Searching The Dungeon

Many OSRIC adventures will involve the characters exploring some enclosed area, be it the dungeons beneath a castle or temple, a system of natural caves and caverns, an abandoned mine, the sewers beneath a city, an enemy fortress, a wizard's tower, or a shrine to some dark god. For convenience, and by tradition, these enclosed locations are collectively called "dungeons", which thus refers not just to a set of man-made tunnels but to any indoor adventure location. Indoor/dungeon adventures tend to have similar characteristics and thus the same set of procedures and suggested resolution mechanics applies to most "dungeon" adventures.

Order of Play:

While exploring in a dungeon, each turn is resolved separately. Experienced GMs sometimes will allow the turns to run together, but this is only recommended after the GM is comfortable with the basic procedures. The order of events in a game turn is as follows:

  • Wandering Monster Check: Every third turn the GM rolls to see if any wandering monsters are encountered (typically 1 in 6—consult specific level key for non-standard frequency of check or likelihood of encounter)
  • Statement of Action: Party caller or individual player describes the activity of the various party members, which are resolved as appropriate by the GM:
  • Move: Up to full move rate per turn for cautious movement (including mapping); 5x normal rate when passing through familiar areas (no mapping allowed)
  • Listen for noise: 1 round per attempt, 10% standard chance for success (adjusted for class (thieves, assassins) and race (elves, gnomes, halflings, half-orcs), only 3 attempts allowed per situation (e.g. door)
  • Open a stuck or locked door: 1 round per attempt, 2 in 6 standard chance for success (adjusted for strength) for stuck door, locked door requires key, //knock// spell, exceptional strength, lock picking, or breaking down door, unlimited retries allowed but no surprise possible after failed attempt
  • Search for traps: 1-4 rounds per attempt (covering one object or location), chance of success determined by race (dwarf or gnome), class (thief or assassin) or free-form verbal negotiation (at GM's discretion)
  • **Casually examine (and map) a room or area: **1 turn per 20-ft×20-ft room or area
  • Thoroughly examine and search for secret doors: 1 turn per 10×10-ft area, 1 in 6 standard chance for success (2 in 6 for elves and half-elves)
  • Cast a spell: See specific spell descriptions in Chapter 2 for casting times and effects.
  • Rest: Typically 1 turn in every 6, plus 1 turn after every combat, must be spent resting (i.e. no movement or any other strenuous activity)
  • Other activities: Duration of attempt and likelihood of success determined and resolved on adhoc basis by GM
  • Encounters: If an encounter (either with a wandering monster or a planned encounter) occurs, the GM determines surprise, distance, reactions, and resolves the encounter normally (through negotiation, evasion, or combat)
  • Book-keeping: The GM records that a turn has elapsed and deducts any resources that the party has used (lost hit points, spell durations expiring, torches burning out, and so on).

Since each turn represents ten minutes of time, characters may combine several actions in the same turn if each is reasonably brief. Thus a character might draw a sword, move up to a door and attempt to open it all in the same turn, for example. Longer actions may take several turns to resolve (such as making a minute search of a 500 square-ft wall) and sensible parties will take steps to guard a character engaged in such activity from unexpected attack.

The guiding principle behind the exploration rules is to maximise the number of meaningful decisions the players take about their actions, and minimise the number of dice rolls between each decision.

Wandering Monsters:

Typically, wandering monsters are checked for every third turn and encountered 1 chance in 6. If a wandering monster does appear, determine the creature involved randomly unless some factor makes it obvious what the party has met.

Some dungeon levels have special provisions for wandering monsters affecting the frequency of checks, the chance of an encounter, or both. For example, in the first level of the Pod Caverns of the Sinister Shroom, the chance of meeting a wandering monster is only 1 in 10.

