Monday, October 23, 2023

Dungeon Skrimishing

TACTICAL COMBAT MECHANICS for Theater of the Mind Dungeon Crawls

Front Piece From the Holmes Edition - 1977

Running skirmish sized combat requires more than a party that can win with limited special abilities (such as a sleep spell or fireball), and must hold players interest by avoiding an endless grind of simple attack rolls. To do this it’s best to introduce some element of tactics. It’s important that, without resorting to true “grid combat”, one has rules for: spacing, ranks, and flanking. With these few concepts one can have simple shieldwall combat that provides both faster and more tactical skirmish size combat while still retaining the basic structure of the rules found in older editions of Dungeons & Dragons or other systems built from the same sources. These rules are also modular, and can be adapted to each table’s time needs, comfort with tactics, desire for combat options and interest in measurement or fine detail … to a degree of course. These rules are still early Dungeons & Dragons combat based on Arneson’s “alternate combat system” and use the same abstracted, simple rounds, initiative and attack rolls every Dungeons & Dragon player is familiar with. A significant advantage for my own games is that this set of additional rules don’t require grid-style combat -- the concepts of line, rank and spacing are largely self-contained, self-relational and intuitive to a degree that with a little practice they are easy to run from even gridless maps or a vague sketch of a random wilderness area. That is, these are “theater of the mind” combat rules.

Simplified and Fantastic Pre-modern Combat. Here is a set of mechanics and a procedure that allow some tactical complexity with the limitations of the fairly simple and abstract combat mechanics of older, exploration focused systems, without the necessity of a grid or measurements. They also badly mimic the basic ideas behind pre-modern combat, based on the shieldwall tactics of classical Mediterranean infantry armies and even more a simplified and vague cinematic interpretation of the warfare between warbands in Europe of the Migration Period and Early Middle Ages. They are undoubtedly incomplete and historically wrong in several ways … but they have worked for me to offer a comprehensible tactical system that doesn’t require a complex grid and token system.

Combat occurs between two lines of armored (and often shielded) combatants facing each other so that each front line fighter limits the number of opponents they face and can avoid being flanked
Less well armored combatants either take up positions behind the front line in ranks to attack over their shoulders with spears and polearms, or extend their side’s line in an effort to flank the enemy line.

RANGES AND DISTANCE The dungeon is almost always a cramped place and dungeon combats tend to take place at very short distances compared to field battles. There is no room for cavalry, artillery, push of pike, or even much for missile fire. Because of this the exact measurements or even the grid of a war game are less necessary and estimated “range bands” can be used if they are easier to imagine and remember. As with turn keeping vs. time keeping, remember that the characters and players are unlikely to know or care if their heavily armored foes at the other end of the hall are 42’ away or 37’ — only if they are in range to charge this round. Instead of calculating the 40’ combat movement rate it’s more efficient to consider distance in terms of simple distance categories: Close (grappling or 0’), Melee (in melee strike range or 5’), Reach (attack range for spears and polearms or 10’), Charge/Medium (Distance that can be closed with a charge attack; 10’ - 40’), and Long (beyond 40’ usually outside of torch or lantern light distance, requires a round at least to closer to Charge distance). The referee should estimate distances based on a quick glance at the map (its distance grid can help, but isn’t absolutely necessary), but for it to work the players need to trust and accept the referee’s adjudications rather than argue for advantage.

These range bands still support existing combat mechanics, such as ranged weapon bonuses, the referee just needs to describe ranges and distances in terms of the immediate concerns of the players rather than distances in feet or meters. Explain “What can attack the characters and what can the characters attack” without the intermediate issue of calculating distances. Some detail and granularity may be lost, but for most combats, especially dungeon combats, these estimations are sufficient, far quicker, free the game form grid combat, and leave less room for meta-gaming tiny distances.

MISSILE COMBAT Bows, crossbows and other long range weapons are extremely dangerous to fire into melees and are usually limited to either an initial volley or two as forces close. Thrown weapons can be modeled in a more interesting tactical manner that somewhat mirrors the use of thrown spears in Hellenistic and Roman combat, or hurlbats and francisca by Northern European warriors such as the gallowglass until the 16th century.

Firing into Melee. In an open field battle where opposing forces advance across the field from hundreds of yards, or in a siege long range missile weapons such as bows and crossbows are deadly and effective… in the close darkness of a dungeon, they are rarely useful for more than a couple of shots before melee commences.

