Which Version of Palace of the Silver Princess Should You Run?

Over on Bluesky I did a readthrough of both the orange cover and green cover versions of Dungeon Module B3: Palace of the Silver Princess. My reading of the orange version was more in-depth, whereas my reading of the green version was more for the sake of comparison. I thought it would be productive to summarize my thoughts in an easily-digested blog post.

For those who don't know, the publication history of this module is a mess. The Wikipedia article about the module sums it up pretty well. Essentially, Jean Wells wrote the original, orange version of the module. For one reason or another, TSR waited until the day before the module's publication (when the copies arrived at the office for shipping) to decide that it was unacceptable. All of the copies were recalled or confiscated and destroyed.

The explanation at the time was that the illustration of "The Illusion of the Decapus" by Laura Rosloff was objectionable because it was overtly sexual and was exposing children to the S&M fetish. This was also in the midst of the Satanic Panic, so another explanation was that the illustration gave credence to the claim that D&D was promoting devil worship.

Since then, the story has changed. It is now widely believed that the actual objectionable image was Erol Otus's illustration of the ubues (three-headed, -armed, and -legged monsters with both male and female heads), which depicted the monsters with heads that were essentially caricatures of TSR executives (notably Brian and Kevin Blume, though accounts vary). Apparently, Wells had wanted to replace this illustration before the module went to print, but was told it was too late.

Whatever the exact reason, the end result was the same. The original version of this module was essentially memory-holed. Many players from that time are familiar with only the revised green version, which is the version that was officially released after the original underwent many significant revisions at the hands of Tom Moldvay. You can read Wizards of the Coast's own flawed hagiography and download the orange version for free here.

The original version was similar to Dungeon Module B1: In Search of the Unknown, with many rooms empty of monsters, traps, and treasure, with instructions for the DM to determine these elements their self (the quality of these empty rooms, in my opinion, is not quite as good as those found in B1).

Not every room was empty, however. The module also contained several intentionally designed encounters with unique monsters, among them the aforementioned ubues and decapus (pictured on the cover of both versions), sentient bubbles, a giant six-legged rat with a toothy duck bill and as many hit dice as an owlbear, an amoeba which disguises itself as a pool of water, paralyzes you, and dissolves you, a giant snake with an enchanting whistle, giant marmoset monkeys large enough to serve as mounts for halflings, and a moss that puts you to sleep and then grows over, suffocates, and devours your unconscious body.

The original module was a relatively straightforward site-based adventure - an ancient palace in ruins, and legends of a giant ruby still hidden within. What will you do with that information? Well, probably go explore the dungeon and try to find the ruby - classic D&D stuff.

In addition to the dungeon, there was a minimally-detailed map of the surrounding region with short entries about settlements and geographic features - a city ruled by a Chaotic baroness (the only living descendant of the titular Silver Princess, who wants the ruby for herself), a shire occupied and terrorized by the baroness's soldiers for unknown reasons, a halfling village which is home to escaped prisoners and slaves, a canton cut off from the barony by a dark and foreboding wood, a swamp inhabited by strange monsters, where spellcasting is unpredictable, a mountain range riddled with abandoned dwarven mines, and another range plagued by storm clouds conjured by an evil wizardess.

The module was filled with mystery. According to legend, the kingdom ruled by the Silver Princess fell into ruin when a stranger arrived at a party to see the Princess's ruby. A red dragon appeared over the valley and burned the land, a man in silver and blue armor upon its back. However, upon exploring the ruined palace, the players may find that the relationship between the Princess and the stranger was not as adversarial as it seemed - or, perhaps, she was corrupted by this sinister interloper. So what exactly happened?

The module does not provide an answer, because the DM was expected to. It is filled with unanswered questions. The land surrounding the Palace is wandered by a tinker in a special wagon that can float across rivers and lakes. Should the players have occasion to visit his home, they will find in his bedroom a suit of silver armor covered by a blanket, and a strange set of riding equipment too large for a horse - intriguing! Inside the dungeon is a chaotic cleric, Catharandamus, who has partnered with Aleigha, a werebear with a ruby sword, who is believed to be another descendant of the Princess. It is possible for the player characters to align themselves with the two, but it's unclear why they're at the Palace and what they're trying to accomplish. And what of the soldiers with the wolf insignia, who can be found on the random encounter tables? Who are they, and what are they doing here?

By comparison, the green version is much more focused and utterly non-mysterious. The region surrounding the Palace is totally omitted. Rather than being an ancient ruin, the Palace was recently besieged by Arik, an evil outsider akin to a godlike entity. The ruby is a vessel of sorts to Arik, which drove some of the Palace inhabitants mad, petrified the others, and trapped the Princess and the dragon-rider (who is unambiguously good in this version) inside the ruby itself. The Palace is held within in a stasis field while Catharandamus (teleported there by Arik's magic) tries to summon Arik into the world.

The party is visited in their dreams by the Protectors, a race of magical beings from the land of Faerie who beseech the player characters to enter the palace and free the Princess. (The Protectors are also in the original version, but they are simply magical guardians of the Palace's tower, warding it against evil so that the party can safely rest within.) Throughout the adventure, the Protectors serve as the voice of the DM, giving the players hints as to where to go next, how to destroy the ruby, and the like. If the Princess is freed, the land returns to normal, and all is more or less hunky dory.

Gone too are the module's original monsters (excepting only the decapus - which loses its ability to create hyper-specific and horrifying torture illusions - and some deadly plants). Instead, the magic of Arik has attracted the usual D&D fare - goblins, hobgoblins, orcs, and the like - to terrorize the place. These creatures were present in the original as well, but it is suggested that they are minions of Catharandamus, which he has brought there as an insurance policy against the berserkers who serve Aleigha. Although the two are aligned, the former does not trust the latter. 

There is a hint of faction play in the original. This nuance is totally absent from the green version. Aleigha and her dwarf henchmen are replaced by a werewolf and some orcs. Aleigha's ruby sword (a magical heirloom which introduces a complication to the evil baroness's claim to the Palace and ruby) is now the Sword of Arik, one of many MacGuffins which can be used to destroy the ruby and yield the module's happy ending. The mysterious tinker, himself a potential complicating piece of the puzzle, is omitted entirely.

I would be remiss not to mention Travis. In the orange version, he is a paranoid old adventurer who tortured and killed his party out of greed and now wanders the dungeon holding "court" - capturing, torturing, and killing adventurers to take their treasure. In the green version, he is instead a Palace guard driven mad by the magic of Arik, but otherwise doing more or less the same.

That all is not to say that the green version is without its merits. What might compel you to run one version over the other? Well...

If you want a classic old school D&D dungeon, run the orange version. In the orange version, the Palace is an ancient ruin filled with odd monsters, tricks, and traps. The party's goal is to go there and find the ruby, doing with it whatever they please. The player characters are not assumed to be do-gooders with any higher purpose to their motivations. It's a classic D&D smash-and-grab operation.

If you want a narrative-driven adventure or your player characters are assumed to be heroes, run the green version. The green version has more of a high fantasy vibe - an unambiguously good princess and her kingdom in peril at the hands of an unambiguously evil villain. There are magical fairies which can serve as the DM's voice to nudge the players along the intended path. If the player characters are victorious, they're heroes of the realm and earn a substantial reward, a day of thanksgiving in their honor, and even a ceremony where they're presented with medals.

