How to build a good dungeon in 5e

[BRC]

Hi, my name is BRC, and I've got a problem.

I like the idea of dungeon crawls, but I don't know how to build or run them in a way that feels fun outside of some very specific scenarios.

I'll admit this may be somewhat limited to D&D 5e and my particular approach to it, but...


In my mind, a dungeon crawl is a series of rooms. Often, there is a goal at the end that the PC's are pursuing, but there should be joy found in exploring the dungeon itself. Rooms should have both challenges and rewards to make it fun to explore. You can do a series of obstacles on the way to the objective, but that alone doesn't really count as a "Dungeon Crawl" in my book. The crawling itself should be fun.

Problem 1: Rewards
I can understand old-school dungeon crawls, where the "Reward' is gold or other relevant treasure. Newer stuff mixes that in with XP.

But I use milestone leveling, and with 5e's minimal support for actually spending gold past early levels, there isn't much to do with finding little bundles of gold scattered about. handing out consumables and other 'Minor" magic items often get lost in the shuffle. So just scattering gold around doesn't feel right.

I've had some recent success with a dungeon crawl where the PC's were seeking components for things that would be directly usable in their immediate future. They were headed to a haunted house full of ghosts, and so the dungeon before then had moss that could be brewed into potions of necrotic resistance, water that, due to sitting in an area charged with cursed energy could become supercharged holy water, grave-silver that could be used to craft wards against the undead, ect. Consumables that were specifically relevant in the near future so they felt good to get and use. My players remarked that getting the moss felt really good because they knew those potions would be useful soon.

But that isn't especially generalizable.That dungeon seemed to be a lot of fun for my Players, but I don't know how to pull off that trick again without setting up the same scenario of "Okay, you've got THIS dungeon, and as part of clearing it you can find things that will be useful in the NEXT dungeon".

Problem 2: Challenges

Combat in D&D takes a long time at mid-high levels, especially if you want the enemies to represent an actual threat to the PC's. There's enough overhead in managing a combat that a "small" fight is often just not worth running at all. The fighter attacks a bit, the wizard maybe throws a spell slot, ect. If the enemy isn't a major threat, it's often not worth throwing resouces at the fight. But Big, consequential fights take too long, and the idea of danger being constant is part of what makes things a dungeon crawl, rather than just "Here's the two rooms with Fights in them".

My current approach is mostly to use "Glass cannon" enemies, low-health, high damage, so the fight can be over quickly and the challenge can be less about "Will this enemy defeat us" and more "How many resources will we lose to this thing before we kill it", often with the enemies specifically designed to ambush. Like above, I just ran a dungeon that I felt made sense to have these sorts of psedo-trap enemies, it was a ghoul warren and so the ghouls could hide in burrows in the walls and floor and then jump out in ambush, get a hit off, and then go down to the PC's (or not, and get another hit). It seemed to work pretty well. But I don't know how to generalize the lessons I learned there to dungeons that didn't fit that very specific theming of "Enemies that don't mind hiding in a hole in the wall for somebody to run by, and don't have enough self-preservation to not attack a fully geared party of adventurers by themselves". Any group big enough to actually present a threat becomes a slog of a fight that kills the momentum of the table, any group that doesn't present a threat is rarely worth the table time to fight, and often won't make narrative sense to attack anyway.

[sithlordnergal]

---Rewards---

Give them minor magical items, mostly useful consumables. Things like potions are great, but also give them ammunition. Magical arrows and crossbow bolts are great, and add in special consumable spell components. Also, add in single use magical thrown weapons. This is in addition to the regular gold they can get. I actually enjoy making a table of random treasure rewards that the players roll on when opening a chest.


---Challenges---

Obviously you want to use some combat, but I like to sprinkle a healthy amount of traps and puzzles into my dungeons. These can use up resources as effectively as combat, sometimes even more so. It can also be pretty stressful for the players. For example, I set up a trap that, when activated, trapped players in a solid Forcecage, and filled the Forcecage with Sickening Radiance. A player was trapped, and they had to figure out how to shut down the trap before the trapped player died.

Another trap that I used involved Prismatic Wall and Ropers that dragged players through the Prismatic Wall. HOWEVER!!! I caution against using this trap unless you have given players a way to get through every single layer of the wall. Otherwise the players will be forced to tank the wall, and that's not a good way to deal with a trap.