Wandering monsters in dungeons should be appropriate to the environment both in type of creatures encountered and encounter difficulty. Traditionally dungeons are organised such that the deeper the dungeon level, the more numerous and deadly the creatures encountered—so a group of first level characters exploring the first level of a dungeon should tend to encounter first level monsters, with maybe the occasional second level one, whereas on the sixth dungeon level, characters might expect to meet third or fourth level monsters. This pattern varies from dungeon to dungeon, however. For example, in the //Red Mausoleum//, an adventure designed for characters level 12 and higher, most monsters are extremely powerful even on the very first level!

At the GM's option, wandering monster checks may be made less often or even skipped entirely, but before doing so it is important to think about the reasons behind the rules, and particularly what the wandering monster check should accomplish. The first purpose served by wandering monster checks is to create an impression of complexity in a "living dungeon" environment without GM needing to create activity schedules or account for every creature in the dungeon at every moment. Monsters in the dungeon will have various reasons for leaving their lairs: some may be on patrol, others looking for food, and still others exploring the dungeon just like the player characters; all of the above and more are represented by the wandering monster die.

The other purpose of wandering monsters has nothing to do with verisimilitude and is purely a rules construct, but an important one: wandering monsters discourage players from wasting time. If there is no chance of meeting a wandering monster, there is no incentive for the players to keep the game moving no reason why they shouldn't hold long conversations about their course of action and methodically check every inch of floor, walls, and ceiling for traps and hidden treasure. Many players, especially those accustomed to computer games that have no in game time limits, will tend towards a "pixel-hunting" approach to play. It is up to the GM, by means of wandering monsters, to discourage this kind of slow play and keep the game moving—otherwise the game will become mired in dull minutiae and nobody will have much fun.

The same principles also apply in reverse, though. The GM should adjust the chance of meeting a wandering monster according to the players' approach. If the party is stealthy, swift, and silent, avoids heavily-trafficked areas and does not stay long in any one place, they should encounter few wandering monsters.

The Role of the Party Caller (Optional):

If there are many players in the party, some groups like to designate one player as "Caller," or party spokesman, and filter communication through that single player. This role should not default to a "party leader" who gives the other players orders and reduces them to spectators! Rather, some groups may find that by having a single player speak for the whole group, potential chaos of each player competing for the GM's attention is reduced and the game should run more smoothly, improving the play-experience for all involved. Therefore, if a caller is used, he or she should consult with the other players and then report the party's actions quickly and accurately to the GM.

The caller can be anyone and need not be limited to, for instance, the character with the highest charisma or social standing. In fact, we suggest that if the party uses a caller, the role should rotate among the players from session to session, giving each a turn.

Movement during dungeon exploration:

Is at the rates listed at the beginning of Chapter 3. This slow, cautious move rate (which works out at a tenth the speed characters move in combat) allows the characters to make a map of their progress, if they wish. When passing through familiar areas or following a map, characters can move at up to five times the normal per-turn move rate (so that a character with a normal move rate of 60-ft could move up to 300-ft per turn if passing through known territory).

Characters fleeing from an encounter may run at ten times their normal per-turn move rate (i.e. at full combat speed). No mapping is possible while fleeing in this manner and a double rest period (see below) is necessary at the end of the pursuit. Since parties will typically want to remain together, movement speed will necessarily be limited to that of the slowest character in the party.

The players should establish, and the party caller inform the GM of, the party's "marching order", i.e. which characters are in front, the middle, and bringing up the rear. In a standard 10-ft wide dungeon corridor, up to three characters may walk abreast, though if any are wielding large weapons such as flails or two-handed swords, this may be reduced to two characters or even one. Characters in the second rank may only attack with a long weapon, such as a spear or pole arm, or if they are firing missiles over the head of a shorter character such as a gnome or halfling.

There are various ways of keeping track of marching order. If miniature figures are in use, they can be placed on some board to indicate where each character is. If miniatures are not in play, most GMs will ask the party to show their marching order on paper. Sensible parties tend to hand the GM a default marching order upon entering the dungeon, and may have standard positions and procedures for other common circumstances as well. A organised group might say to the GM, "this is our formation when opening a door," or "in 20 foot wide passages we move like this," and so forth.

