belated gming advice for atenistra

-my dming style is probably not perfect but here's as much as i can explain about what i do or try to do

scene painting. this is generally my initial goal and what i plan the most for. i write up a long form description of the area using languages primarily related to the five senses i.e. what the players see, hear, smell, etc as they enter a new scene, it's not a bad idea to relate dimensions in exact terms either, especially when the party enters an interior area; i also like to use ambient sound and background music to further inform the players on what kind of scene they're in. we can use language like "scenes" for anything that's not an encounter, which are situations where the party's life is in danger to some degree. even a tense confrontation might be considered an encounter; scenes don't involve much die rolling unless they lead into an encounter. after i finish the description writeup i move on to more mechanical information

dungeon design. this is where mechanical information comes into play. the general instinct is to create a series of fun encounters but i've found that encounters are better when they are natural. the dungeon should have a purpose for being there, the dragon should have a reason to be there; furthermore there's various things we need to know (or don't need) to know about a scene to help inform the party's play. i like to list the contents of each container underneath my long-form description, and likewise describe whatever creatures might be within, their attitudes and treasure; likewise, this is where we describe any traps or general environmental features. a lot of this is based on the average dungeon module, for an example:

22. Peryton Roost <-room title and number rel. to room number on gm map
This large nest on the roof is made of woven sticks, leaves, and feathers intermingled with the sun-bleached bones of sizable animals: deer, bear, or mountain goat, perhaps. <-description

The nest abuts a chimney that leads down to Talis’s bedchamber (area 18) and belongs to a mated pair of perytons. Several times during the day, the perytons leave their nest to hunt for food. There’s a 50 percent chance that the nest is empty during the day; at night, both perytons are roosting here. <-useful information
The perytons are allies of the cult and have a clear view of the portals in front of the lodge (area 1). They cannot see the stables or kennels (areas 2 and 3) from the nest. <-monster information
Combat on the Rooftop. The nest provides reasonable traction and a good perch for the perytons, but the sloped roof is tricky to stand on. At the start of its turn or whenever it takes damage, a creature standing on the roof must succeed on a DC 10 Dexterity saving throw. On a failed saving throw, the creature falls prone on the roof; if the saving throw fails by 5 or more, the creature slips and falls off the roof, taking 2d6 bludgeoning damage and landing prone on the ground. <-environmental features

Treasure <-fat loot
The nest contains bones, old bits of armor, a punctured helmet, and other detritus. Digging around for 1 minute or more also reveals two treasures: an ancient arrow-catching shield and a bag of tarnished silverware worth 100 gp.

this example also brings in maps, which i think are important to D&D but involve a lot of time. fortunately there's a huge amount of maps online and even a market thats beginning to open for D&D battlemaps. they're not hard to make either, and the ones i make are made according to pretty easy to follow photoshop tutorials. however, maps are totally unnecessary; you can draw them as you go just according to the dimensions you have listed, they don't need any detail, black and white lines are fine, and even without dimensions it's easy to run even complicated battle scenes without a detailed battlemap. however, if you want to be exact, it's not a bad idea to have at least a layout of the dungeon on hand

for my games i've been preferring a focus on player-driven goals and player action. however, you've got to start somewhere, so the first scene is often a plot hook. classic D&D adventures start in the tavern, with the party coming together after hearing rumors that drive them to nearby haunted ruins. more complicated adventures involve active factions, like cults seeking to bring about the end of the world. there's lots of good advice in the 5e dungeon master's guide relating to how you might choose your scope, your hook, and so on, and just in general on how you might want to construct an adventure. personally, i prefer to keep to a simple structure: start with a small section of the world, decide the key features: towns, mountains, ruins, swamps, undead-filled crypts; generate hooks and rumors based off the state of the world and the features you just wrote up; once the players really want to sink their teeth into one, draw up the scenes and encounters you need related to that hook, and let the players encounter them at their own pace.

encounter design. the dmg has a bunch on encounter design and it's all pretty useless honestly. if you go according to the CR system a "deadly" rated encounter is something that a fresh (just from a long rest) party can overcome with little trouble so long as they don't do anything stupid. ideally you want the party to feel a little threatened. the best way to do this is by placing things where they make sense according to the world and not according to some principles of encounter design. the creatures should ideally follow naturally from the fiction: they have their own goals, etc

in actually running the encounter, apply a light hand. don't try and make the encounter play out the way you think it should, but still try and keep a good pace. you don't even need to cheat, just consider all the magic items and things at the players' disposal and apply them to your own npcs too. furthermore, reinforcements are always possible; if these monsters aren't enough to challenge the party, then maybe a nearby door bursts open and more charge in. finally watch lots of action movies, play lots of video games etc

well, let's try to go over this and give it some semblance of structure. i talked about how i prefer a natural style of gming. i like to let the players take an active hand in deciding what the game is about. in this sense i try to set the game so i have time to prepare scenes and encounters beforehand. if something excites the players, i'll start to detail it, usually beginning with a general layout and an idea and transforming it into an area map with matching descriptions and details in a word document. i talked a little bit about structure and how you might start with a small area, a few plot hooks and expand out, detailing as the players take interest. i talked about encounter design and the importance of keeping things sensible and realistic (a barracks with 20 bunks will probably have 40 people defending it, for example) over following the dmg's rules which make for unsatisfying and easy encounters. i talked about how it's best to simply follow the rules during encounters and use the game against the players over seeking to force certain outcomes. and i think that last point is actually the key. in running a game it's really more about creating a world for the players to explore. that world will have active conspiracies, cults seeking world domination, dark sorceresses plotting in the shadows, but the players drive their own story, and they are the stars of the show. if you have a certain outcome in mind it's easy to get frustrated when the players are just going along with their characters. instead be ready for anything. there's a lot of good books on this subject, one's called "play unsafe!" and i recommend it along with the dmg (minus the encounter rules, although the new ones in xanathar's guide to everything are worth a look)

is there anything else i should talk about? voices are fun and having very detailed roleplay is fun too but it's also perfectly fine to summarize things or shorthand through scenes. a lot of it is up to each dm's individual style. all that really matters is that everyone is having fun. in order to run a successful D&D game you need to talk constantly about even the smallest things that bother you and negotiate as much as possible. the best advice i can probably give is to just try, see what works and see what doesnt. see what you like and what your players like, see what you don't like and what your players don't like.