A NOTE: On Encumbrance, Treasure and Session Structure

Recently I’ve discussed the importance of resource based risks to classic play, and perhaps offered a reason to use this play style, but I’ve saved the most important element of meaningfully including supplies in your game until now. Risks and tables that offer dire consequences for characters who fail to appreciate them are maybe interesting, but they’re superfluous if they never have a chance to enter play, which requires limiting player supply. The primary way to limit supply is Encumbrance. The amount of equipment, weapons and armor that a character or party of characters can carry in game is important because it provides a clear metric of character strength beyond hit points that makes intuitive sense to any player, and with proper rules can make mechanical sense in a dungeon crawl (see The Risk Economy Part II).

 

THE RUST MONSTER’S LEGACY - EQUIPMENT AS SUPPLY

 

When I was a kid, playing Basic Dungeons and Dragons after school in a friends basement around 1986, one of the game events that upset me the most was an encounter with a ‘Rust Monster’. The propeller tailed, bug thing completely devoured out fighters’ weapons and armor before being killed. A prized +2 sword was reduced to a jagged crumbling shard of metal and both me and the other two players were aghast at the horror visited upon us. The Rust Monster is one of the uniquely Dungeons & Dragons creatures that Gary Gygax invented from a bag of plastic toy “dinosaurs”, but more than any of those others: the Bulette, the Owlbear and perhaps the Umber Hulk, the Rust Monster is a monster tied to the exploration side of the game. It doesn’t do much harm to characters’ hit points, but it destroys the party’s ability to engage in combat with other enemies, dramatically increasing the risk of further adventuring because it attacks equipment rather than hit points. It upset us young players for precisely this reason, because it operated outside the structure of risks and rewards we expected. It was such a scandalizing outrage that I still remember it because equipment (especially that magic sword) is something that D&D players value greatly but view as static, and because its destruction made a great deal of obvious sense.

The rust monster's natural form

 The Rust Monster itself looks a bit goofy, but there’s nothing wrong with what it represents, a novel danger or obstacle for the players to think around. It’s a trick monster, but notably the Rust Monster’s trick, while dangerous, doesn’t represent a danger of immediate death for a character. The loss of equipment hurts a character effectiveness and increases overall risk and is difficult to replace in the adventure locale, but it has an intuitive logic - tools break - and so it doesn’t feel like a gamified and artificial mechanic. There’s an important distinction her though, as much as Rust Monsters, prying open doors, rolling down slopes of jagged scree and wedging moving walls apart make losing weapons or armor make sense, weapons, armor and magic items are generally considered permanent - players don’t expect them to be exhausted, while other items are disposable. Food, light sources (including oil bombs), scrolls, potions and a few mundane supplies like iron spikes are something that players expect to exhaust during the adventure and resupply in town. It’s useful to make a distinction between these types of supplies - semi-permanent equipment and usually the easily exhausted consumables (or supplies).

Risk to character equipment has a history beyond special monsters such as oozes and the rust monster, and the AD&D includes a set of saving throws for equipment based on its material and various types of disasters. Potions boil, freeze and shatter while scrolls survive falls and “crushing blows” with ease. It’s a fairly functional system really, applying both to the loss of player items in trying circumstances and player character efforts to destroy mundane objects: cutting ropes, burning down doors and such.

AD&D Monster Manual Rust Monster
David C. Sutherland III (?)

5th Edition also makes some nods to the possibility of equipment destruction, thought it seems more concerned with players destroying obstacles and dungeon furniture.“When characters need to saw through ropes, shatter a window, or smash a vampire's coffin, the only hard and fast rule is this: given enough time and the right tools, characters can destroy any destructible object. Use common sense when determining a character's success at damaging an object. Can a fighter cut through a section of a stone wall with a sword? No, the sword is likely to break before the wall does.” - 5th Edition Dungeon Masters’ Guide, Pg. 245. Afterwards the 5th Edition Guide provides useful rules for item Armor Class, Hit Points and damage reductions/thresholds to make durable objects stronger.