Missile weapons can always be used normally prior to melee combat and fired from any rank, but the risk of injuring or dangerously distracting one’s allies is quite high.  When firing into a melee (even at enemies in the second or deeper ranks) a natural attack roll of five or under (modifiers don’t count), will strike the ally nearest the target (or alternatively distract them allowing their opponents to strike them) inflicting its damage on the ally.

Optional Rule: Reactive Thrown Weapons Thrown off hand weapons such as hurlbats, plumbatas, piling, or throwing knives, which can be used in reaction to and attack. Held in the off hand these thrown weapons allow a trained Fighter or Thief a ranged attack as a new enemy moves to engage them in melee.

A Reactive Attack is made just like a normal attack, but interrupts the initiative sequence, and allows the combatant with the drawn thrown weapon to attack prior to an enemy moving into melee.  This attack can only be made immediately prior to the enemy’s first attack or charge, thrown from a foot or two, it is otherwise as a normal attack. Reactive attacks are not allowed as an additional attack on the combatants own action (though the throwing weapon can be used as normal if missile combat is an option). A reactive attack with a thrown weapon does not provide time for the combatant to draw an additional weapon, pick up a shield, brace against a charge or otherwise perform any additional acts prior to the opponent's action.

Lines are one or more combatants armed with a melee weapon who controls an area and prevents up to two enemies directly in front of them from passing them.  While it’s possible to break or flank a line, an enemy armed with a regular melee weapon cannot pass it or attack anyone except for the 2-4 enemies directly in front of them in the enemy line. In a dungeon skirmish lines are often anchored by a wall or other obstacle, and so become impossible to circumvent (or flank). Each human sized combatant takes up and can protect 5’ of space (or half a map square) (See Fig 1.), unless they are in a doorway, in which case they can cover up to 10’ of space (See Fig 2. This means that two defenders are required to form a line across most corridors.

A combatant in a line formation can attack enemies in the 15’  in front of them, including the right or leftmost enemy in the next line segment. This means that the most enemies a combatant in a line will face directly is three (Fig.1) ... well four if they are in a doorway.

Defenders A & B make a narrow line, attacked by 1-5
is in melee with 1 & 2.
B is in melee with 1, 2 & 3
has been FLANKED by 5
is not in melee combat

For those defending doorways this is significant, the lone defender is able to prevent the enemy from flanking, but still faces multiple (up to 4 plus any from the enemy's rear rank) attacks each round. For larger groups attempting to block an advance it is always better to defend behind the doorway allowing and attack single (See Fig. 2)

While these rules describe specific distances and can be used easily on a typical map of 10’ squares they don’t need to be. The ideas of line (along with rank, flank and the other rules here) are general enough that a referee description can replace a grided map. A single combatant can form a line in a doorway, two can form one in a corridor, while a large room might take several, but exact distances aren’t necessary. 

The decisions about line length can be the referee’s alone, as long as they are consistent, and clear to the players before the characters form up.  If a room is too wide for the party’s melee combatants to form a line, then the referee needs to make this clear before they attempt it -- it’s something that would be obvious to anyone with any knowledge of melee combat after all. Yet, while it might take an unmanageable eight fighters to make a shield wall in a huge gallery (or far more in an open field) this problem is one for the players to solve with retreat, another formation such as a square, or spells and items (walls of fire or a pool of burning oil will protect most flanks) -- it’s not the referee’s job to offer simple situations, especially to mid-level parties.

Lines protect characters (or monsters) that are less capable in melee combat, but become tactically meaningful with bonuses for flanking and can become sufficiently complex to allow some interesting decisions around rank, missile use and of course sorcery. 

Optional Rule: Stances. Sometimes combatants may need to focus more on defense or attack based on their tactical situation and by adopting an offensive or defensive can boost either at the cost of the other. This is especially effective at holding lines or breaking them and can make a second rank far more effective. 

Each round a combatant may decide to shift their stance between the following three options:

Defensive: AC +2/No Attacks.
Balanced: No change.
Offensive: AC -4/+2 to hit and damage.

Charging and Bracing. Combatants in “Charge range” - at least one square or 10’ from the target can charge, gaining an advantage to attack. Defenders armed with the right weapons can brace themselves against charges and gain the chance to inflict devastating damage prior to the charge striking. 

A charge can be launched from up to the maximum movement of the attacker (or alternatively by an attacker in charge range with very fast creatures having the ability to charge from long range) on their initiative. Charging grants a +2 to hit and double damage and provides a two point AC penalty.