[BRC]

---Rewards---

Give them minor magical items, mostly useful consumables. Things like potions are great, but also give them ammunition. Magical arrows and crossbow bolts are great, and add in special consumable spell components. Also, add in single use magical thrown weapons. This is in addition to the regular gold they can get. I actually enjoy making a table of random treasure rewards that the players roll on when opening a chest.

The issue is my campaigns tend to go pretty long. As a result, each player accumulates a massive list of consumables that they then forget about.

With my recent dungeon, where the consumables were focused towards their next goal they got excited about picking them up. Normally it's like "oh, that's neat" and then chuck it on the pile.

There's also the issue with Narrative. There's lots of reasons why sellable loot might be scattered around a place, but for magic items, I can only stretch things so far narratively to have random adventurer-useful consumable magic items scattered about.

---Challenges---

Obviously you want to use some combat, but I like to sprinkle a healthy amount of traps and puzzles into my dungeons. These can use up resources as effectively as combat, sometimes even more so. It can also be pretty stressful for the players. For example, I set up a trap that, when activated, trapped players in a solid Forcecage, and filled the Forcecage with Sickening Radiance. A player was trapped, and they had to figure out how to shut down the trap before the trapped player died.

Another trap that I used involved Prismatic Wall and Ropers that dragged players through the Prismatic Wall. HOWEVER!!! I caution against using this trap unless you have given players a way to get through every single layer of the wall. Otherwise the players will be forced to tank the wall, and that's not a good way to deal with a trap.

I do use traps, but I don't love them on principle. big setpiece traps are cool, but lots of "okay roll to see if you spot the trap. Okay roll to disarm it" isn't especially engaging for me. Traps often don't feel like a challenge as they are a punishment for exploration. Even a short fight gives the players opportunities to make choices. Traps can either fall into the category of "This is just a random obstacle that will take a lot of table time to resolve" or "Roll some dice, take some damage".


I guess I could play more with non-hidden traps that can allow the Players a chance to think of cool solutions using their toolkits rather than just throwing the disarm at it.

Although the biggest issue I have is usually narrative. Lots of traps make sense if the dungeon is a vault or ancient tomb, someplace nobody is supposed to get without a lot of work. But if the Dungeon is a place where people live and work, they're not going to have active traps all over the place. Memorizing which stones not to step on only gets you so far.

[BRC]

I keep talking about this "Dungeon I just ran", I guess I should clarify.


My PC's are currently on their way to deal with a haunted manor house occupied by a very powerful ghost. The haunting means the house is full of ghosts actually, not just the one they are trying to exorsize.

The whole area is cursed so there's lots of undead. Prior to cursing it was a major burial ground (Pre-curse, the place was actually blessed so nobody buried there would come back as undead, so there's a whole necropolis graveyard with catacombs). The house itself is difficult to access due to the Ghost blocking it off with barriers, but they learned that the inhabitants of a Ghoul Warren had managed to secure a tunnel into the house.

So the PC's went to the Ghoul Warren in Gallow's Hill, where historically criminals would be hastily buried after execution. Now, all the criminals buried there were rising with the curse, full of malice and hatred.

I established that, due to the curse-magic that suffused gallows hill, certain things in that area gained unique properties. Water that had been sitting in Gallows Hill absorbed the curse magic, when that water was blessed it reversed and become incredibly potent holy water. Fungus growing under Gallows hill could be turned into potions of necrotic resistance, or burned to create a miasma that distracted undead. Grave-Silver that had been marinating in the curse could similarly be blessed and turned into potent tools against the undead. Powdered wizard bones could be used to force ghosts into corporeality. None of these were things the undead inhabitants of the hill especially cared for or recognized as valuable, so the PC's could find them scattered about the burial chambers and catacombs of Gallows Hill.

Meanwhile, the inhabitants themselves, mostly maddened ghouls wrangled into direction by the more powerful of them, and mindless vengeful specters, were pretty ideal as dungeon-crawl enemies. The Ghosts functioned more as living traps than anything, jumping out when the PC's poked around on things. The Ghouls lay in ambush but gleefully threw themselves at the PC's in the hopes of tasting living flesh. (note, I hate D&D ghoul's paralysis theme, finding it unfun. These were different statblocks) They hit hard, but were fragile enough that combat in the narrow tunnels never devolved into a slow slugfest. The Traps were either triggered by, say, a skeleton buried in the catacomb walls with a trigger attached to it's jawbone, or something the ghouls would find inconvenient but not dehabilitating.