The burdensome nature of these rules (or punitive one if used in every situation where they might apply - do all the items in a PC’s pack need to save after every blow, after every battle?) makes them something that often gets forgotten in play, but exact method (the saving throws above, or perhaps a simple X in 6 chance of breakage) is unimportant and can be streamlined or applied only in extreme situations. The special revulsion and horror that I showed towards the rust monster as a young player shows that risks to character equipment remains a valuable tool for a GM who wants to expand risk while attacking something other then character HP, but like most serious risks, if a character would have a chance to evaluate risk of breaking an item then the player should be forewarned.

SCARCITY CREATES VALUE

While both equipment and consumables are fitting resources for obstacles and dungeon denizens to attack, destroying equipment is more serious, something for special situations, specific obstacles or rare foes. Consumable supplies however deplete over time, and the player expect to expend them either through exploration or to overcome obstacles. Both kinds of supplies provide the designer or Game Master alternatives to directly attacking character survival by depleting hit points, while remaining intuitively understandable to the player, and threatening (memorably so with special items) enough to prompt player decision making and planning. For consumables the long term threat of losing a hard to replace or special item is lacking, but consumable items are also usually survival items and also of obvious importance to continued adventuring and the character’s margin of safety.

If the expectation that consumables will be depleted during an adventure is obvious, the method for making them a meaningful consideration is less so. In the past I’ve written about reducing light using an exploration/overloaded encounter die or otherwise enforcing timekeeping limitations, but supply reduction becomes meaningless if there’s little limit on the amount of supplies a party can carry. Yes, classic games have numerous traditional work-arounds: henchmen, magical bags of holding, and permanent light sources - but each of these resources offer clear risks. Henchmen get eaten by grues, flee or quit when treated poorly or ignored. Magical items are hard to find (consider getting rid of bags of holding and that every magic weapon might not be a light source) and are still pieces of equipment that can be lost or destroyed. So while these player methods to avoid supply issues are effective they also have in game logic and can be thwarted by the environment. They tend to be solutions that require experience as well - and there’s nothing that says the risk of running out of supplies can’t diminish over time.

At the beginning though the solution that keeps supply important, and places time limits on exploration is encumbrance - how much the characters can carry around with them. Contemporary and classic systems often include encumbrance mechanics, but they tend to be more complex than many players want to deal with, and so are abandoned. Instead I suggest implementing a simple slot based encumbrance system.

Most encumbrance systems for Dungeons & Dragons have been weight based, the coin system of Basic D&D and pound based systems of later editions. There’s an intuitive sense to weight based encumbrance - we often talk about how many pounds or kilograms we can lift or carry, but no encumbrance system simple enough to use regularly is going to be able to simulate weight distribution, ease of access and carrying unwieldy items in a meaningful way. True simulation isn’t the goal of encumbrance (or any game mechanic), and the purpose of encumbrance is to make provisioning questions (what to bring on an adventure) and supply risk questions (what to use when) important. Slot encumbrance does this very well.

From 1981's Basic D&D - Errol Otis

Slot Encumbrance: The simplest system of slot encumbrance is to allow one item per point of character Strength. Armor, weapons, equipment, supplies, potions, scrolls, and treasure all count towards encumbrance (though usually coinage and reasonable amounts of ammunition doesn’t). This system works well enough, but it’s far more difficult for low Strength characters. Alternatives include 10 items plus or minus Strength bonus, or 10 items plus Strength divided by 3. All will give characters between 7 and 16 items of carrying capacity.

For any items over the listed amount the character is encumbered, and depending on system suffers penalties to movement, attack, defense and skill rolls. Carrying more than 1 and ½ or double the number of items set by Strength is impossible.

Significant Items: One issue with slot encumbrance that often comes up is determining what exactly counts as a significant item. Does every torch or flask of oil? A stick of chalk? A canteen of water? What about treasure? Players will find creative and often convincing ways to justify items having no or limited encumbrance. Remember however that the goal here isn’t simulation, rather it’s to create a risk of scarcity and enable tracking supplies without expending too much player and game master attention on them. A significant items is then defined not by its size or bulk, but by its obvious usefulness.