If you want the Palace to be part of a larger world, run the orange version. The orange version includes a regional map with additional locations like the evil barony and its surrounding settlements, and mysterious wilderness locations like the Abaddon Woods, Misty Swamp, and Thunder Mountains. There are multiple factions suggested by the module's descriptions of these locations, which place the Palace within a larger context and give the players other activities to pursue before or after finding the ruby. The locations are minimally detailed, so you will have to do some legwork to bring them to life.

If you want the Palace to be self-contained (or want to drop it into an existing sandbox), well, you can probably run either version, but I would still lean towards the orange version. Both versions describe a decent dungeon, but I lean towards the orange one because of the Palace's age in that version, which lends itself towards being plopped down as just another ruin on the map. In the green version, the calamity is extremely recent and the kingdom is in immediate peril as a result. It seems harder to just drop that into any old sandbox.

If you want a dungeon that "makes sense", well, neither version really does. Whichever version you use, the dungeon layout is somewhat nonsensical and "random". The green version feels more "palatial" and provides names to all the rooms, which makes it easier to imagine what those places were originally used for, but there's also weird details like tapestries, mosaics, and paintings depicting the Princess and the dragon-rider. The dragon-rider apparently arrived at the Palace the day before Arik struck, so why is there so much artwork of him? This makes more sense in the orange version, where an indeterminate amount of time passed between the dragon-rider's arrival and the ruin of the Palace.

If you want to make the dungeon "yours", run the orange version. Several rooms in the dungeon have space for the DM to add monsters, traps, and treasure. This ensures that no two DMs' version of the dungeon will be the same. Likewise, many characters' motivations are not made explicitly clear. You will have to come up with these yourself, but by the end you will have your own private Palace of the Silver Princess.

If you want everything figured out for you, run the green version. This version's keying of the dungeon is much more complete. You won't have to do any prep beyond the usual or rolling to find out at the table, and you won't have to figure out what any NPC's "deal" is.

If you want original monsters, run the orange version. As I've already mentioned, this version of the module has some truly horrifying, memorable, and silly original monsters. I like the living bubbles that drown you, but if a six-legged duck-billed giant rat as tough as an owlbear is too goofy for you, you may not like this version as much.

If you want your standard D&D monsters, run the green version. While goblins, hobgoblins, and orcs aren't anything new, that doesn't mean they're bad. There are merits to every dungeon containing unique monsters, but there are also drawbacks. If your setting is otherwise plagued by monstrous humanoids, maybe you want your Palace to be overrun by them as well. It might be odd if your campaign features the usual D&D fare, only for this one location to randomly have very weird creatures that don't show up anywhere else in your game.

If you want faction play, run the orange version. The faction play in this version of the module is not especially strong, but it does exist. In the green version, basically everything in the Palace is a minion of Arik and united against the player characters. In the orange version, there is at least tension between Catharandamus and Aleigha and their respective minions. The ubues are a faction of their own and may play the party against Catharandamus, align themselves with him, or even help the party against him. That's not to mention the faction play you can have between the evil baroness and the other settlements in the region, plus Aleigha, who may also be the rightful heir to the Princess.

If you want to make Travis an imposing foe, run the green version. Travis the insane adventurer/palace guard is probably my favorite NPC here. I really like the idea of this crazy old fighter ruling a corner of the Palace like a kingdom, torturing intruders, holding court, and training giant apes to guard his treasure. In the orange version, Travis is a lone operator, a 1st level fighter, and has just 4 hit points. Shame! In the green version, he has 16 hit points, is a 3rd level fighter, and commands a retinue of orcs which he dresses in guard uniforms. Cute! He apparently hangs out in his office writing gibberish reports on the conditions of his orc troops, mimicking the behavior of the guard captain he slew. This is welcome additional detail. 

You could also easily take the suped up Travis and plant him in the orange version along with his orc guards. I probably would not have Travis wandering the dungeon all on his lonesome and relatively easy prey for adventurers - I'd have him fortified in his lair, sending his orcs out to round up intruders. It would be very cool if the orc guards captured the player characters and brought them before Travis for judgement.

To conclude, here's a random miscellany of other neat features from both versions:

• If you want your player characters to be able to hitch a ride on a wagon that floats across rivers and lakes, run the orange version (or add the Tinker to the green version if you're dropping it into a sandbox).
• If you love paralyzing your player characters and depriving them of air, run the orange version. Many monsters in the module have paralyzing attacks. There's a whole section in the introduction with drowning rules based on what armor the character is wearing, and many monsters and traps try to drown the characters or suffocate them.
• If you want snakes in unusual places, run either version.
• If you want "good" boxed text, run the green version. The orange version mixes the typical "at a glance" information with stuff that the players would only learn once their characters investigate an object in the room (the contents of buckets and barrels, for example). The green version's boxed text is much more consistent with what you'd usually expect from these descriptions.
• If you want your lady thieves to be HOT and evil and not just "above average-looking" and friendly, run the green version (or run the orange version but make the lady thieves HOT and evil).
• If you want your player characters to be able to capture and sell or tame a young cave bear, run the orange version.
• If you want a trap to drop 20 pounds of glitter on the player characters, run the orange version.
• If you want a harp made of ice that can magically calm any beast, run the orange version (or assign this property to the same harp in the green version, which in that version is just another MacGuffin that can destroy the ruby, rather than a more generally useful magic item).
• If you like strange eggs in unusual places, run the orange version.
• If you want to teach your players a lesson about robbing a Palace they're supposed to be saving, run the green version. When they free the Princess at the end of the module, she will get angry at them if they looted the Palace on their way, demanding that they return the stolen treasure and giving them a lesser reward. For a game like D&D (especially this era of D&D), this is quite cruel!

Overall, it probably sounds like I prefer the orange version...and I do! In fact, reading both versions has convinced me that there's little if any reason for the green one to exist at all. Obviously, Jean Wells was screwed over by TSR, thrown under the bus, and her work appropriated and basically erased entirely by the revision. 

The orange version is certainly incomplete, with unstocked rooms, dungeon history, and important NPC motivations left to the imagination of the DM, but this is pretty consistent with the trend in modules published at the time. There seems to be a belief that Moldvay was needed to "finish" the module, regardless of the other justifications for its burial, however I would argue that he does a poor job. This is not intended as a dig at Moldvay, who was put in a position which is probably not conducive to good adventure writing. 

What results is a module with an entirely different feel than the original. It is less unique and weird. It wields clumsy narrative devices in service of a black-and-white, linear adventure of unambiguous heroism which feels more at home among the content published by Wizards of the Coast today, rather than among the B series adventures of the time. 

The fault is not with Wells or Moldvay, Rosloff or Otis, but with TSR themselves. Whatever the actual reason for the reaction to the orange version, the failure is not one of imagination or execution, but of development and oversight. If there was objectionable material, it should have been identified before the module went to print, and it should have been the original authors, illustrators, and editors who were left to determine how to correct the issue, not another writer brought in to hastily patch and retcon the initial work in the eleventh hour.

As I have demonstrated, there are reasons why one might wish to run the green version over the original. I may have even convinced you that the green version is right for you, but I have not convinced myself. For me, it's Jean Wells's Palace of the Silver Princess all the way.

Creating a New Regional Sandbox for B2: The Keep on the Borderlands

Over on Bluesky, I recently finished up my readthrough of Dungeon Module B2: The Keep on the Borderlands - actually, by the time this post sees the light of day it will have been over a month ago since I finished it!