Plus, the layout of the place (A catacomb), meant that it was easy to have that classic dungeon crawl feel of lots of twisty passages and little rooms, rather than building a living, functional space that also serves as a dungeon.

The raid on Gallows Hill went quite well in my estimation, the PC's were enjoying scouting about, dealing with traps and ambushes, and recovering the variety of little loot-packets scattered around, knowing that it would all be useful, not in the nebulous future, but when dealing with the haunted house. Part of this thread is me seeing what worked there and trying to figure out how to generalize it beyond the specific narrative circumstances found there. I can't just keep running Ghoul Warrens and Haunted Houses forever.

[clash]

Let me preface this by saying I generally run a series of smaller dungeons and past of a general campaign, that being said the same principles apply I think.

In my current game many of the monsters are not humanoid and have no reason to have usable lot on them. My simple solution to this is adding a in game currency of sorts collected off monster corpses. Call it magic dust. So monsters always have at least money loot and it fits the narrative.

Second, I introduced an entire crafting system where weapons, magic items, clothes, gadgets etc can be crafted during a long rest, with the appropriate training, but they all require money to do so. That alone keeps my players pretty perpetually excited about money rewards and unfortunately less excited about other rewards 😂

[Sorinth]

Having loot that the players want is for sure going to provide more excitement then random loot. So yes having potions of resistance prior to a threat where they will come in very handy is good but can be overplayed if it's a linear, however you can always make them more side quest things. So instead of picking up the super holy water being along the main adventure path if they got the info that by taking a 2-3 day detour then it's going to be inherently more rewarding because it was a player choice.

By providing more ingredient/recipe style rewards feel like they might work well for you. You can be more generous with the amount they find since they still need to go do an adventure/dungeon before they get the item and by letting the players choose whether to pursue the magic item they will be more invested. So if you hand out 5 magic item recipes even if they basically ignore 4 of the 5, they will still find the whole thing rewarding because they actually want the one they chose to pursue. And perhaps in some as of yet unplanned adventure one of the 4 they ignored will be a good fit and now instead of handing out the item you can provide one of the two ingredients which reminds them of the recipe and gives them the option of preparing properly by finding the last ingredient somewhere along the way.

Also if the players have nothing to spend money on then why not just change that? Like I get the comment that there's minimal support for spending money past the early levels, but why is that stopping you from providing an outlet for spending gold? Your the DM, if you want players to gather gold and spend it on buying bigger/better magic items then allow them to do that, give the players downtime and ways to spend money on training. Perhaps the Bard knowing that many spells rely on having a common language will want to hire a tutor to teach them a new language, it costs them 2000 gold and their downtime but now they've gained proficiency in that language. Maybe they save some random NPC chef who they can then hire to accompany them on their adventures, it costs 50gold a day + lodgings, this NPC won't go into dungeons with them but will stay in camp/town and they provide access to the Chef feat.

There's plenty of ways to make money useful even for higher level PCs.


For your combat, yes fights can take lots of time, this usually means dungeons will take more then 1 session to complete is that an actual problem for you? Like for sure in games where the players aren't consistently there then you want each dungeon to be completed in the 1 session which means smaller dungeons and not a lot of fights, but if it's a regular group is it actually a problem?

[Mastikator]

Consumables

If players aren't using consumables then the game isn't hard enough. IMX players will use healing potions when they need them, and save them when they don't. Some consumables are only useful outside of combat as prep, so let them prep.

Give powerful monsters consumables that buff them further into the stratosphere. If the players are facing a medusa and her gorgon mount then give her two potions of speed, and the first thing she does in combat is to drink one of them. They then loot a potion of speed of her. Also maybe give her two or three greater healing potions, and make her drink one in combat.

Magic items

Sometimes magic items aren't useful to anyone in the group. So make sure that the largest, most prosperous city in your campaign has an auction house or museum or university that is willing to buy and sell magic items. Be sure to not treat magic item shopping like mundane item shopping. Have a master smith of legend, (and/or a powerful smithing guild) who the players can commission to craft +1 weapons and armors. Both Xanathar's and the DMG has guidelines for how much they should cost, the players should at best get a 50% resell value.

I think the rule of "there shouldn't be magic shops" philosophy isn't good for the game, and is not actually recommended by the DMG. If buying arrows is like going to the supermarket and buying milk, then buying a set of winged boots is like buying a Monet painting. You can do it if you have the money and connections. The players acquire the gold in the dungeons, and the connections from the quest givers.

