The great news is you already know what you want to achieve with your one-shot. Teach them the mechanics of both how the game works, and how you will run it. From there, everything else flows.
World building is for campaigns. So lets instead talk room design.
So first of all, make their characters for them. Let them decide who they will play - literally put the sheets in the middle of the table at the start of the game and say ‘pick a character’. Or describe them and see who gets excited. The reason to do this is so you know what their capabilities are. Make their skill sets diverse - the fighter should not have a sideline in Rogue skills unless your group is so small as to not have a rogue. If there’s no wizard, give the cleric a sideline in arcane studies but don’t have a PC tread on another PC’s area of expertise.
Everything on that character sheet is a tool for you to use.
Envision your game as a series of rooms. Maybe they’re linear, room 1 leads to room 2, leads to room 3 and so on. There is nothing wrong with this. It requires substantially less work on your part which means more time spent on the individual rooms. With more experience you can make some of them branch - maybe they can go backwards and not just forwards, or move between a set of parallel rooms before reaching a point of no return.
Room is abstract. Room is a challenge, an encounter, any situation where failure is both a possibility and where doing so will have meaningful consequences. Room is not a skill check to get some information about another room - but room can be a tense negotiation.
You set the tone of your game. Especially when you’re new, it can lead to quite a cautious, languid pace as you nervously relate the contents of your notes and try to remember all the things you need to remember. That means your players will become quite relaxed and treat things as if they have all the time in the world.
Don’t give them that luxury. Run them ragged. Introduce timers.
A timer isn’t a ‘do it or die’ its a ‘hurry the fuck up, you don’t have time to discuss this in a committee’. Lets take a typical example of a 1st level dungeon room. The PC’s are being swarmed by kobolds crawling out of tunnels in the rock - the group they’re fighting at the start of the room isn’t a major threat, maybe one or two kobolds per player. But…in 1d6 rounds, 2d6 MORE kobolds will arrive. In 1d6 rounds after that, ANOTHER 2d6 kobolds will arrive. Make the timer visible. Get a big dice and roll it in front of them. Tell them ‘you can hear the approach of reinforcements in the tunnels, and there are a LOT of them’ and then count it down for them as the rounds go on.
This makes the encounter time sensitive. They’ll be more concerned with solving the challenge you put in front of them to survive it and advance than with brute forcing the situation because they flatly cannot do so. Your kobolds are endless, their resources are not.
Each room should have a challenge, an objective. The timer in the example room above forces the PC’s not to dick around, or suffer the consequences. So what is it you want them to hurry up and do? Well, lets look at our characters’ sheets. Lets say we’ve got Bob the Fighter whose a fuckhueg tank, Sheila the Rogue whose a swashbuckling trap master, John the Cleric who we’re going to give a wand of cure wounds spells too, because playing the party bandaid box is the least fun experience ever, and Kara the Wizard who has a deep well of useful information you can feed the party through her knowledge skills.
So lets say that at the end of our room there is a door. The objective is simple - get through it before the kobolds overwhelm the party. What’s stopping them? Lets say this room is one where Sheila gets to show off. So there’re a couple of problems. First of all, there’s a chasm between the party and the door. Lets add a rope Sheila can use to swing across. Lets also say that to ensure only one person does this, and therefore its most likely to be Sheila, that the thing the rope is anchored too is not that stable. When she does swing across, it breaks. Nobody else can do that now, but Sheila is on the right side.
Since we’ve taken away the only route across the chasm, we need to add something she can do to get the rest of the party across. So lets add a draw bridge, which was raised. Sheila needs to lower it so that the party can get across. Maybe the mechanism has been disassembled, and she needs to find a key part - this lets her get some real use out of her perception. When she passes, she finds the mechanism in a pile of refuse. If you want to complicate the situation still further maybe its difficult to reach - climbing could be required. Perhaps its on a ledge in the chasm itself. This is also a way to get the other PC’s involved in the challenge - because teamwork. Tell Sheila spots the mechanism and that she could try to climb down to it and back up, but it’d a lot faster to find some way to bring it to her. Hopefully she calls out to Kara ‘Can you levitate that *points* to me? I need it to lower the draw bridge!’. Kara levitates the mechanism to Sheila. They now feel like a well oiled machine of badassery and all you’ve done is have one of them jump across a chasm, roll a perception check and cast a spell. Bob and John are amusing themselves with kobold squishing.
The mechanism restored, Sheila lowers the draw bridge, the players retreat across it.
Key Concept: Interactive Environment
Make shit in your rooms do things. This is doubly useful for starting out players because odds are you don’t own a ton of dungeon tiles or terrain or anything like that. And you’re probably not a digital artist who will render lovingly designed maps with oodles of detail. But you don’t need too!
Less is more. If you put a hulking great pillar in the middle of an otherwise empty room, it draws the eye. So use that. Make it part of the encounter.
So lets say in this room the only bits of terrain OTHER than the ground on which the PC’s are standing on and trying to reach is this drawbridge. Right away this tells them that they need to be focused on this drawbridge. We already know that solving the ‘missing mechanism’ challenge is the key to the room, and that our rope swing was telegraphed by being the only OTHER specifically represented/mentioned thing.
So our party is now on the right side of the chasm, but the draw bridge can also be used by the kobolds! This is only half over! So lets make the drawbridge a bit less sturdy. Maybe its frame is rotted, and it wouldn’t take too much to break it - sending the bridge tumbling into the chasm. Time for John and Bob to go to work!