One thing that struck me about B2 is the regional map, which looks like this:

Not included in the image above is the scale of the map, which is 100 yards per square. Nothing wrong with that, except:

Huh. It takes an hour to walk 300 yards. That doesn't sound right. A commenter on my thread pointed out that this is probably an error:

That certainly makes more sense, but I actually kind of like the larger scale produced by the error. If we just count squares and determine how long it takes to get from place to place using the originally printed guidelines we get the following travel times (assuming 8-hour travel days):

• ~1 hour and 40 minutes from the fork in the road to the Keep
• ~16 hours (2 days' travel) from the Keep to Area 1, the Mound of the Lizard Men
• ~32 hours (4 days) from the Keep to Area 2, the Spiders' Lair
• ~10 hours (1.25 days) from the Keep to Area 3, the Raider Camp
• ~6.5 hours (.8 days days) from the Keep to Area 4, the Mad Hermit's lair
• ~16 hours (2 days) from the Keep to the Caves of the Unknown
• ~16 hours (2 days) from the Keep to the Caves of Chaos

Maybe it was intended for the player characters to be able to walk from the Keep to the Caves of Chaos in 2 hours and 40 minutes, but I like my players to have to work just a little bit more to get to their destination. It gives them a chance at having random encounters which produce interesting and unexpected results and complications, and I want that to happen. I like the larger scale.

I don't however, like the map. I don't like counting squares. So I made a new map:

This map more or less yields the same travel times as those above, and I'll show my work.

The scale is 6 miles per hex. Plains hexes take 1 hour to enter and 2 hours to cross (same with roads if the player characters are on one, but in this case it doesn't matter because the only roads go through plains hexes). Forest hexes take 2 hours to enter and 4 hours to cross. Heavy forest hexes take 4 hours to enter and 8 hours to cross. Forested hill hexes take 3 hours to enter and 6 hours to cross. Forested mountain hexes take 4 hours to enter and 8 hours to cross. Fen hexes and wetland forest hexes take 8 hours to enter and 16 hours to cross. Rivers take 1 hour to cross.

So:

• From the crossroads to the Keep is 2 hours
• From the Keep to the Mound of the Lizard Men is 16 hours (6 hours on the road, 1 hour off road, 1 hour to cross the river, and 8 hours to enter the fen)
• From the Keep to the Giant Spiders' lair is 15 hours (2 hours on the road, 1 hour off road, 1 hour to cross the river, 4 hours across the plains, 1 hour to cross the other river, 4 hours across the plains again, and 2 hours into the forest) instead of 16
• From the Keep to the Raider Camp is 11 hours (2 hours on the road, 1 hour offroad, 1 hour to cross the river, 4 hours across the plains, and 3 hours into the forested hills) instead of 10
• From the Keep to the Mad Hermit's lair is 6 hours (3 hours across the plains and 3 hours into the forested hills) instead of 6.5
• From the Keep to the Caves of the Unknown is 13 hours (10 hours on the road, 1 hour offroad, and 2 hours into the forest) instead of 16
• From the Keep to the Caves of Chaos is 15 hours (12 hours on the road, 1 hour offroad, and 2 hours into the forest) instead of 16

I had to make some concessions due to the self-imposed restraint of using hexes as a measurement tool and placing the locations of interest perfectly in the center of the hexes, but most of these travels times are just an hour off from the numbers I originally calculated. The exception is the Caves of the Unknown, which are pretty close to the Caves of Chaos in the original module. At this scale, I can't properly recreate the intended distances (I'd probably have to include them both in the same hex, which I don't want to do).

Now, I think the Mound of the Lizard Men, the Giant Spiders' lair, and the Raider Camp are a little anemic as presented in the original module, so I want to juice them up a little to make them feel like proper AD&D monster lairs using that edition's numbers, since that's probably how I'd run this module. 

My version of the Mound of the Lizard Men contains 40 (!) lizard men (the original contains 7 males, 3 females who fight similarly to males, and 8 young who do not fight - AD&D does not make this distinction, so all 40 of mine are simply fighting lizard people). I would flesh out the lair to make it a muddy warren of burrows and dens, forcing the players to actually get in there and try to use the space to their advantage. My lizard men do not simply come parading out to be killed. I imagine this environment would play similarly to the Caves of Chaos themselves, with tight corridors and lots of united enemies which can quickly be mobilized to respond together to threats.

I'm okay with leaving the spiders basically as is, but I would change them to AD&D's giant spiders (the module uses black widow spiders, which aren't explicitly a monster type in AD&D). AD&D's giant spiders have a bit more hit dice, better AC, and do less damage. Otherwise they're basically the same.

B2's Raider Camp contains a dozen chaotic fighters - hardly enough to pose any real threat to the Keep. To make them a more proper and dangerous faction in the sandbox, I'm going to treat them as brigands (Chaotic Evil bandits). My Raider Camp has 100 brigands led by five 3rd level fighters, three 4th level fighters, two 5th level fighters, two 6th level fighters, a 7th level lieutenant, and a 9th level leader with six 2nd level guards. They have nine important prisoners and 30 camp followers and slaves.

I want to bring this region more in line with the wilderness stocking described in AD&D's Appendix B. I have 49 hexes total, and 16% of them should contain either a settlement, stronghold, or ruin - that's seven or eight hexes. I already have the Keep and the Caves of the Unknown, which I'm going to classify as a dungeon/ruin, since that's more or less how it's described in the text. I'm classifying all other locations, including the Caves of Chaos, as monster lairs. That leaves me with five or six areas of inhabitation to add.

Let's go with six. I rolled up two single dwellings, a village, two more strongholds (a deserted keep with a monster and a totally deserted tower), and another ruin (a tomb). Let's add those to the map:

I cluster the settlements around the Keep along the river. The ruined strongholds are off to the east, near the caves. These areas may have once been cleared of forest, but since the strongholds fell into ruin, the forest has reclaimed them. The ruined keep (now inhabited by a will-o-wisp) is the one farther to the east. The ruined tomb is to the northwest of the Keep on the Borderlands. Roads still lead to all three ruins, and the Keep probably charges a toll to use them.

Next I want to determine if I should add any more monster lairs. My rule of thumb is that 10% of plains hexes, 20% of forest and hill hexes, 30% of deep forest and mountain hexes, and 40% of wetland hexes should have monster lairs. As of now I have six empty plains hexes, eight empty forest hexes, five empty hill hexes, one empty deep forest hex, two empty mountain hexes, and nine empty wetland hexes. 

I roll and determine that I need one monster on the plains, one in the forest, one in the hills, and four in the wetlands. I place these and get the following:

Here is my final key:

• 01.01: Brownies
• 01.02: Tomb Ruins
• 01.05: Dwelling
• 01.07: Raider Camp
• 02.01: The Mad Hermit
• 02.03: The Keep on the Borderlands
• 02.04: Village
• 03.05: Dwelling
• 03.06: Ghouls
• 04.06: Giant Weasels
• 05.03: Deserted Tower
• 05.05: The Mound of the Lizard Men
• 05.07: Giant Spiders
• 06.02: The Caves of the Unknown
• 06.07: Tribesmen
• 07.01: Green Dragons
• 07.02: The Caves of Chaos
• 07.04: Deserted Keep (will-o-wisp)
• 07.05: Wolves
• 07.07: Beholder

01.01 Brownie Burrow: This forested hillside burrow is the lair of 11 brownies, chief among them Leongath. They are friendly towards adventurers and particularly helpful to those that are Lawful Good, offering to make or repair mundane items. They were friends of the Mad Hermit (02.01) - before he went mad - and can guide the party to his lair. 