If the party's position is for some reason unclear to the GM, he or she is well within his or her rights to determine who is where by means of a die roll.


May be performed in most places, often at doors before opening them. Unless the entire party is still and quiet (no chattering or clanging around), and unless headgear such as helms are doffed, the listener will not detect any noise save the very loudest.

Thieves and assassins have an enhanced chance to hear noise (see "Thief Skills" in Chapter 1). Characters of all other classes have a base 10% chance. This should be modified by race; elves, gnomes, halflings, and half-orcs have a base 15% chance.

Normally the GM rolls this die in secret, because the player has no way of knowing whether no noise was heard because of the roll or because there was no noise to hear. A character who fails (or thinks he or she has failed) to hear noise may try again, each attempt taking one round. However, no more than three attempts may be made the same character before the strain becomes too great and no further listening attempts will succeed until the character has rested for at least one turn.

If the check is successful, the GM should decide whether there is in fact any noise to be heard. Some monsters, such as bugbears, are stealthy and cannot be detected by listening. However, generally if there is some monster in the area and a "hear noise" check is passed, the party should gain some clue about what it is. Clever players whose characters speak various monster languages may gain valuable information from overhearing snatches of conversation—but the GM should be careful only to describe what characters can actually hear. So the GM would not normally say "you hear a giant spider," but rather "you hear a scuttling, rattling sound" as the creature climbs to a suitable spot from which to ambush the party.

Don't forget, monsters can hear the party in the same way as the party can hear them!

Listening for noise as often as possible, at every door and intersection, is an understandably common tactic, because it's one of the easiest ways for players to improve the odds in their favour—so as to be able to make better-informed decisions about their actions. This is fine in moderation. However, if the pace of play slows considerably, diminishing the excitement and reducing the adventure to dice-rolling, the GM should discourage the players from endless listening attempts. Emphasise the inconvenience of donning and doffing helmets and headgear while the rest of the party stands around doing nothing; and if play is still slow, employ tricks that circumvent listening, e.g. silent monsters or phantom noises (perhaps due to strange acoustics in the dungeon or magic). In extreme cases the GM can place traps and monsters that specifically target listening characters, but before it gets to that, the GM should speak frankly to the players and explain that while some degree of caution is good play, carrying things to extremes only makes the game less fun.

Balance this against the lethality of the dungeon. In extremely dangerous areas, the players should not be punished for taking due care.

Opening doors:

Is not normally difficult; the player (or party caller) states the action and the door is opened. However, in some dungeons many doors are stuck and must be forced open. Doors may be locked, braced, jammed, spiked shut or otherwise held fast (by means of a wizard lock spell, for example). Stuck doors may be forced by brute strength (see the strength ability in Chapter 1 for chances of success). Locked doors will need a key, a thief or assassin to pick the lock, or some may be broken down with axes or battering rams. When designing the dungeon, the GM should note which doors are normal, stuck, locked, etc. as well as the locations of any keys.

Attempting to force a stuck door takes one round per attempt and, depending on the size of the door, more than one character may try at once. Thus, two characters could simultaneously try to force a 6-ft wide door—each character makes a check and success by either indicates the door opens. If the first attempt fails, additional tries may be made at no penalty except for time and noise. Attempting to force a stuck door, and particularly multiple attempts on the same door, is noisy and may increase the odds of meeting a wandering monster. In any event, a failed attempt to open a stuck door will prevent surprise on any creature on the other side of the door.

Lock-picking attempts by thieves and assassins are handled in Chapter 1 and take between 1 round and 1 turn per attempt(depending on the complexity of the lock). 1-4 rounds are typical.