Some exceptions and deviations of course also make sense. Generally I limit these to ammunition (a quiver of 20 - 30 arrows or a case of bolts is one significant item), throwing weapons (a bucket of javelins, brace of tomahawks or bandolier of throwing knives - about 2-4 of them), torches (3-4 per bundle - making them more efficient than a lantern despite their shorter burn duration), a set of thieves tools (see Incidental Resources below) and of course purses of coins (each 100 - 1,000 coins or fraction thereof). Some items of course are insignificant, and this is itself what makes them valuable - gems and jewelry for example, but also items like letters of marque, mementos or correspondence from lost loved ones. Otherwise consider it a significant item and count it towards encumbrance - potions, scrolls and flasks of oil are especially important to include because they are extremely useful and need to be tracked.

Incidental Resources: Without slot based encumbrance a crafty player will write all sorts of small, light and useful items on their character sheet: chalk, charcoal, whetstones, grease, rags, black pepper, marbles, vials of acid, flasks of wine, small hand tools, paper, etc. Some of these (bags of marbles, lard or pepper big enough to impede pursuers for example) should definitely be considered significant items, but others don’t need to be. Additionally, there’s always the question of items that seem reasonable for dungeoneers to carry, but which a player might not include on their sheet - a whetstone or extra bowstring for example. One way to approach these sorts of items is to simply assume that characters have them tucked in a pocket or at the bottom of their pack, but sometimes an item is a bit too odd, a bit too useful, or the player’s rationale for why their character is carrying it is a bit too convoluted. When this happens it’s sometimes useful to view equipment less granularly, less as a collection of individual items, a list recorded and accounted for on the character sheet, then as a supply of useful items that the character - a skilled and practiced adventurer - has stockpiled against the risks of the mythic underworld.

This abstraction isn’t a new idea - it begins with the concept of “Thieves’ Tools”. Mentioned repeatedly in Basic Dungeons and Dragons, the AD&D Dungeon Master’s Guide, and the AD&D Player’s Handbook, the nature of “Thieves’ Tools”, “Thieves’ Tools and Picks” and “Burglar’s Tools” isn’t explored in detail. All sources mention lock picks, rakes and tension bars (a bit of an anachronism given the crude nature of the early modern locks one might expect in medieval fantasy), but what about other tactics of the historical house breaker? Do thieves tools contain a crowbar? acid for burning out locks? Historical burglary gets stranger then this though, the thieves of medieval Cairo employed special wall breaking tools (an iron claw or hook) and most crucially a tortoise with a candle affixed to its back - to release into the darkened target site where it’s wandering would draw attention away from the thief.

Do thieves’ tools include a tortoise and a candle stub? If a player insisted that there’s did I’d be tempted to allow it, as it’s such an evocative and fantastical historical detail. However, it’s also pretty bizarre and not entirely obvious so a check of some kind would be appropriate. A 3D6 vs. Wisdom (did the character remember to pack the weird stuff in their kit?) to check for plausible but rare objects in a thieves’ or other kit of tools (medicine, engineering, scribe etc.) and for each item add a d6 (so the second odd item in that kit would require a 4D6 check). Of course kits don’t contain large amounts of a specific item, or multiples of the same item - a thieves’ kit might have a little can of oil with a long spout to grease hinges, but it’s not going to be enough to refill a lantern for more than a turn, or to use as an oil bomb and even if the kit holds one distracting tortoise, it will never hold a pair.

This system may seem overly generous, the Dungeon Crawl is premised on resource depletion and allowing players to think up items they need when they need them seems antithetical to this. However, this may again be a place to make a distinction between supplies, which the players and characters know will be exhausted over time, and tools or equipment. The locus of play in the Dungeon Crawl is puzzle solving, figuring out schemes and clever tricks to overcome obstacles. Allowing incidental items in adventuring kits encourages this - it doesn’t avoid the risks of supply but rather encourages creative solutions to puzzles and obstacles.