If made aware of the settlements south of the Keep on the Borderlands (02.03), the brownies will eventually migrate there, settling in the surrounding farmlands (01.05, 02.04, and 03.05) before eventually taking up residence in various homesteads.

01.02 Tomb of the Druids: A dungeon for 1st level characters. This forested ruin is little more than a burial mound surrounded by ancient standing stones. Here the druids of the area were entombed. Many died of natural causes, but many more were slain in their conflict with the Cult of Evil Chaos. The dungeon is inhabited by undead and cursed vermin, and the sacred burial grove deep within is guarded by Balaesus, a former druid reanimated as a ghoul out of his desire for revenge. He is hostile to all intruders, suspecting them of being tomb robbers or cultists. 

If the Mad Hermit (02.01) is killed and brought here for interment, Balaesus will become friendly (he assumes that the Cult of Chaos is responsible). He is somewhat knowledgeable about the Cult and knows about their shrine within the Caves of Chaos (07.02).

01.05 Hale Homestead: A peaceful farm tended by Edrin and Mara Hale. They mind their business and are unfriendly towards adventurers, mistaking them for potential brigands. Their son, Tamsin, was captured by brigands and taken to the Raider Camp (01.07) while wandering the fields across the river (02.05).

01.07 Raider Camp: See Dungeon Module B2: The Keep on the Borderlands, but use brigands. There are 100 normal brigands with five 3rd level fighters, three 4th level fighters, two 5th level fighters, and two 6th level fighters. They are led by Sobek (lieutenant) and Geppert. Geppert is protected by six 2nd level guards. The brigands have 30 camp followers and slaves tending to their camp.

The brigands have seven light horses and three medium horses. Groups of ten mounted riders will venture forth into the plains (02.05 and 02.06) daily to snatch up villagers and homesteaders who wander across the river. They have acquired nine prisoners in this way and use them to demand a ransom from the Keep on the Borderlands (02.03).

• Geppert: 9th level fighter. He is neutral towards adventurers and is willing to entertain them on the off-chance that they will be useful in raiding the Keep and its surrounding lands, or as middle men in securing and delivering a ransom from the Keep.
• Sobek: 7th level fighter. He is unfriendly towards adventurers, suspicious that they will reveal the camp's location to the garrison at the Keep.
• Prisoners: Nine in total, includes Tamsin Hale of Hale Homestead (01.05) and several villagers from Southron Village (02.04).

02.01 The Mad Hermit: See Dungeon Module B2: The Keep on the Borderlands. The hermit is a former druid, captured, tortured, and driven mad by the Cult of Evil Chaos in their conflict. He has encountered Gungroyeth of Wyrmwood Cave (07.01) at a distance in the forest east of his lair (03.01 and 03.02) and believes the dragon is a demon conjured by the Cult to devour his soul.

02.03 The Keep on the Borderlands: See Dungeon Module B2: The Keep on the Borderlands. The Keep controls a number of settlements to the south (01.05, 02.04, and 03.05). Several villagers have been captured and taken to the Raider Camp (01.07) and are being ransomed. The Castellan would rather pay adventurers to root out the raiders than pay the ransom. 

Patrols have spotted Gungroyeth of Wyrmwood Cave (07.01) in the forest to the east (03.03) and have heard reports that he flies over the eastern road (05.04, 06.03, and 07.03). The dragon is viewed as a threat to the Keep's survival, and adventurers are encouraged to seek out its lair and slay it.

(I would also make the Castellan at least 9th level, since that's the level at which fighters in AD&D can have strongholds.)

02.04 Southron Village: A peaceful village within the domain of the Keep on the Borderlands (02.03). It has a mixed population of about 900 and is ruled by Lady Althaea, a half-elf aligned with the Keep's Castellan. The village is a gathering place where homesteaders from the surrounding lands grind wheat, bake bread, and trade for goods. The village controls the road leading north to the Keep.

Sir Garrick commands the village militia (about 90 can be mustered in times of urgency) and leads patrols across the river, clashing with brigands from the Raider Camp (01.07). Several of Southron's homesteaders have been captured by the brigands and are being held for ransom. Lady Althaea would prefer to pay the ransom to ensure the safe return of the villagers, but the Castellan will not allow it.

NPCs and their henchmen:

• Delmare: NG human fighter 1. She can be recruited as a henchman by a non-Evil character.
• Reynrielle Smugbottom: N dwarf fighter 1/thief 2. She and Hurlbert dream of getting access to the Keep's inner bailey and stealing everything that isn't nailed down.
• Hurlbert: CN human thief 1
• Vanelis Silentsmile: CE elf fighter 1/magic-user 1/thief 2. He is a secret agent sent by the Cult of Evil Chaos to spy on Southron.
• Shaeris Summerpot: LG halfling fighter 2. She is friendly and can be convinced to aid the party on a sufficiently good-intentioned quest, such as freeing prisoners from the raiders.
• Damiane: LE human fighter 3. She and Xiomara will fight as mercenaries against the raiders if the price is right.
• Xiomara: NE human fighter 1
• Lady Althaea: NG half-elf fighter 8/thief 11. Loyally serves the Castellan, but will recruit adventurers to free prisoners from the Raider Camp.
• Sir Garrick: LG human fighter 7. Wants to make a name for himself in battle with the raiders.
• Squire Padgett: NG human fighter 3
• Magus Awarnach: LN human magic-user 3

03.05 Vetch Homestead: An unassuming homestead on the edge of the swamp. Harlan, Ilyra, and Corvin Vetch live here, eking out an existence at the edge of the lands controlled by the Keep on the Borderlands (02.03). They are friendly towards adventurers, offering hospitality to those on their way northeast or into the swamp.

Corvin is a restless teenager and will want to accompany adventurers on the road as far as the fork leading to the ruined tower. It is difficult to convince his mother and father to allow this, since lizard men (05.05) and wolves (07.05) prey upon travelers on the road, and a dragon (07.01) has been spotted flying over the area.

03.06 Blackfen Hollow: The lair of three ghouls - priests of the Cult of Evil Chaos who were captured and executed by the Castellan of the Keep on the Borderlands (02.03) and dumped here without a burial. They prey upon any who wander into the fens, and have already devoured Ilyra of Vetch Homestead (03.05).

04.06 Blackfen Den: The lair of five giant weasels - two adults and three young (50% grown). The adults' pelts will fetch 2,000 gold pieces each if brought to the Trader in the Keep on the Borderlands (02.03). The young are not yet totally wild and can be trained as hunting animals or guards if taken from their parents. If gifted to the Castellan of the Keep, the Castellan will grant access to the Keep's inner bailey.

05.03 Aramis's Tower: A dungeon for 7th level characters, This ruined tower was once the stronghold of Aramis, a cruel and evil human magic-user concerned principally with strange experiments and arcane power. Aramis lived and died long before the Keep on the Borderlands (02.03) was built, and the forest has since grown to reclaim his ruin. While the tower has collapsed, the mazelike dungeon beneath, where Aramis conducted most of his experiments, is largely intact. 