Challenge

This ties back to consumables. I think glass cannons is the right approach for mid/high level, but you either need more combats between long rests so that they have to tap into their consumables, or harder combats so they have to prep for them by using consumables. Instead of 1 fire giant and it fire elemental friend vs the party of 4 level 9s, make them face 1 fire giants, 1 fire elementals and their 10 magna mephits (all who rush in ready to explode in the PC's faces) but they get to prep and strategize. Now those potions of fire resistance get drunk.

[Sorinth]

Consumables

If players aren't using consumables then the game isn't hard enough. IMX players will use healing potions when they need them, and save them when they don't. Some consumables are only useful outside of combat as prep, so let them prep.

Give powerful monsters consumables that buff them further into the stratosphere. If the players are facing a medusa and her gorgon mount then give her two potions of speed, and the first thing she does in combat is to drink one of them. They then loot a potion of speed of her. Also maybe give her two or three greater healing potions, and make her drink one in combat.

Magic items

Sometimes magic items aren't useful to anyone in the group. So make sure that the largest, most prosperous city in your campaign has an auction house or museum or university that is willing to buy and sell magic items. Be sure to not treat magic item shopping like mundane item shopping. Have a master smith of legend, (and/or a powerful smithing guild) who the players can commission to craft +1 weapons and armors. Both Xanathar's and the DMG has guidelines for how much they should cost, the players should at best get a 50% resell value.

I think the rule of "there shouldn't be magic shops" philosophy isn't good for the game, and is not actually recommended by the DMG. If buying arrows is like going to the supermarket and buying milk, then buying a set of winged boots is like buying a Monet painting. You can do it if you have the money and connections. The players acquire the gold in the dungeons, and the connections from the quest givers.

Challenge

This ties back to consumables. I think glass cannons is the right approach for mid/high level, but you either need more combats between long rests so that they have to tap into their consumables, or harder combats so they have to prep for them by using consumables. Instead of 1 fire giant and it fire elemental friend vs the party of 4 level 9s, make them face 1 fire giants, 1 fire elementals and their 10 magna mephits (all who rush in ready to explode in the PC's faces) but they get to prep and strategize. Now those potions of fire resistance get drunk.

I think a big part of the problem with potions is that it takes an action to use, so sure you can prep but once in combat it's almost never worth it. I think BG3 showed it was much better by changing that to a BA.

In terms of magic item shops, it depends a lot on the campaign world. It can make sense for common/uncommon magic items to be easily bought or sold. But generally I agree that buying and selling most items should be more of an event, and often it won't be a gold transaction but a trade of some sort, land or a business like a mine in particular can be good options for the DM to use when buying a magic item from the PCs. But I think a magic mart is preferable to collecting gold and having nothing to spend it on.

[aimlessPolymath]

Rewards

There are a few angles I can see taking with this:

-You could create new gold sinks- not necessarily lifestyle costs, but maybe opportunities where a challenge could be easily bypassed by spending X gold (spending gold to throw a big party; opulent gifts for officials to sway them; hiring people to go off and do things for the party, etc.). If you try lifestyle costs and purely narrative effects of lifestyle don't catch the player's attention, you could tie minor mechanical benefits to a sufficiently high lifestyle (having 1-3 more Hit Dice per day if you spend X/Y/Z gold per month?).

-Rather than relying on gold, you could add different types of loot. If the players have a base, offer nice-looking furniture to furnish it with (my Shadowrun game has seen players steal two widescreen TV sets so far). Art and gems can be more 'interesting' than pure gold. Perhaps special crystals which could be spent to perform big magic rituals? I could definitely see a situation where players require 20 [reagents] to perform a big ritual spell, and enter a dungeon where those reagents can be found as loot- in that situation, the reagents effectively replace gold as the driver to search deeper, and they could return once they found enough (or if they take extra, upgrade the ritual)

-Provide pseudo-magic-items with vaguely defined properties, and allow players to use them later as macguffins. Instead of a flask of dragonblood being treated as a potion with a defined effect, you can tell the players "This flask does fire-related things. If you need a source of fire magic, you can use it with a skill check to make something happen."