Instead of attacking, a combatant may brace themselves on their initiative and wait for attack. A braced defender gives up their initiative in exchange for a defensive or responsive attack.  Bracing grants an attack following the first charge attack against the braced defender that inflicts double damage if it hits successfully.  Unless the braced defender is armed with a spear or polearm.  With a spear or polearm bracing becomes more effective, providing the defender with a reactive attack BEFORE the charge hits them and allowing such a reaction to all opponents that charge them or anyone next to them in their defensive line. It is not a good idea to charge a line of braced spearmen.

Flanking is the threat that creates the necessity of lines… a combatant is flanked whenever they are being attacked from more than one direction (from the rear or side). A shorthand for this can be when they are in melee with more than three enemies (See Fig 1.) Flank attackers can of course bypass the defender into the rear of the line to attack those they are defending in normal melee … or attack the flanked defender with a +4 to hit for double damage (as per a Thief’s Back Stab). This should quickly crumble flanked lines.

Optional Rule: Champions. Fighters over level 4 are traditionally heroes… they can parry and dodge any attack, spinning and rotating, leaping and shiting lines of defense and attack with nearly supernatural skill. This is a powerful ability of fighters who can be at the center of a large battle without penalty and better live up to their class name.

For each level above 3rd a fighter can confront an additional attacker without penalty (4 at 4th, 5 at 5th … until at 8th level they can defend from every angle at once). This ability also applies to large, powerful and unnatural enemies over 4HD such as ogres, dragons, manitichora, and bear-owls -- all of whom are immune to flanking. A considerable weakness of normal animals, even those with high hit dice such as bears and elephants, is that this rule doesn’t apply to allowing low level hunters to overcome them with numbers! Though the referee should consider if larger creatures take up more “frontage” in a battle line, and so can defend against a larger number of attackers.


Ranks are additional lines of combatants behind the first. While only those in the front rank can engage in direct melee combat, those in the second rank (also only two per line segment) can only attack the enemy front rank it they have spears, polearms or similar reach weapons, and are immune to melee attack from the enemy front rank unless they are armed with the same. (Fig. 2).  Ranks beyond the second cannot act in combat except to cast spells or use ranged weapons (see Firing into melee rule), but they are available to step into holes in the forward ranks if an ally falls.

A defends a doorway from enemies 1-5
is in melee with 1, 2, 3 & 4
5 can attack A from the second rank with a reach weapon

When someone in a line is struck down, their opponent may elect to step into their space in the enemy line. This threatens the enemy line with the danger of breaking and the defenders on either side of the break risk being flanked and likely overwhelmed. A second rank prevents flanking as the attacker pushing into the gap made in the front rank is in melee with the second, where there is no second rank, the first rank is exposed to flanking. (See Fig 3).

A, B & C fight 1-4
prevents 1 from FLANKING B
is FLANKED by 4 who has stepped out of their line to flank.

Breaking Lines.
It’s also possible to break lines without killing defenders in the front line, by pushing through them with strength and aggression: leaping, pushing or trampling those in the line to displace and disorder them. This line breaking attack is made with a regular attack roll at -2 for each rank in the enemy formation. Worse the line breaker must go last that round with anyone attacking them gaining a +2 to attack. If the line breaker succeeds they not only damage their target, but have pushed the entire enemy file back one rank. The defenders can try to dislodge them the next round, pushing them back into their own lines with a similar line breaking attack.

This situation can lead to a push and pull of combat, and encourages combatants in the rear ranks to add their numbers to the file defending against enemy efforts to break their line or where their own forces are trying to push through the enemy line.

Rank Swapping and Retreat.
Another advantage of fighting in ranks is that injured front line combatants can retreat back through their lines and fresh ones from the rear ranks can take their place. On their initiative a combatant can displace the combatant in a file in front or behind them with a successful 3D6 roll under DEX.  If the roll fails any enemies in melee with either combatant get a free reactive (or opportunity) attack against them.