A pair of umber hulks has moved in, drawn by the strange lingering magical energies. They lair in Aramis's vault in the dungeon's deepest level.

05.05 The Mound of the Lizard Men: See Dungeon Module B2: The Keep on the Borderlands. There are 40 lizard men here. The lizard men prey upon those straying from the road to the north (04.04 and 05.04), lurking in the river to snatch them and carry them back to their lair. They are aware of the ghouls of Blackfen Hollow (03.06), but avoid the place.

05.07 Giant Spiders: See Dungeon Module B2: The Keep on the Borderlands, but use giant spiders instead of black widows.

06.02 The Caves of the Unknown: See Dungeon Module B2: The Keep on the Borderlands. In the spirit of the original module, this area is left undetailed so that the DM can fill it in themselves!

Okay, fine. Let's say it's a dungeon for 1st level characters. My Caves of the Unknown were a dumping ground for Aramis's experiments. Throw in a secret passage or collapsed tunnel leading to Aramis's Tower (05.03), perhaps. The most powerful creature here is an ochre jelly which has been festering in some kind of alchemical sump. There's all sorts of other weird stuff here. Happy?

06.07 Poison Needles Tribe: This tribe are the original inhabitants of the grasslands surrounding the Keep on the Borderlands (02.03). When the Keep was being constructed, they clashed with the garrison and were ultimately forced off their land. To avoid the Keep's frequent patrols, they retreated across the southernmost river. When they came under attack from the Raider Camp (01.07), the tribe fled east into the fens, where they now live in their village.

There are 100 fighting men and women in the village, plus a mix of about 100 noncombatants and children. There is an additional 1st level druid, ten 3rd level fighters, ten 4th level druids, four 4th level subchiefs, and three 6th level druids. They are led by the head druid, Killian, and their chief, Ronan.

• Killian: N human druid 8. Suspects adventurers of being raiders or hired soldiers from the Keep. If convinced otherwise, he will be helpful. He wants the raiders to be driven out, but believes in peaceful coexistence with the Keep if the Castellan will agree to leave the tribe alone.
• Ronan: N human fighter 5. Advises Killian to be suspicious of armed adventurers and other fortune-seekers. He favors an isolationist approach to outsiders.

Giant spiders sometimes wander into the tribe's fens from 05.07. The tribe hunts them and harvests their venom, which the tribe's fighters use on their weapons (save versus poison or be killed). 

07.01 Wyrmwood Cave: This forested cave beneath the roots of a great tree is the den of a mated pair of very old dragons, Gungroyeth and Rustathultox, and their very young offspring, Zergosh. Zergosh is well fed and is in a deep sleep within the lair, while his parents are awake and active, alert to threats from the Caves of Chaos (07.02), the Lawful inhabitants of which often bring the dragons tribute to avoid their wrath.

• Gungroyeth: Large, powerful, and deceptively stealthy. He is of average intelligence and does not speak, but goes out into the forest to hunt for food to bring back to his child and mate.
• Rustathultox: Sleek and regal in appearance. She is very intelligent and capable of speech. In addition to the language of dragons, she speaks the common tongue and the languages of goblins and kobolds. She is the schemer of the pair, remaining in the lair to guard and educate Zergosh. She keeps detailed records of gifts given by the clans of the Caves of Chaos. 
• Zergosh: A disproportioned and overfed very young dragon. Due to his healthy diet, he is often asleep. When awake, he is tutored in the ways of "domain management" by his mother. He is not yet capable of speech.

When hunting, Gungroyeth ranges as far west as the forest east of the Mad Hermit (02.01) and the Keep on the Borderlands (02.03), and as far south as Fenside Cave (07.05). The dragons are largely indifferent towards adventurers, so long as the adventurers do not try to harm the dragons or steal their treasure. They desire the treasures of the Keep, Aramis's Tower (05.03), and Gerwald's Folly (07.04), and wish to drive the Chaotic creatures (who do not bring them tribute) from the Caves of Chaos.

Gungroyeth and Rustathultox each have 8 hit dice and 56 hit points. Zergosh has 7 hit dice and 7 hit points. Like all dragons, they can be subdued and sold. Zergosh, being young, small, and unintelligent, can be sold for just 1,400 gold pieces. Gungroyeth, being larger but unintelligent, is worth 28,000 gold pieces. Rustatholtox, being the only dragon capable of speech, is worth 33,600 gold pieces.

Zergosh is the most likely to remain subdued, followed by Gungroyeth, then by Rustathultox. They are unlikely to serve a Good master for long - for example, if sold to the Castellan of the Keep, they are likely to break free and unleash havoc on everyone inside. Rustathultox, being intelligent, is likely to usurp her master rather than remain subdued - for example, if sold to the Raider Camp (01.07) or the Cult of Evil Chaos, she is likely to overthrow Geppert or the Evil High Priest and rule in their stead.

07.02 The Caves of Chaos: See Dungeon Module B2: The Keep on the Borderlands. The Lawful inhabitants of the Caves (kobolds, orcs, goblins, and hobgoblins) peacefully gather once a month to contribute to a tithe which is then delivered to the dragons of Wyrmwood Cave (07.01) so as to avoid their wrath. The Chaotic inhabitants (bugbears, gnolls, and cultists) do not contribute, which has put the whole valley at risk and heightened tensions among the monster clans.

07.04 Gerwald's Folly: A dungeon for 7th level characters. This ruined keep was built by the magic-user Gerwald to rival Aramis's Tower (05.03). In his attempts to surpass Aramis, Gerwald summoned Barachiel Vex, a pit fiend which eventually slipped its bonds and slew Gerwald. Vex has remained here, studying Gerwald's research to conceive of a way to open a gate to bring his infernal legions to the mortal realm.

Since Gerwald's demise, the ruin has been reclaimed by the forest, and a lonely spirit (a will-o-wisp) now haunts the surface ruins. It is the soul of Gerwald, bound to this world by hatred of Vex. It is friendly towards adventurers, though still very much Chaotic Evil. It attempts to lure adventurers down into the subterranean ruins to confront the fiend, often leading adventurers directly into danger, more so out of blind determination than a desire to see the adventurers perish.

Gerwald's spirit will not aid adventurers in combat except against Vex, though it still understands the languages Gerwald knew in life and may be convinced to help.

07.05 Fenside Cave: This riverside cavern is the home of a pack of seven wolves and their eight cubs. They are desperately hungry and attack any travelers which seem weak. 

They range as far west as the outskirts of Vetch Homestead (03.05), but mainly attack travelers on the road at 05.04 and 06.03. They can be found near Aramis's Tower (05.03), the Caves of the Unknown (06.02), and Gerwald's Folly (07.04). They are occasionally preyed upon by lizard men from the Mound of the Lizard Men (05.05) when ranging into the latter's hunting grounds.

The cubs are not yet totally wild and can be trained as war dogs or hunting beasts if taken from their parents. If gifted to the Castellan of the Keep on the Borderlands (02.03), the Castellan will grant access to the Keep's inner bailey.

07.07 Bogwatcher Warrens: This mazelike marshy crater and its warrens are home to Mirexath the Bogwatcher, a beholder more ancient than any creature in the region. It is hateful and aggressive and will attack any it encounters without listening to reason, though it doesn't travel far from its lair. Despite its lengthy residence in the area, it knows little of what goes on in the region.