-One of the big benefits of gold is that it's fungible- you can provide 100 gold in 1 place, or 10 gold in 10 different places, so even small amounts are valuable. One odd option would be to replicate this fungibility by having groups of items which, when brought together, have significantly greater value- like a set of gems which have been scattered, but form a magic item when all collected.

Consumables
I agree with the idea that most aren't worth an action in combat- there's also something of an issue where if players have a big pile of consumables, it may not be clear when they need to use the items to succeed. This is particularly an issue with the classic approach of 'give out a large number of low-impact consumables', where any given consumable may only have a small impact on a fight. I think a smaller number of powerful consumables would end up getting used more.

IMX, consumables get used the most in a few specific ways:
-To extend the effective adventuring day. This happens when players are forced to keep going (and can't rest) but are running on fumes, so they see that they need to draw on extra resources.
-When the consumable does a really interesting thing, providing an entirely new option that's otherwise unavailable (Using invisibility to sneak into someone's home and spy on them) Interestingly, I think this happens more the weirder the consumable is, because it's more likely to let the party try an approach they otherwise couldn't.
-During a tough-looking fight, if the consumable is very powerful relative to party options (using a scroll of fireball to blow up a crowd of enemies when the party does not otherwise have access to fireball).
-Prior to a climactic story fight that's expected to be tough, if the consumable provides an ongoing buff for the fight.
-If the consumable is obviously well suited to the fight, and it's very clear that the party will have a lot of trouble without using the item (ex. potion of see invisibility vs. invisible foe). This needs to be fairly obvious, because I think a lot of fights are designed to be beatable without spending items.

I rarely see consumables used in winnable fights, or (oddly) when the party is caught off-guard or surprised. I don't really see consumables used to extend the adventuring day or proactively conserve resources before they're low (i.e. I don't really see people quaffing a damage buff potion before a fight so that they defeat monsters faster and save resources on healing- the potion gets saved 'for later').

Combat Encounters & Challenge:
This is rough, yeah. I honestly do like the 'glass cannon' approach for short-duration fights designed to drain resources, and I think it's probably as close as you can get to the old-school dungeon crawling paradigm. Some other ideas:
-As a variant of the 'trap' enemy, hit-and-run enemies that deal some damage, then retreat to strike again later.
-"Glass Knives" who apply pressure by inflicting longer-lasting effects, like draining spell slots, inflicting exhaustion, causing disease, or other long-lasting conditions.
-The classic answer of applying time pressure or environmental threats to reduce the frequency of rests.

Re: traps in particular, I tend to imagine most traps in places where people live acting in two main phases:
-The majority of the traps are inactive 99% of the time, or otherwise harmless to defenders (ex. a trap which releases a monster that is trained not to attack the defending side, a blade which strikes high enough to decapitate Medium creatures but doesn't hurt the Small inhabitants).
-There are a small number of traps which, when activated, alert everyone to turn on the inactive traps (or turn them on automatically).

Most of these traps would be designed to either split up invaders and place them in a bad spot, or drain their resources- they aren't necessarily lethal, but they put them in a position where the defenders have a significant advantage. Think of medieval murder holes- they aren't a 'trap' in the D&D sense, but they give the defenders a position of overwhelming advantage over the attackers by pouring oil or firing arrows from a position of safety.

[BRC]

Rewards

There are a few angles I can see taking with this:

-You could create new gold sinks- not necessarily lifestyle costs, but maybe opportunities where a challenge could be easily bypassed by spending X gold (spending gold to throw a big party; opulent gifts for officials to sway them; hiring people to go off and do things for the party, etc.). If you try lifestyle costs and purely narrative effects of lifestyle don't catch the player's attention, you could tie minor mechanical benefits to a sufficiently high lifestyle (having 1-3 more Hit Dice per day if you spend X/Y/Z gold per month?).

-Rather than relying on gold, you could add different types of loot. If the players have a base, offer nice-looking furniture to furnish it with (my Shadowrun game has seen players steal two widescreen TV sets so far). Perhaps special crystals which could be spent to perform big magic rituals? I could definitely see a situation where players require 20 [reagents] to perform a big ritual spell, and enter a dungeon where those reagents can be found as loot- in that situation, the reagents effectively replace gold as the driver to search deeper, and they could return once they found enough (or if they take extra, upgrade the ritual)

-Provide pseudo-magic-items with vaguely defined properties, and allow players to use them later as macguffins. Instead of a flask of dragonblood being treated as a potion with a defined effect, you can tell the players "This flask does fire-related things. If you need a source of fire magic, you can use it with a skill check to make something happen."