When breaking contact and fleeing without the benefit of a rear rank to retreat past the retreating combatant is always subject to a reactive attack and must roll 3D6 vs. DEX to make their escape at all. Optional Rule: Squad Initiative. A change to the standard D6 based side based initiative that I have found helpful for larger parties and combats, especially with online play where it can be rolled or placed in chat and easily ordered at the start of each turn is "Squad Based Initiative." This system is a variation of individual initiative, with each player rolling initiative at the start of each round. This also means that one can more easily add initiative modifiers to individual characters or monsters based on conditions or abilities stats (I give a +1 bonus to INT 15 and over and -1 to INT 5 and under.) The referee of course roll initiative for their enemies. The difference that makes this suitable for larger combats with dozens on each side is that rather then rolling for each character, NPC follower and foe these rolls count for multiple creatures. Each player rolls a single roll for their character and any henchmen or followers that they lead -- a squad. Usually by mid-level this will be a group of 4-10. The referee rolls initiative for similarly sized groups of foes, each "squad" of 5-10 regular combatants or groups of weaker foes each assigned to led by any more powerful leader types. Action takes place by squad, players directing "their" henchmen at the same time as their character. I find this system quick and efficient as it reduces planning and discussion among the party each round about what order they act in, curtails ill feeling around players that always want to act first or declare the party's strategy and sets an order of action where each player knows exactly when they need to contribute and which the referee can enforce by calling on whoever is first to act.

Dungeons & Dragons was likely originally intended as a supplement to larger scale wargames, with dungeon crawls existing only as a sort of “special mission”. I take this reading from both the use original intent to use Chainmail (mechanically a fairly typical war game of the 1970’s) as its combat system, and reading through “The First Fantasy Campaign”, which, whatever its flaws or additions, captures the spirit and structure of Dave Arneson Blackmoor.  Blackmoor of course chronicled both the exploration of Blackmoor Castle, and the larger conflict between various forces on the scale of a larger military campaign.  Descriptions by Arneson, Gygax and others also suggest that even after the full scale fantasy wargame campaign structure was set aside (or became the last “tier” of play - often called domain management) that the party size for early RPG campaigns was considerable, with 8-10 characters accompanied by many henchmen, hired guards or even orcs overawed into service. With parties of 10 - 50 (largely consisting of semi-disposable henchmen) the encounter sizes of hundreds of orcs or bandits listed in OD&D make far more sense. This alone isn’t a reason to use more complex skirmish based combat, but provides context for certain rule decisions that are at the core of all Dungeons & Dragons editions, and also points towards a type of scenario that is not widely used in newer adventure design.

While it was common in the OSR period (approximately 2006 - 2020) to discuss “tiers” of play - starting with dungeon exploration at the early levels, wilderness travel in the mid-levels and domain play at higher levels, like many OSR ideas there’s an element of desire and idealization in this idea, or at least simplification.  As much as the Basic and Expert Box sets of Moldvay and Cook  suggest two tiers with their sample adventures and content, and the later BECMI adds to these with domain management, world politics/crisis and finally an immortal tier, these labels aren’t absolute. Even in adventures leading to immortality such as level 30 to 35 “M3 Twilight Calling” there are dungeon crawling interludes, and this is even without considering such classics as “Queen of the Demon Web Pits” (Q1), the D series by Gygax (“Descent into the Depths of the Earth” etc) or “Tomb of Horrors” (S1). In the other direction there is also a tradition of low level content about wilderness travel (such as B8 “Journey to the Rock”), or even siege and skirmish combat (B12 Queen’s harvest or early parts of B10 “Nights Dark Terror”).  While these elements are often underrealized, or trivialize the hard earned spells and abilities of higher level characters with arbitrary tools (especially if Gygax was involved, or more fairly they are tournament modules).  Sure B2 Keep on the Borderlands and X1 Isle of Dread offer fairly contained “tiered” experiences, but as excellent as both are they are part of a larger continuum of design that is far less strict and limited.

While there certainly are distinct types of scenarios there’s little reason to think that one can’t run a wilderness adventure, or even a domain based scenario for love level characters and likewise no reason to think dungeons are inappropriate for higher level characters.  Incongruities and mechanical weakness in Dungeons & Dragons and its progeny are often the result of the unintended consequences of its creation and evolution, rather than the intention of its creators. One can combine and mix dungeons, wilderness journeys, domain management and leading troops in skirmish scale (still under 100 combatants).

Yet the idea of tiered stages for classic Dungeons & Dragons campaigns has a great deal of traction … because it’s not a bad idea.  Tiers present a way of looking at and at least partially resolving the most significant issue of post Greyhawk Dungeons & Dragons ... a steep power curve that makes combat ever more of a slog or massacre and trivializes many low level exploration challenges.