Here's a final updated map with some of the monster lair icons changed:

This updated and expanded sandbox definitely distorts some of the intention of the original module. The purpose of the wilderness encounters in the original is to simply provide additional points of interest for the players to investigate. None of them are larger, more complex, or more dangerous than the Caves of Chaos. By beefing up the number of raiders and lizard men in their respective lairs, I'm only reducing the number of viable low level adventure sites in the region.

On the other hand, I've codified the Caves of the Unknown as a 1st level adventure site and have added an additional dungeon for 1st level characters (the Tomb of the Druids). I've also added two 7th level dungeons (Aramis's Tower and Gerwald's Folly). Here is also a breakdown of monster lairs (and the Keep) organized by the level of the most powerful monster found therein:

• Level III: Brownie Burrow, Blackfen Hollow, Blackfen Den, Mound of the Lizard Men, Fenside Cave
• Level IV: Mad Hermit (mountain lion)
• Level V: Giant Spiders
• Level VII: Raider Camp (Geppert), Keep on the Borderlands (Castellan), Poison Needles Tribe (Killian)
• Level VIII: Wyrmwood Cave
• Level X: Bogwatcher's Warren

If we also take for granted the module's suggestion that the Caves of Chaos are fit for 1st to 3rd level characters, than the progression through adventure sites in the region would look something like this: 

Tomb of the Druids/Caves of the Unknown/Caves of Chaos > Brownie Burrow/Blackfen Hollow/Blackfen Den/Mound of the Lizard Men/Fenside Cave > Mad Hermit > Giant Spiders > Raider Camp/Keep on the Borderlands (if the players wanted to say, kill the Castellan)/Aramis's Tower/Poison Needles Tribe/Gerwald's Folly > Wyrmwood Cave > Bogwatcher's Warren.

That's a solid amount of stuff for levels 1 to 3 and level 7. Getting from level 3 to 7 might be a bit challenging, as will getting from level 7 to 10. However, that isn't to say that characters must be of a certain level to confront the corresponding monsters. Level III monsters can appear on the 1st level of the dungeon, level IV and V monsters on the 2nd or 3rd level, level VII monsters on the 5th level, level VIII monsters on the 6th level, and level X monsters on the 7th level. Thus is might be appropriate to say that this sandbox is fit for characters level 1 through 7.

I'm pretty satisfied with this. It's scope creep, but a good kind, where the play area contains a variety of play environments with different levels of challenges, allowing the player characters to grow within it. If I run Keep on the Borderlands at some point, I'll be using this - my Keep on the Borderlands.

Play Environments, Power, Complexity, and Agency in D&D

Location types exist in close proximity to one another but serve very different gameplay functions.

There are basically four main location types in D&D (and D&D-like games): settlements, strongholds, dungeons, and lairs.

You might disagree with this taxonomy. The Welsh Piper's Hex-Based Campaign Design identifies two other types of "major encounters" (or locations): religious orders and natural phenomena, as well as many more "minor encounters" (or locations): camps, beacons, construction sites, battlefields, crossings, gathering places, and more. Sachagoat's Re-Inventing the Wilderness identifies not just towns, lairs, and dungeons but also "scenes" and "utilities". Hexographer has icons for oases, geysers, windmills, wineries, and graveyards. 

You can't really boil all the locations player characters might encounter in their adventures down to just these four broad categories. But these are, I believe, the "core four" locations around which a D&D-like game tends to be structured. They are the major locations characters will be interacting with again and again. 

There might be one settlement or stronghold which serves as a home base, a single megadungeon that props up the whole campaign, and one big bad monster lair such as a dragon's den or an orc camp which threatens the entire region until it is dealt with, or there may be many settlements to familiarize oneself with, many strongholds whose rulers must be obeyed or subverted, many dungeons to haul treasure out of, and many monster lairs to clear out or otherwise deal with.

Three of the four location types are those on the AD&D DMG's Inhabitation table:

Lairs are not included because it is assumed that these will be discovered when monsters are randomly encountered in their lair based on the % in lair statistic. I prefer to stock my sandboxes with these lairs ahead of time, which makes them set locations. Thus, when I create a sandbox, I'm populating it with these four location types. First I'm checking each hex for human inhabitation (including ruins). Once all areas of inhabitation are determined, I check the remaining hexes for monster lairs.

While I'm not going to argue that Gygax was right about everything or that his approach to designing D&D should color our perception of all such games in the present and into perpetuity, I do find this distillation of location types into a manageable few to be appealing. There are certainly others, but to me these feel very much like the most important, central locations in a campaign. (I replace religious orders, as defined by the Welsh Piper, with strongholds ruled by religious character-types like clerics, druids, paladins, and monks. And how many groves, stables, windmills, and wineries are keystone locations in the average campaign? I think it's not many.)

For further reading, I recommend my post on creating minimalist sketches of these four location types for the initial stages of sandbox prep. I think that's a good primer for better understanding the four location types and what I consider to be the most important information to know about each before diving into more detail-oriented prep.

What interests me about these four location types is what gameplay utility and experience each provides - that is, what is each location type for and how does it feel to interact with each?

Since my primary focus at this time is AD&D, I will be analyzing this topic through that lens. Not every D&D-like game will have settlements filled with unhelpful NPCs and burdensome taxes, strongholds ruled by character-types of a specific level with a specific number of men-at-arms, dungeons which conform to that particular early D&D logic, or lairs containing hundreds of orcs, but AD&D has come to color my perception of D&D and its derivatives as a whole, and many of the ideas here are still broadly applicable to similar games.

Settlements

I've already written a great deal about settlements in AD&D, the purpose they serve, and the vibe they give off. In short, they are incredibly useful places where player characters can accomplish mundane tasks such as reprovisioning, recruiting henchmen, and gathering information, as well as more fantastic efforts such as acquiring spellcasting services from high-level NPCs or even auctioning off captive dragons and selling monster organs.

While being useful, settlements are also often unfriendly, oppressive, and incredibly dangerous places where player characters are strangers and viewed as threats to the powers-that-be, where relentless taxes check their rampant accumulation of wealth, NPCs are sensitive jerks, diseases flourish, powerful factions war with one another in the streets, and almost everyone you encounter is trying to trick you, call the guards on you, rob you, fight you, or even kill you for looking at them funny.

Due to their nature as densely packed social environments with labyrinthine rules, odd taboos, and resourceful individuals and groups with independent interests in the player characters and what they do, settlements are immensely complex environments where the players must carefully consider and prepare for each move they make. 

It is this element of settlements which enables them to provide their other gameplay function (aside from sheer utility): political intrigue and high-level faction play. Above all, settlements are a test of the player's mastery of the complex social fabric.

Settlements are the lifeblood of an adventurer, allowing them to turn treasure into gold, gold into experience points, experience points into levels, NPCs into contacts, contacts into information, and information into further adventure. They also don't exist solely for the player characters' pleasure and utility. While they might serve as a home base to the characters, they are also highly dangerous. A wise player of a low level character will spend as little time as possible within a large settlement like a town or city before turning their attention back to those environs in which the adventurer truly belongs.

Dungeons

Dungeons are the low level adventurer's true home. Considering that low level player characters in a town or city might stumble into a high-level NPC who has a problem with them, a demon that's escaped a cleric's or wizard's control, or a vampire that wants to suck their blood, they really shouldn't be there if they can avoid it - or at the very least, they should avoid going out for a walk at night as much as they can. 

