behaves like a coin flip (do the players or the enemies act first?), and then you’re into it.
Why Roll Once?
The biggest advantage the one-roll system has over the two-stage attack is speed. Further, with two rolls, you usually have two different sets of modifiers to add to the results; cutting this away means less to memorize. With a one-roll system, that complexity isn’t necessarily removed. Through clever design, you can hang that particular coat elsewhere.
Consider FFG’s Star Wars: Edge of the Galaxy. You grab those weird numberless dice (pictured below) and roll just once to attack. There’s still a push-and-pull among the dice as you interpret them, the hits and misses canceling each other out, the narrative coalescing in your mind. The mechanical richness of the two-stage attack is still there, but now it happens in two phases (roll + math) instead of four phases (roll + math, roll again + math).
Looking Forward
Combat resolution mechanics have a huge impact on the speed and “crunch” of the game. Because there is momentum away from D&D 5e and a dizzying array of other systems available, I think there’s greater moves toward innovation and less of a feeling of the two-stage attack being the “default.”
Could this trend away from WotC and D&D, and by extension its classic design features, lead to newer games having less complexity and tactical crunch? Perhaps, but not because of this. More salient in that equation are the changing desires of the TTRPG community.
Do you think that modern games have a different sensibility about combat? What do you think the future portends for the two-stage attack roll?
Simple mechanics do not imply a simple game; that’s a basic game design principle, and if you don’t believe me, look at chess. There are many, many other ways of adding nuance beyond rolling more dice.
†