A final mechanic that can improve the use of incidental resource items/tools is a limit on their number (perhaps replacing the die check to see if the PC has a specific item). Each kit has three slots (blank below it on the character sheet) that the resourceful player can fill in with incidental items as needed. However, once filled in these are the odd items in that character’s kit. If the character wants additional rare items, they will need to carry a second kit. This of course works well with encumbrance, and means that specialist type characters (thieves, rangers, engineers, healers) will begin to carry more and more tools and equipment as they become more experienced.

Usage Dice: Inventory and encumbrance can be further simplified by the introduction of usage dice. This is a mechanic simplifies inventory tracking, much like the the exploration/overloaded encounter die simplifies timekeeping and compliments . Bundled resources (most obviously ammunition) don’t have a specific number (a quiver of 30 arrows is simply a quiver of arrows) of items or uses but rather a “usage die”. This obviously works for magic items as well: wands seem especially interesting to attach a usage die to.

After each battle (for ammunition) or use (for magic items) the player rolls against the usage die, which varies by how many of an item is carried together. On a ‘1’ the item is either exhausted, or the usage die size goes down. Generally a 1D6 or 1D8 usage die that steps down is effective for plentiful ammunition like arrows of quarrels, while a 1D4 makes a good counter for javelins. This system is abstracted, and depends on viewing combat in an abstracted manner as well - with a combat round as a pass or exchange, lasting far longer than is popular in contemporary games. Abstracted combat assumes that an archer is firing many arrows (even if only having one chance a round of hitting effectively) each round, just as the swordsman is exchanging several blows with his enemy.

 

THE LURE OF GOLD

Tracking supplies and encumbrance isn’t just important on the “risk” side of the “risk v. reward” calculation, it impacts what the characters can recover from the dungeon as well. Simply, treasure has encumbrance value, and carrying it off imposes risk and compels player decision making. When and what treasure to take? How to move it? Where the party might cache treasures to continue exploring? What to leave behind? Are there supplies worth sacrificing to carry treasure? In a decently sized dungeon these considerations all add to play, providing more places for player decision and interaction with the setting.

Encumbrance also encourages clever GMing by creating a hierarchy of treasure with portable gems and jewelry at the top and heavy unwieldy items at the bottom. This is not a new concept, and in one of the passages that (around the same time as my rust monster umbrage) made me think beyond simplistic hack and slash play was this amazing passage from the AD&D Dungeon Master’s Guide:

“A pair of exceedingly large, powerful and ferocious ogres has taken up abode in a chamber at the base of a shaft which gives to the land above. From here they raid both the upper lands and the dungeons roundabout. These creatures have accumulated over 2,000 g.p. in wealth, but it is obviously not in a pair of 1 ,000 g.p. gems. Rather, they have gathered an assortment of goods whose combined value is well in excess of two thousand gold nobles (the coin of the realm). Rather than stocking a treasure which the victorious player characters can easily gather and carry to the surface, you maximize the challenge by making it one which ogres would naturally accrue in the process of their raiding. There are many copper and silver coins in a large, locked iron chest. There are pewter vessels worth a fair number of silver pieces. An inlaid wooden coffer, worth 100 gold pieces alone, holds a finely wrought silver necklace worth an incredible 350 gold pieces! Food and other provisions scattered about amount to another hundred or so gold nobles value, and one of the ogres wears a badly tanned fur cape which will fetch 50 gold pieces nonetheless. Finally, there are several good helmets (used as drinking cups), a bardiche, and a two-handed sword (with silver wire wrapped about its hilt and a lapis lazuli pommel to make it worth three times its normal value) which complete the treasure. If the adventurers overcome the ogres, they must still 'recognize all of the items of value and transport them to the surface. What is left behind will be taken by other residents of the netherworld in no time at all, so the bold victors have quite a task before them. It did not end with a mere slaying of ogres . . .” - AD&D Dungeon Master’s Guide. pg. 92.