Similarly, they have no armies with which to contend with stronghold rulers or enough renown to be welcomed into the Keep on the Borderlands's inner bailey, and they are not strong enough to venture into the unpredictable wilderness and start clearing out dragon dens and orc camps. Dungeons, on the other hand, are much more suitable environments for early career adventurers.

Dungeons are "balanced". The 1st level of the dungeon is fit for exploration by 1st level characters. The monsters therein are drawn from a handful of tables including creatures that such characters could reasonably fight or otherwise overcome or engage with. It gets more dangerous the further down they go. Unlike the town or city, the dungeon is a simple environment with rules which are fairly easy to grasp. 

Obviously, that doesn't mean the dungeon is safe - there are monsters, tricks, traps, and other nasty things. Higher level monsters can wander up from deeper levels of the dungeon. A trick stairway might turn into a slide that sends the party down more levels than they wished to descend. A falling ceiling trap may instantly kill an adventurer of any level. They might get lost within, never to escape.

The value of a dungeon to an adventure is that it is a relatively simple environment with easily understood rules and structure. It is a place to collect treasure and experience points early in the game when the player characters are most vulnerable. Dungeons do not usually contain groups of monsters which require a small army to uproot, and there is not usually a powerful ruler sticking their nose in the party's business and exerting their will upon them.

That's not to say that dungeons are without intrigue. Just as settlements offer complex political faction play, dungeons have factions too - they're just (usually) smaller. The factions in dungeons aren't guilds of thieves and assassins with high-level cutthroats at their disposal or merchant consortiums that control the levers of power - they're groups of 2d4 hobgoblins, 4d4+2 kobolds, and d6+6 orcs (at least initially).

While factions and social play are important in dungeons, this is not the primary challenge of these locations. Rather, dungeons are a test of the player's mastery of a space. The primary mode of play in these environments is exploration. 

Play in dungeons is concerned with figuring out and dealing with whatever's in the next room. If there are monsters, how do the player characters overcome them, circumvent them, or collaborate with them? If there's a trick, what does it do? If there's a trap, how does the party disarm it or get around it? If there's treasure, how do they identify it and get it out? 

More advanced is not just confronting the dungeon room by room, but understanding the broader picture and how the room fits into the greater space. How does the room connect to others? Can the party use that information to plan an ambush or a tactical retreat? Where is it safe to rest? Where are the secret doors? The stairways? The hidden dungeon entrances and exits? What is the fastest route to the party's goal which allows them to minimize encounters with wandering monsters?

Exploration is key to settlement play as well, but like much of settlement play, the medium through which information is acquired is social. Players learn where to find an inn, where to buy equipment, and where to go to recruit henchmen by asking around. They learn how to buy and sell smuggled goods, where the assassins' guildhall is, and how to infiltrate the palace by leveraging their connections. They don't (usually) do it by mapping all the buildings on graph paper, looking for spatial discrepancies, and knocking on the right 10' section of wall (although they could, I suppose). Navigating the physical space of the settlement is less important than navigating its social dynamics.

Strongholds and Lairs

I feel compelled to address both strongholds and lairs together, since the two are difficult to parse. Is not a stronghold the lair of its ruler? Is not a monster's lair a place where it fortifies itself against intrusion by adventurers? A group of bandits, berserkers, brigands, or dervishes might rule a stronghold as their lair. A group of demihumans or humanoids or even a dragon might lair in a deserted stronghold.

It is worth pointing out that there are different kinds of monster lairs. What exactly a lair looks like is determined by the type of monster that lives there. It could be a cave, a camp, a village, a castle, or something else entirely. The most important distinction is not so much the type of place but the quality of creature therein - namely, is the monster intelligent or not, and does it appear in numbers or with minions?

The lair of an owlbear is a relatively simple - though still dangerous - environment, while a lair of orcs is more complex because of their greater numbers, ability to organize to repel invaders, and respond intelligently to repeated attempts to infiltrate their home. Whether they live in a cave, fortified village, or ruin will make a difference, but not a huge one. A dragon's lair might be comparable to an owlbear's if it's a dimwitted variety of dragon. It will be more complex if the dragon is intelligent or cunning, and even more so if the dragon has enlisted a small humanoid army as worshipers or mercenaries.

The idea for this post initially came to me after reading dungeon modules B1: In Search of the Unknown and B2: The Keep on the Borderlands. I felt that B1 is a good example of a prototypical dungeon (itself a deserted stronghold) and that B2's titular Keep and Caves of Chaos are good examples of a stronghold and a lair (or series of lairs), respectively. 

But reading through the Caves of Chaos, with their large homogenous populations of allied monsters who can all be mobilized to ward off intrusion together, along with their tightly-packed corridors where control of strategic chokepoints would prove essential, I wondered, "Is this really that different from a stronghold?" Whether the players choose to clear out the Caves or ally with their inhabitants to instead lay siege to the Keep, does the actual gameplay experience of doing either of those two things differ all that much?

Gus L. writes about this element of Gygaxian design better than I ever could in his post on All Dead Generations, Gygax's Fortress. In it, he writes about how many of Gygax's iconic adventurers like The Keep on the Borderlands, the Against the Giants series, and the Vault of the Drow (all of which detail what I would classify as "lairs") put siege-style gameplay front and center. Siege, you say...like the thing you do to a stronghold?

This is not to say that there is otherwise no overlap between all of these location types. The ruler of a settlement might dwell in a fortified tower. The Keep on the Borderlands is a place to provision and gather information as much as the Village of Hommlet. A settlement might contain a ruined manor which is essentially a dungeon, and a dungeon might house a large group of squatters or other inhabitants which form a de facto settlement within. Monsters might lair in settlements just as they do in the wilderness, and a wilderness elf enclave is a settlement of a kind. A ruin could have once been a stronghold. Monsters lair in ruins.

But the relationship between strongholds and lairs strikes me as different because they feel so similar when you consider how the players interact with them. If the players wish to roust an evil ruler from their castle, it's going to look very similar to clearing out the Caves of Chaos - a protracted series of forays into a defensible position held by an organized, (mostly) united enemy.

Much of this can be chalked up to aesthetics. A stronghold is a stronghold because it is a tower, keep, or castle. It has a gatehouse, battlements, oil cauldrons, ballistae, catapults, or whatever. A lair is a lair because it's a cave, an informal camp, or a hut in the woods. But it isn't always that - hobgoblins and orcs sometimes lair in above ground villages with ditches, ramparts, palisades, gates, and guard towers. Many giants and aquatic creatures like locathah and tritons live in castles.

You could also attribute the difference to ideology. A stronghold is a stronghold because the creature who rules it is viewed as a person or as part of the established order, regardless of their alignment or relationship to their neighbors. A lair is a lair because it is inhabited by crude monsters who are outsiders, without rights or the privilege of simply living. It's essentially Law versus Chaos.

That's all well and good (or bad, I guess), but it's unsatisfying to me because I'm primarily concerned with the gameplay function of these locations - what use do they present to the players, and how do the players engage with them? From that perspective, how are strongholds and lairs different?