-One of the big benefits of gold is that it's fungible- you can provide 100 gold in 1 place, or 10 gold in 10 different places, so even small amounts are valuable. One odd option would be to replicate this fungibility by having groups of items which, when brought together, have significantly greater value- like a set of gems which have been scattered, but form a magic item when all collected.

I could. I guess this requires me to set that up at the campaign level, which I've failed to do. I have occasional opportunities to spend gold, but the ranges from the DMG on how much a magic item should cost are pretty wide. As a result gold loot has mostly been an afterthought. My players care more about narrative rewards than statistical ones for the most part (They're the sort to spend lots of gold to secure food and housing for random NPC refugees for example), so it's been fine. Occasionally I'll give them a treasure hoard or make up how much something costs, but it's not generally super relevant. It does mean I can't fall back on that as a game mechanic for situations like this where I want to hand out a small yet relevant reward.


I guess if I want to run dungeon crawls more frequently, I need to build out a campaign-level infrastructure to handle fungible or semi-fungible rewards like this.

[JellyPooga]

One of the joys of a dungeon is that it's, well, a dungeon. It's not a city to explore with folk to talk to and laws to abide, nor a wilderness to navigate to horizon's end, but a discreet and limited area within which, largely speaking, the only civilisation is that which you, as a character, bring with you.

(1) Being discrete means there's a limit to what can be explored, so exploring all the things becomes a reward in and of itself. That's not to say that the players are going to enjoy finding every last dead-end necessarily, but it does mean that they should have fun just finding out how far that blank space goes. Cities and towns are too big and the wilderness is bigger still and all of them are filled with an awful lot of nothing interesting (trees, residences, plebs, etc.), but dungeons...dungeons have lots of interesting things like monsters and treasures, so player want to find secret doors and extra levels that let them carry on exploring and seeing what interesting thing is around the next proverbial (or literal) corner, even if it means heading back to town and coming back another day. Rewards don't have to be gold or items or xp...it can be a reward to simply have more dungeon to explore.

Giving a dungeon an end boss or goal gives it purpose, but every good dungeon has a figurative door or portal that asks the PC's if they want to see more; the passage into the Underdark, the portal to another Plane, the trapdoor or ladder to the next level, the subterranean lake that leads to the underwater temple and so on and so forth. Make the transition obvious, to keep "completing" the dungeon discrete and achievable, but dangle the carrot of possible future adventure, with commensurate challenges and rewards. You can also offer plot hooks as rewards; whether it be a dying prisoner with a quest to offer, a fragment of an artefact begging to be completed or some snippet of legend or information that leads the players to the next adventure...offering the players the choice of what they do next is a mighty reward indeed.

---

(2) Being outside of civilisation means the characters bring their own sense of justice/morality and part of doing a dungeon delve is also exploring the extremes of the characters' personality traits and how they react under pressure. Another significant reward you can offer is not only diversity of encounters (" Gee. What fun. Another mook-squad to fight "), but encounters that challenge not only the characters stats and abilities on paper, but their actual character.

A bunch of combat encounters and traps and puzzles is great for a couple of dungeons, but it doesn't actually engage who the characters doing all this stuff are. When you start throwing some moral decisions into the mix and test more than their atrack bonus or skill modifiers, forcing the players to make some hard choices...that can also be a reward, because it reveals and explores character traits and leads to character growth and development. Test their loyalty, their greed or their compassion, for example. Play the devil's advocate, pass notes, be secretive, see how far you can push the player as much as the character. Get them engaged with who theor character is now and ask them who their character is going to be after they get back to civilisation.

This engagement with the player's characters is as much a reward as, if not more than, any magic sword. In the dungeon is also the best place for it, because it isn't overshadowed by such things as laws or just wanting to handwave travel from point A to point B.

[Mastikator]

I think a big part of the problem with potions is that it takes an action to use, so sure you can prep but once in combat it's almost never worth it. I think BG3 showed it was much better by changing that to a BA.

In terms of magic item shops, it depends a lot on the campaign world. It can make sense for common/uncommon magic items to be easily bought or sold. But generally I agree that buying and selling most items should be more of an event, and often it won't be a gold transaction but a trade of some sort, land or a business like a mine in particular can be good options for the DM to use when buying a magic item from the PCs. But I think a magic mart is preferable to collecting gold and having nothing to spend it on.