Ignoring, or reveling in, the ‘High Gygaxian’ language, the passage here describes a theory of treasure placement the is organic and evocative, but also appear at odds with the treasure tables famously included in every D&D edition. Closer consideration shows that Gygax’s “Ogres in a hole” example doesn’t really venture far from a treasure table. In the Monster Manual Ogres have a varied chance of treasure. Each individual carries 20 - 80 Gold pieces with a large treasure cache in the ogre lair (protected by a family of ogres, including ogre children that fight as goblins - the inclusion of non-combatant monster children is an interesting artifact of early D&D worth discussing at some point). Because of the relatively low chance for various classes of treasure (types of coins etc) used in the AD&D treasure tables it’s hard to obtain an average for ogres - no one class of treasure has a greater than 50% chance of appearing. However, ogres have a lot of treasure options - mixed hoards (Treasure type B) maxing out at 5,500 GP value of so in coins and several gems and jewelry (both of which vary greatly in value), gem caches (Type Q) and buckets of potions (Type S).

It’s easy to see where one could get a result of 2,500 GP in mostly copper and silver coins with 50 GP of gold and a few lower quality gems and jewelry from the tables. The interesting element is how Gygax interprets this hoard result. Instead of providing coins, gems and jewelry, much of the hoard is transformed into art objects, while more or less retaining the encumbrance issues associated with each class of treasure.

While this is a great example of how to make treasure interesting, as description that works with setting (plundered trade goods), as objects the characters may covet (the fancy but non-magical sword) or even adventure hooks or setting information (who did that necklace belong, or who generally has been suffering from these ogre raiders?) It also creates complexities for the players - how to carry off the valuable unwieldy hoard?

The 5th edition of Dungeons & Dragons doesn’t emphasize treasure, and while it’s treasure tables include some evocative art object and other interesting treasures, but they have a smaller scale - fewer gemstones, a small variety of valuable items, and lower values, especially at high levels, then the 1st edition. This reflects 5th Edition’s view that “Adventurers strive for many things including glory, knowledge, and justice. Many adventurers also seek something more tangible: fortune.” 5th Edition Dungeon Master’s Guide. pg. 133. In short, treasure in modern Dungeons & Dragons is an ancillary goal behind the intangible impulses of epic adventure leading to heroism. There is nothing wrong with focusing on heroism as a play goal, but it creates its own set of necessary mechanics and problems - which often run counter to the classic dungeon crawl playstyle. The core player facing mechanic that 5th Edition uses to promote heroic conflict as the locus of play is experience points.

Treasure in 5th edition provides no experience points for characters. Its immediate value is limited to equipment it can buy, which is quickly exhausted unless a GM includes the sale of magical items. Treasure is effectively stage dressing in the 5th Edition, something included because it has always been part of Dungeons & Dragons, but nearly optional to play. Contrast this with early editions where treasure is the primary source of player advancement - defeating monsters in AD&D provides far less experience then their normal treasure caches. For example the two ogres in Gygax’s example above provide 900 XP (or 4.5 x the XP needed to level from 1st to 2nd) while in AD&D they offer approximately 470 XP (or .23 the XP needed for a fighter to rise from level 1 to level 2). The treasure of the exemplar ogres however (say 2,500 GP worth) brings the experience point reward for the encounter to almost 1 and ½ the experience needed to level a single fighter from level one to level two.

The question many contemporary players may have about this change is “why isn’t that a good thing?” I don’t suggest it’s good or bad, but it certainly encourages a specific kind of play, assuming one’s players are motivated by gaining levels. It encourages combat and to the same extent discourages exploration. As I’ve mentioned before this shift is reflected in almost every aspect of 5th edition, from character creation and abilities that emphasize combat prowess to an adventure design philosophy that focuses on the “encounter”. Again, there’s nothing inherently wrong with a game about slaughtering monsters to gain power - and if players want to focus on tactical combat 5th edition’s design choices seem ideal. However, the dungeon crawl is a playstyle about exploration and puzzles - not tactical combat. Yes, tactical combat can be a big part of it, but the early editions’ rules around experience mean that recovering treasure (ideally without the risk of combat - a serious risk to character survival in early editions) is the goal of advancement minded players.