One way to think of it is how the locations project power. Strongholds control a certain area around them and collect revenue from the settlements within their domain. They send out patrols of men-at-arms led by high level fighters. The stronghold's ruler might emerge to challenge the party to a joust, demand a tithe of treasure or magic items, or ensorcel the party to send them on some quest. Monsters in their lairs, on the other hand, just wait around for the party to come kill them and take their stuff.

But this is a disservice to monsters, and not entirely true. Monsters with a lair in the area should be included on wilderness encounter tables. Orcs can go out on patrol too, mounted bandits will surely raid any inhabited areas within their sphere of influence, and flying monsters like dragons can control and have influence over the entirety of a small region, even demanding tithes of treasure from settlements within their domain much the same way a stronghold's ruler might.

You could also make a distinction between the two on the basis of their relative locations. Strongholds are generally found in inhabited areas, whereas monster lairs are found in the wilderness. Strongholds are known locations and readily accessible to the players should they have the means to take them, whereas monster lairs are hidden and remote, requiring scouting to find and wilderness travel to reach.

This too does not hold up to scrutiny. More dangerous monsters are generally found in the wilderness, yes, but monsters lair in inhabited areas as well. Humans, demihumans, humanoids, giants, and even vampires can establish their lairs in close proximity to settlements, strongholds, roads, and cultivated farmland. If one of these creatures lairs in a fortified camp or stronghold, its location is likely known, and it is probably well within reach of player characters launching forays from inhabited lands.

You could attribute the difference to logistics. Taking a stronghold requires a small army and siege weapons - but clearing some monster lairs will require these as well. According to the AD&D DMG's Appendix C, a stronghold will have its ruler, up to five henchmen, and as many as 64 men-at-arms led by four low-to-medium level fighters. On the other hand, a gnoll lair can have up to 200 gnolls, 10 leaders, a chieftain, 20 guards, and either three trolls, 16 hyenas, or 12 hyaenodons. Humans, demihumans, and other humanoids also appear in numbers up to the hundreds with leader-types and (often) other monsters in their service. 

These will need to be faced with armies as well - and in many cases, armies which are larger than those needed to deal with strongholds. Facing a monster lair in the form of the aforementioned hobgoblins' fortified village or giants' castle will likewise require the use of siege weapons, just as strongholds do.

The distinction might also be a matter of politics. Strongholds are part of the established order - they are fortified positions in which armies can be mustered, which control a surrounding area and collect revenue from its inhabitants. Powerful people within the social fabric of the campaign will generally care one way or another about what happens to a stronghold, and they will be sticking their noses in the player characters' business if they get wind of a plot to usurp a local ruler.

But that's not to say that people won't care what happens to a monster lair. Goodly folk will likely be appalled if they hear that the party led an army to massacre a village of elves, and a powerful Evil magic-user who rules the local town will be none too pleased to find that the party has slain a manticore with which the magic-user was secretly aligned. Similarly, people's feelings about the fate of a stronghold can go both ways - fury if the party burns down a monastery inhabited by peaceable monks, or jubilation if they overthrow a tyrannical warlord.

The difference is utility. A stronghold is real estate which the player characters may now control. They are probably not going to take up residence in a cave cleared of goblins (and would not be able to raise an army there or collect revenue from it in any case), and might not be able to make much use of a castle proportioned for inhabitation by giants, but a conquered stronghold is a readymade base of operations for the player characters going forward, which will come with both perks as well as obligations.

All of these locations exist on a gradient which is perhaps best viewed through the lens of power, complexity, and player agency. Dungeons are relatively simple environments where players have a high level of agency despite their characters not having much power. There are factions within, but they are small. A local ruler might have opinions about what goes on in the dungeon and may lay claim to what is hauled out of it, but their ability to exert control over it is limited. Players of low level characters by comparison have much less agency in highly complex settlements, where powerful factions and high level NPCs rule, influence, and exert themselves upon those weaker than them.

Strongholds are somewhere in the middle. They are fortified, ruled by high level NPCs, and involved in regional politics, but they are not so formidable that they cannot be overcome by medium level characters, and their position (often) on the borderlands of civilization means that the more powerful rulers of settlements' influence over them will be somewhere between that of nearby dungeons and that of the settlements themselves. However the players address them will require some consideration and planning because they are of medium complexity.

So what then is the role of monster lairs? Because they can vary so much, they fluctuate along the gradient. They are stepping stones from one level of power, complexity, and agency to the next - the owlbear den is probably more dangerous than the dungeon levels the player characters have explored up to that point, but it is not any more complex (it might even be simpler), and the players have considerable agency in addressing it because it is unlikely that anyone will retaliate against them for clearing it out. These lairs give the players the experience (both in terms of experience points as well as accumulated knowledge) to later tackle more challenging locations like strongholds.

The lair of hundreds of organized men, demihumans, or humanoids may in some cases be an even tougher nut to crack than a stronghold ruled by a character-type, but past experience with such a stronghold (and the accumulation of revenue and mustering of an army in such a place, should the characters be able to hold it) will prepare them for the challenge. It is a more complex environment than the owlbear den because the monsters there will be true factions of their own - meaning powerful individuals will have opinions about them, which in turn will limit the players' agency. Like the owlbear den before it which provided the bridge to engaging with stronghold play, this type of lair will be the launchpad into high level play in towns and cities, with the player characters now powerful enough to assert their agency there.

We can summarize the complexity/agency gradient as a semi-linear progression as follows, from low complexity/high agency environments which require only a low level of player character power to engage with to high complexity/low agency ones which require greater player character power: dungeons > simple monster lairs > strongholds > complex monster lairs > settlements. 

There are of course more complex and challenging dungeons (or simply deeper, more dangerous levels of the same dungeons in which play begins), as well as simpler settlements (a thorp and a city are very different environments). We could complicate the progression like so: dungeons/simple settlements (thorps, hamlets, and villages) > simple monster lairs > moderate complexity settlements (towns)/moderate complexity monster lairs (where monsters are intelligent/faction-like but not huge in number (a clan of ogres, for example)/strongholds > complex monster lairs (a clan of 100 orcs, for example) > complex settlements (cities).

This is reflected nicely in my sandbox for B1: In Search of the Unknown. In that post, I anticipate that the players will begin by tackling Quasqueton, then venture into the wilderness to deal with the less complex monster lairs (the leprechaun, the giant eagles, and the owlbears), then graduate to the brigand stronghold and the slightly more complex/challenging ogre lair and adventurer camp, before finally being able to confront the evil illusionist who rules the local town.

None of this is prescriptive - the illusionists' stronghold is right in the middle of town, and the adventurer camp, brigands' castle, and ogre den are nearby. There's nothing stopping the player characters from checking those places out in the very first session, but they will face considerable challenge should they choose to try confronting them. This challenge might be readily apparent to the players, but if I'm conscious of it as well, I can be sure to signpost the danger in rumor tables and interactions with NPCs.

Regardless of how exactly we define it, understanding this progression is key because it informs how campaign play is structured. Player characters begin with a relatively small amount of power and a limited ability to affect the world. They begin play in less complex environments, which they can affect without entangling themselves in overwhelming complications. As they accumulate power, their ability to affect the world grows, allowing them to tackle more challenging environments while being equipped to deal with the complications which arise.

Each of these locations is essential to a standard D&D-like game not only because they provide different gameplay experiences to the players and utilities to the characters, but because each is a stepping stone to the next type of adventure environment, allowing the campaign to evolve in scale alongside the player characters' growing ability to affect the setting.