For added tension, bring in a monster when your timer hits zero this time. Something big. Something they have no way of defeating normally - choose something appropriate to the encounter. If they’re fighting evil guards maybe they have an ogre mercenary, if its the kobolds again perhaps an umberhulk. Either way, if it gets to them they’re mincemeat. They’re new to the game, so TELL them this. ‘You’ve heard stories of these creatures.’ (Roll knowledge for more information about its actual capabilities) ‘If it reaches you, there is no way you could defeat it!’
If they don’t immediately seize on the bridge idea, give it a couple of rounds and then do a perception check for Sheila every round until she figures it out. Then tell her, or pass her a note ‘You notice/recall the sorry state of the drawbridge supports. They could be destroyed, sending the bridge into the chasm’. so that Sheila is still the focus of this room, even though everyone’s had a chance to contribute.
They break the drawbridge, they’ve learned to use the environment as well as their basic abilities and they’re free to move on to room 2.
The interactive environment can be changed. Perhaps there’s a trap that no longer works - but a skilled rogue could turn back on. That marauding troll Bob and John are barely holding off starts eats 4d6 of fire damage every turn. Maybe there’s a magical obelisk that used to be used by wizards for defense. Kara can power it up by sacrificing a spell slot and zap the big bad monster with it.
This basic formula works for all kinds of rooms. You’ve got a challenge they need to solve, you’ve got ways they can use the room and their abilities to do it, and you’ve got something bad that happens if they waste too much time.
D&D is a heroic game. It strongly weights ‘appropriate’ encounters in favor of the party. Largely because if they die in a ‘challenging’ encounter your game is over. Keep that in mind. But don’t let it rule you.
There’re three ways, in the above approach to room design to make things easier or more difficult.
Shorten the Timer - This means they afford fewer ‘wasted’ rounds. Failed checks will mean more, time spent not dealing with the central challenge will be punished more severely
Increase the actual difficulty - I like to keep difficulties uniform for my rooms. On the understanding that a 3rd level character will have a maximum bonus to their skill of X regardless of what it is, the challenge should be surmountable for them. If a wooden door has a DC 15 to pick the lock, but 30 to break it down because the rules say so, breaking doors down becomes a high level activity. Unless the higher one is radically faster, this is completely arbitrary. Id rather give them the option, because perhaps Sheila is doing something OTHER than picking the lock, and Bob is the one who realises they need to get through this door and uses his boot to accomplish the job. Use their bonuses to set difficulties. If your party has a +5 to their skill checks, then an average result is 15. They’ll typically spend about two rounds on each DC 15 checks - some might happen in one, others might take three or more that’s dice for you. But on average. So if you want to make it more difficult, make the difficulty 18. Or even 20. They’ll have a 35 or even 25 chance of passing these checks, they could still do it in one round but it could easily take four or more.
Make the timer consequences more severe. 2d6 kobolds is an average result of 7 kobolds but as few as 2 and as many as 12. I recommend against increasing the numbers - the action economy in D&D is the most powerful tool for difficulty that exists and it will make your turns long as you roll lots of dice. Instead, beef them up. Maybe these kobolds have magical arrows that deal additional fire damage. Maybe they bring a shaman with them that casts a few buff spells. Maybe kobolds aren’t the only things down here…
Key Concept: Action Economy (also, boss fights)
This refers to the number of actions a character can take in a turn. Controlling the number of actions taken is hugely powerful in D&D. Its why Haste and Slow are some of the best spells, and why boss monsters are such a minefield.
The game turns on randomisation. If you roll twenty dice you could get 20 hits, you could get 20 misses. Bonuses vs difficulty will tilt things one way or another but the game is still largely random. That means rolling more dice is pretty much always better than rolling fewer.
A big boss monster is common in video games but it doesn’t work well in D&D because of the action economy. A single creature still only has one of each action a turn. Haste it, and it has one extra. The party has one of every action. Each. This is a one sided fight, in which your PC’s will employ the crudest form of rope-a-dope and tank and spank your boss.
The traditional solutions are to add more HP, or make it more deadly. Neither are good solutions. Chewing through a massive pool of HP is boring. If the outcome is basically certain because y’know, action economy, then this is just a dice rolling exercise. Its going down, there’s no real suspense, it might make a cool cut scene but actually doing this in a video game would be a chore. Once you broke the back of the problem you’d be grateful if the game just went ‘Well done, you got this’ and showed you the cinematic of the fight rather than making you ‘press x now’.
MOAR DAMAGE is equally lame. Because randomisation will get you either no appreciable difference to your players, because they won’t be hit enough to really notice it, or will end up killing them because they cannot keep up with the spikes.
If you like boss fights (and I do) then don’t have it be one thing. Have adds. They become a sort of timer. The green forest dragon might have a breath weapon but unless he uses it every turn (and that itself gets boring after a while, I mean yes its optimal for the dragon to stay airborne where its well protected and use its breath weapon but…boring encounter) he’s still limited by action economy.
So what if he has servants? Maybe kobolds. Maybe some kind of bog-pygmy that just happens to use the stats for kobolds because you’re lazy and why reinvent the wheel? Its got oodles of these little shits. They swarm the PC’s, they slow them down, they niggling bits of damage, impose penalties. Now your dragon can’t be focused down with the full might of an entire party every round. They’ve got other things to worry about. Maybe the dragon can’t use its breath weapon for fear of hurting its little friends. Maybe it doesn’t give a damn and nukes them anyway - possibly right before your timer ticks down and it gets a fresh wave. You decide, but boss fights feel more epic if its not just chewing down a hitpoint pool every round.