Chopping down a door with axes or by other means is time-consuming and noisy. It takes a full turn at least to chop down a standard-size door, during which time several wandering monster checks should be made. Naturally, the party will have no chance of surprising any creature on the other side.

Furthermore, once a door is opened, it is usually difficult to keep it open, or for that matter to keep it closed. OSRIC has a double-standard that while adventurers may have a hard time opening doors in dungeons, monsters have no such trouble and can open doors automatically unless the players prevent them. The usual way to hold a dungeon-door open or closed is to wedge it with iron spikes. Even then there is a small chance (at the GM's discretion but often around 20-30%) that a spiked door will slip.


A key element of dungeon exploration; but it is one of the most controversial and misunderstood aspects of the game. If not handled carefully, mapping has huge potential to slow down the game and mire it in frustration.

When designing the dungeon, the GM should map it out on sheets of graph paper, showing the rooms, chambers, corridors, stairways, doors, traps, and other features in relation to one another. As the party moves through the dungeon the GM describes to them what they see and, assuming they have light and proper equipment and are not moving too quickly, the players may choose to draw a map of their own based on the GM's descriptions.

It is important to understand the purpose of the players' map. The goal is not to create an exact copy of the GM's map, but to keep a record of which areas are explored and which not, to allow the party to find their way back to the entrance and, on subsequent expeditions, find their way back to where they left off. If the dungeon is small or simple in layout the players may not need a map. Even if the dungeon is larger or more complex, a "trailing map" with lines for corridors and squares for rooms and chambers, maybe with marginal markings showing length or size, is almost always enough. Only in the most labyrinthine of dungeon levels, with rooms and corridors tightly packed together, are players likely to find making a strictly accurate map rewarding.

On such levels an accurate map can help the players deduce the locations of secret rooms, show them when they're circling back into areas they've already explored by a different route, or even alert them to some trick—a teleporter, shifting room or wall, sloping passage, or the like. Parties keeping a trailing map, or no map at all, may miss hidden treasures or not realise they have gone astray until hopelessly lost, but careful mapping might quickly reveal something is amiss, allowing the party to backtrack and correct their course or search for a solution. These areas are the most difficult to map, but also the most rewarding and fun, since mapping this sort of level can lead to tangible positive results.

Many players hate mapping, considering it a fun-killing burden, and these players will often try to get the GM to design simpler dungeons or even to draw the map for them. The OSRIC GM should avoid these "solutions"; play goes quicker if a player maps. Encourage the players to map appropriately—i.e. only when necessary and use a trailing map where possible.

The GM should make mapping easier by giving effective verbal descriptions: quick, accurate, and reporting only what the party actually sees. Visualise the dungeon in your mind. Describe things in distances rather than squares.

The players may show you their map and ask if it is correct. Comply only if there is a major error that would be obvious to someone in the dungeon (such as a triangular-shaped room where the party entered via the apex but drew their map as if they'd entered from the base) or if your description was faulty—and in the latter case try to make your descriptions more accurate in future.

In a particularly complicated setup—a room with lots of odd angles, for instance—a quick GM-drawn sketch may be helpful. Do this rarely, and never directly on the players' map.

The players' map represents an actual in-game object. If the players at the table are making a map, then a character must also be making one. This has several corollaries: the party must have light (they can only map what they see) and mapping supplies (something to write with and something to write on), they must be moving slowly and methodically (no more than standard exploration speed), and measuring the size of a room takes time (1 turn per 20-ft×20-ft area is suggested). Perhaps most importantly, if something happens to the map in-game, it happens to the players' map as well! If the mapping character dies and his or her body is left behind, if the characters are captured and stripped of their equipment, or if a jet of acid or a green slime destroys the map, the GM should confiscate it. If the party wants backup copies, the players must actually draw them. If the entire party dies in the dungeon, the only way their maps will survive is if copies were left on the surface.

Clever GMs will see adventure-creating potential here. Maps are a valuable asset for NPCs as well as PCs; map-buying, selling and trading could be rife, and maps found in treasure hoards potentially more valuable than gold.