Treasure for XP: One of the classic elements of Dungeon Crawl games, and early Dungeons & Dragons more broadly is the idea that recovering treasure provides the bulk of Experience Points. There are many useful side effects of this system, broadly it’s predictable, easily comprehensible and doesn’t favor one solution to obstacles over others (either combat or meeting GM imposed goals). More importantly here, it also amplifies the Resource Management/Supply aspect of the Risk Economy.

When characters gain XP through treasure accumulation (there are several slightly varied ways to handle this - but here the important idea is that recovering treasure obviously and directly benefits the characters) it creates incentives that work in tandem with exploration play. Not only does the tricky task of treasure recovery offer an immediate supply complication but a focus on treasure for experience also:

• Puts exploration, setting interaction and information gathering at the center of the play experience (rather then combat) by treating challenge types equally. With players seeking wealth determining what challenges (combat, puzzle, social) will best provide it at the lowest risk the locus of play shifts from tactical combat. With combat dethroned from its position as 'the' player activity there is more room for player retreat and even less need to 'balance' encounters.
  
  
• Demphasizes combat, as violence provides no mechanical advantage over negotiation or trickery. Demphasis allows encounters with creatures that are very dangerous because there is no assumption that every encounter is a test of combat ability and no inherent mechanical benefit in destroying powerful foes.
  
• Adjusts character goals by making them very clear and more open to player decisions.
   
• Has a neutral (or perhaps even negative) moral valiance that is very natural to contemporary players in that everyone understands the struggle to make money for survival - but is open enough to allow for a range of moral play from dastardly evil to saintly good.
• Creates diegetic (story) freedom, by simplifying goals and decoupling them from a specific narrative (XP for combat could do this as well - but since it favors one type of challenge over another and creates a narrative of relentless bloodshed it tends to require a heroic narrative to justify the sheer amount of murder involved). Players don't have to guess what the GM has structured as a story to move the game forward and will likely create their own goals based on interaction with the world and its factions.
• Simple and clear mechanical explanation for how level advancement works, giving players metrics on determining reward vs. risk. This allows GMs greater freedom in obstacle design as reward is not coupled with the completion of a specific set of acts.

Replacing combat XP with treasure XP is simple. The GM will need to think about how much treasure is included in a dungeon level, and what challenges (including simply distance from an entrance) protect or conceal each cache if he wants to meter character leveling, but this isn’t difficult. Treasure placement is a Dungeon Crawl design principle I hope to address after completing these essays on the ethics of play (two of three major topics are covered as of this post), but it should be relatively straightforward to adapt to later editions after a quick glance at early edition treasure and experience tables or a well designed dungeon crawl (I recommend Gygax’s B2 - Keep on the Borderlands, but there are plenty of other wonderful options, including third party projects like Patrick Wetmore’s Anomalous Subsurface Environment).

 

SUPPLY IN PRACTICE

Ending the Session: One of the common frustrations with Supply (or any other depletion mechanic) is reconciling it with session length. The party enters the dungeon and explores it at one pace, but the players’ time ticks away at another, and it’s a rare table that can sustain the day long session that exploring a large dungeon might take. Outside convention play or there’s rarely enough time to fully explore a dungeon of any size in a standard 2 to 4 hour session, which is about as long as seems manageable online and is plenty of time for most players to have their fill of Dungeons & Dragons for a week or two. This is one of the reasons to use small encumbrance limits - because larger limits prevent resource constraints from having meaning within the length of an individual session.

At the end of this session, a party may only have explored as few as five complex locations, or as many as thirty, but they are unlikely to have ferreted out the secrets of a full sized (40 - 100 rooms) dungeon location. This impacts play in two ways: player availability and resupply. In the 5th Edition Dungeon Master’s Guide little thought seems to have been made regarding session length, but the solutions offered for missing players all suggest that adventures are likely to exist as continuous narratives. See “Missing Players". pg. 235 - 236. A missing player is cause for radical solutions such as another player or the GM playing the missing player’s character. While not explicitly stated this seems to suggest that play between sessions stops and picks up in the moment. Starting and stopping sessions in media res creates difficulties for the classic dungeon crawl or any other play style that depends on resource depletion or limited resources (even HP & spells). When sessions stop in the midst of action and the next begins where the last left off there are often complications. Players are absent (taking resources with them), a previously absent player joins (fresh and fully equipped), or it’s hard to remember exactly how many resources (supplies, HP, memorized spells) were left at the end of the previous session. Yes, great attendance and record keeping eliminates these problems - but like most systems of play, demanding perfection tends to make a game far less fun.