After around 5th level Dungeons & Dragons combat as it’s usually played becomes increasingly focused on burning through combat spells, making or failing saving throws, extremely high AC to avoid high damage monster attacks, and sometimes an endless slogs chipping away at huge HP pools or enemy hordes.  Eventually combat challenges all being to either end the first round because the party burns its resources or drag on for the entire session because the opposition has to be numerous and powerful enough to overcome that initial burst of destruction. In both cases the characters are likely to want or need to rest and recover after any significant combat in a way that they won’t at low level (low level combat tends to be so “swingy” that characters are either killed, flee, or succeed with relatively few casualties and there’s usually plenty of time left in a play session after it.)

Worse, exploration challenges start to fade as many of the tensions of low level dungeon crawling are easily mitigated by level 4 or five. Magic of course is the prime example of this with spells such as “knock” make opening sealed doors trivial, or “detect traps” turning them into a nuisance.  Utility spells like “rope trick”, “create food and water” and “continual light” entirely remove the issue of supply or even returning to the safety of a haven after each season. Likewise magic items originally designed to mitigate other dungeon challenges, such as “bags of holding” or even just the ability to hire a large number of henchmen, mean that weighty treasure no longer offers a challenge either.

There are orthodox and well known ways to make mid-level combat and dungeons more interesting. These methods are often employed in classic tournament modules and the others mentioned above, even without heavy-handed limitations on spell casting or turning ability.  The first is to limit the number of encounters to a few set pieces with environmental elements or to trick monsters who can be treated as a puzzle themselves.  Likewise mid level dungeons largely dispense with navigation and exploration in favor of faction intrigue and puzzles, which can be better designed with player abilities and spells in mind.

However, there’s another option -- remembering and reflecting on the original intentions of Dungeons & Dragons as a skirmish game, and not return to them (an impossibility) but inventing rules like those above that allow greater tactical decision making and also simplify larger scale combats to make them faster (reducing the number of actors each round, and offering combat mechanics that allow for a quicker resolution of larger battles once one side has been outmaneuvered or their formation is broken.) Skirmish combat and tactical complexity also offers a sort of intermediate set of concerns for mid-level in dungeon crawling, management of a larger party that includes a considerable number of low level (or 0-level) retainers.  It is not an absolute or the only solution to the difficulties, but another tool to make mid-level adventure design more interesting.

I’ve used a variation of these rules for many years to run house-ruled OD&D for several years. They aren’t the only rules available of course and likely aren’t to every taste, but they allow one to run somewhat tactical combat with reasonable speed, using only “theater of the mind”.  A typical dungeon map helps, especially for complex situations, but no battle maps are necessary.  It also manages fairly large battles with rules that attach easily to the existing rules for OD&D, Basic (B/X) or even I suspect AD&D.  They are one set of options.


  1. This is interesting stuff. I will need to mull it over and come back to it.

  2. Firstly, these rules are great. I was wondering about some kind of "range bands" mechanics for theater of mind combat, especially to use on online play. Because of scheduling, mostly of the games I have run are online sessions and I am not tech versed enouhg to run all the prep necessary to use VTTs with easyness, so instead I use some "baroque techniques" like maps with simple layouts, with a "fog of war" added via GIMP. I will test these skirmish rules with my set of house rules.
    Secondly, I know that are are enough "fantasy heartbreakers" out in the wild, but your interpretations of the OD&D/Classic rules are very inspiring I'd like to see the once announced CRAWL, published. The subclasses from your old blog are one of the best ways to add diversify to the character classes without mechanical complexity, as is your version of cleric's magic and the flavorful divine mishaps; so it would be a wonder to see the full extent of your rules/procedures.

  3. Thanks - I use a similar method for online games - screen sharing a drawing program with a fog of war on it that I erase as the party explores. These rules have been testing with this system pretty extensively over the years, and while I am perhaps not as rigorous with flanking as I should be at times, they've generally worked quite well.

    CRAWL is unlikely to ever be shared, I'm generally opposed to making more retro-clones. Instead, like these rules, and in the spirit of MOSIAC strict (though without the strictness) I am trying to put some of CRAWL's ideas into blog posts and adventures. These for example may go into a lower mid level adventure I'm working on (Levels 3-6) and currently calling "The Glass Forest". I may of course hit an obstacle and set it aside. Plus I need to do the layout for Bones of Bronze before I really get into it, but it should be ready for play testing soonish as I am writing it up and mapping it while I await my editor.


Old Games

Let’s talk about old tabletop roleplaying games - specifically the kind of games played in the 1980’s and recently depicted in the nostalgia...