Searching for hidden treasure, traps, secret doors, and whatnot:

A common activity. Looking for secret doors is a time-consuming process, taking a full turn for each 10-ft×10-ft area searched. Even so the chance of success is small: 1 in 6 for most characters, with elves and half-elves having an innate advantage (translating to a 2 in 6 chance). Searching for traps is best done by dwarfs, gnomes, thieves, or assassins—chances for success are as described in Chapter 1. A search for traps generally takes 1-4 rounds, but it is also limited to a specific object or small (no more than 5-ft×5-ft) location specified by the player: "I search for traps on the door", "I search for traps on the treasure chest", "I search the area directly in front of the throne for traps," etc.

The GM may allow "negotiation-based" searching for secret doors or traps, in which, through careful questioning and described actions, the players may achieve a bonus, or even an automatic success, on a search. For instance, players may tap along a section of wall listening for the echo of a hollow space. If such a space is discovered, the players may describe their attempts to find and trigger the secret door they know is there—perhaps looking for loose or ill-fitting stones, suspicious indentations or cracks, wall-sconces that may turn or pivot, etc. The same approach can work for traps as well.

The GM must adjudicate these negotiated searches. Perhaps they have no effect and the die roll alone decides success or failure—which certainly helps keep the game moving, but may strip away too much of the players' ability to immerse themselves in the situation. Perhaps a careful description can give a bonus to the standard check, or perhaps the description might trump the die-roll entirely—if the player is able to describe a search in such a manner that the GM feels would definitely find the objective. The downside to this is if the player's description is off-base (searching in the wrong place, via the wrong means, etc.) the GM might actually reduce the chance of success.

These detailed, negotiated searches generally take a long time in-play (more than the standard times listed above) and may increase the odds of encountering a wandering monster. This is, of course, deliberate; without some incentive to keep things moving players might tend to conduct the most thorough searches possible, describing every inch of every room in minute detail, and dragging the game to a grinding halt.

Disarming traps is normally a job for a thief or assassin (with chances of success as shown in Chapter 1) and takes 1-4 rounds per attempt for a simple trap. A complex trap may take a full turn to disarm. Other characters usually have little to no chance of success, though again careful questioning and attention to detail may create exceptions. For instance, a player may be able to surmise that wedging a pressure plate to prevent it from depressing, or stopping a vent with beeswax to prevent gas from issuing from it, could circumvent a trap.

These sorts of "negotiation-based" solutions to traps are wholly at the GM's discretion. Some GMs encourage and reward this sort of play, but others will discourage it, perhaps feeling this slows down the game too much, or circumvents the intended role of the thief class. It is important that the players and the GM discuss this issue to make sure everyone's expectations align—that the players aren't expecting purely roll—based resolution of traps when the GM is expecting them to play out each attempt, or vice versa.

Traps neither avoided nor disarmed will normally trigger 50% of the time. When designing the dungeon, the GM should define each trap by its nature and effect (see the preceding pages on "Traps"). In areas designed for first level characters, damage should not normally exceed 1d6 or at worst 1d10 and "instant death" effects should be avoided. Lower dungeon levels, on the other hand, are designed for experienced players with high-level characters who should have many ways of dealing with traps, as well as more hit points and better saving throws, so more dangerous and deadly traps may be in order. Even so, the GM should typically allow some kind of saving throw or other way of mitigating the trap's effects.

Some OSRIC groups enjoy even more lethal traps—such as those that cause death with no save. Placing these is a matter for the GM's judgement. Do you wish to encourage the players to raise zombies or call forth unseen servants or summoned monsters and send them ahead? Very lethal traps will probably lead to such behaviour, and in some groups there is a place for this kind of play. Others prefer to avoid it.