For these reasons, which compliment the spatial puzzles of location based exploration, it’s useful to insist that each session is a unique delve, ending at a haven - a town, a camp outside the dungeon, or safe zone within. Doing this not only allows players and their characters to come and go, but also encourages player choice as changing goals and adapting to last session’s discoveries is far easier. It encourages scouting and planning for specific monstrous threats and makes figuring out the spatial secrets (alternate entrances, short cuts, secret doors) of a location more valuable as characters seek to return to areas they haven’t explored. It’s possible to use a table (much like the one I posted for escaping the dungeon without light) that makes camping in the mythic underworld extremely dangerous, but a less punitive and informal approach seems to work better in practice. Giving your player notice towards the end of the session that they’ve reached as far as you can go today and that you’ve only got enough play time left to run a quick escape (and it should be quick if they backtrack through already explored areas - a random encounter check or two is sufficient even for all but the most dangerous areas). Remember with this approach that just as players have an option to resupply, prepare, recruit, train, research and carouse in between sessions the denizens of adventure locations will react to intrusions - bringing in allies or replacements, setting up guard posts, barricades and traps, or even fleeing entirely depending on their nature (though usually full restocking should take several sessions).

An Open Table: One advantage of always beginning your game in a haven is that you can swap party members in and out. The absent are deemed to be unavailable for the specific session, hanging about town. This means it’s possible to play with a larger rotating group of players and entirely disposes of the difficulty of introducing replacement characters or new characters between sessions. It also lets new players try your game out for a session or two, gives players with busy schedules a chance to play when they can and makes for varied parties of adventurers with different goals and interests - something that can provide a great change of pace between sessions.

There’s nothing new about this “Open Table” Approach, it’s the one taken during most of the early formative campaigns, but it also makes stories focused on a specific PC’s backstory or scene based narrative campaigns more difficult. The party is recreated each session, as a new expedition with the players that are available for that specific session. This works exceptionally well in online and club play, where there might be a large pool of players that can’t make a weekly game, but can make an appearance once or twice a month. Because characters will level at different rates, this style of play tends to work best for dungeon crawl games, where combat challenges aren’t assumed to be fair and where solving puzzles doesn’t require specific levels of character skill or competence. Because of the characters varied power levels an open table also benefits from using a system with a relatively flat power curve, like those of early Dungeons & Dragons, where a low level character can still benefit the party and where high level characters don’t entirely overawe even low level opponents.

Troupe Play: A more contemporary alternative version of the Open Table style of play, one that allows a changing cast of players but runs less risk with varied character power levels and more use of ongoing story elements is the idea of troupe play. The players can change often but the characters remain similar because they are all part of the same organization, thus sharing goals and a larger story about that organization’s fate. A group of mercenaries, a troupe of traveling actors, a gang of thieves, a college of wizards, an order of templars - there’s a lot of possibilities.

Players each have their own characters within the troupe, and the GM may include an NPC leader of some kind to offer direction (though this has a significant potential to stifle player choice and planning). Alternatively the players have access to a pool of characters - but no ownership of them. Players who play a lot will likely return to their favorites (though a GM might require specific characters take a number of sessions to recover between adventures). The majority of characters in the troupe aren’t special - regular mercenary soldiers for example: low level, with mundane stats and equipment, and after each session they survive the player using them can update their sheet and add a bit about their personality, so that the next person to play them has something to work off of. More important figures in the troupe will appear over time, either starting as characters with better stats, unusual classes and equipment, or developing through play because a persistent player plays them multiple times. Either way the troupe structure creates a shared set of interests for the characters that provides direction to new or infrequent players.