Indeed, it also made some potions last until long rest and made them mutually exclusive. A potion of fire resistance lasts until you finish a long rest but if you drink a potion of lightning resistance the fire resistance ends. It's a decent balancing factor, you have no incentive to stack them up and then use them all at once, instead you have an incentive to continuously use a few as you need them.

[Easy e]

@Jellypooga- Point 2 is an excellent point and rarely gets used properly. In theory.

In practice, it just allows Players to indulge in their worst murder-hobo-"isms" in a (relatively) consequence free environment.

[JellyPooga]

@Jellypooga- Point 2 is an excellent point and rarely gets used properly. In theory.

In practice, it just allows Players to indulge in their worst murder-hobo-"isms" in a (relatively) consequence free environment.

It does depend on the kind of game you run, granted. If the game is little more than a series of tactical skirmishes connected by loose narrative, then offering character development in lieu of more material rewards could fall flat in the players' eyes. As GM, however, it is your duty to set the tone of the game as much as it is the pace, story and encounters, which to my mind includes how the world reacts to such things as "murder-hobo-ism"...what you allow the players to get away with will be reflected in how they react to the setting/campaign you present them, in much the same way that players might react differently to a gothic horror setting vs. a magipunk mystery setting.

[rel]

Some Background
5e, like most iterations of D&D works best when the PC's are navigating a dungeon overcoming challenges and liberating loot from the inhabitants.
I find the system struggles if the party isn't doing some variation on that theme as the minute to minute gameplay loop.

That said, I don't think 5e is a great fit for an OSR style delve into a megadungeon, counting torches and mapping the twisting halls on grid paper to avoid becoming hopelessly lost in the sprawling dark.

And no iteration of D&D works well with conventional dungeoneering gameplay continuing all the way to level 20. The basic gameplay loop of punching monsters, bypassing obstacles, and securing loot doesn't change, but the trappings around it need to.
At some point the PC's need to hang up the 10-foot pole and pitons and stop shuffling through actual 10-foot wide corridors to confront a monster guarding a pie.

The Advice
A solid foundation for a 5e dungeon crawl (or adventure in general) is the adventuring day. Six to 8 challenges with a few short rests breaking things up.
Thus the first question should be why can't the PC's run off to long rest at the local tavern every 5 minutes, and why they can take a few hour long breaks throughout their delve.

Pacing established, you need to come up with some obstacles the PC's are likely to face, then re-imagine them in terms of interesting puzzles, traps, monsters and obstacles of appropriate difficulty.
If said re-imaginings make sense, then great, but always prioritise interesting gameplay over concerns like realism.
If you can keep things internally consistent and believable then great, but amazing verisimilitude is unlikely to carry the gaming session if the actual encounters are boring.

Now comes time to arrange the encounters (sometimes it's easier to sort out arrangement before setting out encounters), 5e has very minimal rules for exploration, so drawing a full map isn't necessary, a simple flowchart is fine unless your gaming group really likes the old ways.
Either way, multiple paths, hidden areas and enough foreshadowing to allow the players to make meaningful choices about what to tackle and in which order are all but essential.

In terms of loot, making custom magical items is really easy in 5e, so those are the easiest way to go.
A player should be able to track 2 to 3 interesting magical items, so a party of 4 can be given about 10 before they start getting forgotten. If you level every adventure or 2, and get 1 such item per adventure, that approach can get you through most of a campaign.
As an added bonus, custom items can be used to increase the effectiveness of those PC's that were built poorly or are muggles.

Other interesting loot options are plot hooks for new sidequests, background info on the world, allies and followers, minor powerups like more HP, low light vision, magicl rituals or the ability to cast a cantrip.

[JonBeowulf]

Regarding the relative uselessness of gp in higher-level play, one of the things lost from the earlier editions is rules for PCs building strongholds and dominions. They were crazy expensive construction projects that gave players something their character was building for and justification for hoarding as much coin as possible. They provided a reason their characters were doing all that dangerous stuff, beyond "because it's game night". They created innumerable plot hooks, let the players add something permanent to the game world, and required as much gp as the player wanted to spend making it into whatever they wanted.

[tchntm43]

I suggest studying the adventures in Tales of the Yawning Portal, as these are all "dungeon" based adventures, some of them with quite long and extensive dungeons. They are all very fun to play and you can learn a lot about how to make every room and corridor unique and fun.