Casting spells is detailed in Chapter 2. Many spells, particularly the various detection and divination-type spells, will make the job of exploration easier. It is up to the players to decide the ideal balance between these "utility" spells and those oriented towards combat or healing. There are circumstances in which a timed //knock// or //locate object// spell may prove just as crucially life-saving as yet another Sleep or Cure Light Wounds.

Rest periods are typically necessary one turn out of every six, one turn after each combat, and double-length (two turns) after an evasion or pursuit. Parties that stay in the dungeon for several hours and are not able or willing to return to the surface may spend an entire "night" holed up within the dungeon to recover spells. During these periods the party cannot move, nor may they perform any other strenuous action (though passive activities such as mapping should be allowed). Players should be aware of when these rest periods are coming up, and make sure their surroundings are as inconspicuous, or at least defensible, as possible. A small, out-of-the-way room with a single door that can be spiked shut could be a good location to rest in for a single turn or an entire night; in the middle of an open corridor or near a stairway to a lower level is likely a bad place for even a short rest period, and often a suicidal place to spend several hours.

Occasionally, by accident or design, characters will not take these required rest periods and attempt to press on regardless. If this happens, everyone in the party is fatigued. What this means, and what sort of impact it has on the characters, is left to the GM's discretion but likely consequences are a reduced movement rate, penalties in combat, temporarily reduced ability scores, and morale reductions for any NPCs who are accompanying the party. The longer the party goes without resting, the worse these effects become.

Other actions are defined in several of the race and class descriptions in Chapter 1. For instance, a dwarf can attempt to determine depth underground, a gnome can try to determine direction, a paladin can detect evil, a ranger can attempt to follow a set of tracks, and so on. Unless otherwise specified, these actions take one round per attempt.

Beyond these sorts of pre-defined activities OSRIC has no specific system for resolving most other tasks. This is intentional—the player characters are heroes, and should be able to do most mundane things without a roll.

Certainly the authors could have included a skill system covering activities such as "horse riding" or "swimming", but doing so is actively detrimental to heroic gaming. Had we included a "horse riding" skill, characters would start falling off their horses. This strikes us as unnecessary, in the context of heroic adventure gaming, so if you seek a generic skill system for your game, seek it elsewhere. Success at most horse-riding tasks (for example) is automatic.

Where a player character tries something beyond the mundane, the GM should determine the chances of success on an adhoc basis. The GM should look at the circumstances and the character's class, level, race, and ability scores and make an informed judgement about his or her chances of success. This could be a flat judgement—"you succeed" or "you fail"—but is more commonly a die-roll of some kind. The GM should usually tell the player what the chance is, ask the player if they still wish to proceed, and if so allow the player to make the appropriate roll him- or herself. However the GM always has the right to roll the dice on behalf of the player, or in secret, if the GM feels the situation demands it.

In determining the ad hoc chance for success for various tasks, it may be helpful for the GM to look at other similar tasks that have already been defined. For instance, the strength-based chances to Open Doors and/or Bend Bars, the magic user's intelligence-based Chance to Know Spells, and the constitution-based roll to survive System Shocks can all be extrapolated to cover a wider variety of situations. The same applies to saving throws, which consider class and level rather than just raw ability, so that high-level characters will be generally more successful than lower level characters, and each class will tend to have areas of speciality (clerics better at tasks that require a save vs death, Magic users at tasks that require a save vs spells, etc.)


The GM should set up some simple system for book-keeping and may wish to delegate some tasks to the players. The present author, for example, keeps a piece of scratch paper by his books and makes a tally mark when each turn has elapsed, enabling him to see at a glance when to roll for wandering monsters, when the next rest period is required, and when the party has run out of lantern fuel. In extreme cases, if the campaign has grown so large that ten or more players per session is typical, an assistant GM can help—the assistant, or apprentice, GM helps the main GM with book-keeping and organisation, and may help the main GM design new dungeons and adventures, eventually becoming either a co-GM or branching off into a separate sub-campaign.