I'm a new DM who's trying to create a world for my first campaign. I was wondering, from a prep standpoint, do I need to have the entire world completely fleshed out before session 1, or can I just start the game with a basic framework (names of cities, major powers, general locations) and only have in depth info (npcs, plots, etc.) for the first city/area that my players start in?
Also, for those who have created their own worlds and run campaigns through them, what are the most important pieces of information to have when describing an area and what are some things that seem important but could really be left out or improvised?
for me, all I really prepare in advance is the starting town, and a few quests and just have a random generator for names on my phone. The PC's will generally tell you what they want to do you can always set up dungeons, ruins, forests between seasons. if you plan too much they will never follow everything you want them to
Build "just enough". You don't have to keep everything in your head, just some broad ideas.
First, start with a session 0, where you talk about your world and give the players info as to what they can be and what rules you are using.
Then, create their first area and a bit more. In your head (or on paper) you can have a lot more fleshed out ideas-wise but slowly grow that organically as the players explore. Keep fleshing out things a bit more as you get used to your players and things solidify in your head. At least for me as DM I feel I provide a broad outline of a world, but the character interactions are what turn the world from an idea into actuality.
To answer your question, for me I like the following:
for the overall story, what is the theme/idea overarching plot
you don't have to have all of that at once, just an idea of what problems you want your players to try to solve
for each area, who is in charge, what are they like, what are some basic rules.
build up encounters and drop plot clues along the way.
IE you explore a barbarian tomb and find an old tablet with a poem
The players may or may not have an immediate use for that, but they will think about it, and it will get in their collective conscious so that you can then introduce a puzzle that the tablet helps with.
if you understand your own broad strokes, then that makes improv a lot easier, as you already have a basic idea
Hope that helps!
Rollback Post to RevisionRollBack
"An' things ha' come to a pretty pass, ye ken, if people are going to leave stuff like that aroound where innocent people could accidentally smash the door doon and lever the bars aside and take the big chain off'f the cupboard and pick the lock and drink it!"
I'm a new DM who's trying to create a world for my first campaign. I was wondering, from a prep standpoint, do I need to have the entire world completely fleshed out before session 1, or can I just start the game with a basic framework (names of cities, major powers, general locations) and only have in depth info (npcs, plots, etc.) for the first city/area that my players start in?
Also, for those who have created their own worlds and run campaigns through them, what are the most important pieces of information to have when describing an area and what are some things that seem important but could really be left out or improvised?
Hi there, welcome to DMhood!
First, no, I don't think you need to have your entire world developed before starting. Like you said, a framework of places, people, and events are the main parts you need for running the game. My advice is to always just prepare content that the group will interact with in the game--don't spend an hour writing a page about the cyclopian frost kraken goddess of the northern sea if the first 10 sessions are meant to take place in the far-southern desert.
Second, talking about areas and locations. The most important thing to note about an area is its problems. Anytime the group shows up somewhere, it should become apparent quickly that something is wrong (doesn't have to be clear what exactly). In light of that, the group should know where they can go to find out more information--is there a tavern, a temple, a townhall, or a keep? Next, who's in charge and what do the people look like? Is the area patrolled by professional uniformed soldiers, or defended by local militia with farm tools? Is everyone a poor peasant, or do they seem well-off? Lastly, is environment. You don't have to turn into Tolkien and give in depth descriptions of the shades of green seen in the forest and the grass, but just let everyone know what environment surrounds them. tl;dr: When describing an area point out clues that show where the action is, note what types of buildings can be visited, describe how people look or what station they seem to hold, and explain what the surrounding environment is like. Doesn't have to happen in this exact order, but I think these are the most vital pieces of information to be given.
Oh, one more thing and this is very important. Do not go into this thinking it will be like Critical role, Dice Camera Action, or any of the D&D Streams. If you put that kind of pressure on you or your players and it doesn't happen that way you might think something is wrong.
When you first start your game, have the map of the local area figured out, anything beyond that is just a description. From there, as you get more settled in, flesh out what you might need in the very near future. Once you're feeling good about all of that, finish the rest of the map in your spare time.
I run a game in a home brew world called Loromir. The group started out in the city of Talmond, I had the entire town mapped out, down to which NPCs lived where, which dieties were worshiped at the local hall, what the major NPCs wanted politically, and which farms and ranches had what crops or animals. From there I mapped out the region for what would be about 2 days travel from Talmond. The coastal islands they could reasonably reach, the forests, mountains, caves, and other geographical items were all plotted. I made sure that any structures that would be important to the main quest, and side quests, were present. The reason for this is because I knew the players were going to spend a lot of time in the town, it's the location where the players were going to get their feet wet and start to learn more about the game. The other reason is, I kind of go overboard with details once I start...some of that information wasn't necessary.
Any near by cities were noted, not fleshed out, just notes. A few lines that would be common knowledge for travelers, as well as some basic information about how the town functioned. This information could be trickled down to the players if they asked about it, if they found a map, or if they needed the information due to a quest. I didn't need a map, I didn't need to know anything more than the description of the town and any key NPCs that might be there. However, when the party decided it was time to leave Talmond, that's when I sat down and mapped out the town they were headed to. I didn't need to map out any of the other near by towns because they weren't important yet, just their basic details.
Once the players gave me an idea of where they were taking their party, I could anticipate their destinations and map accordingly. I could map out days or weeks of journey ahead of them, giving me the ability to flesh out details and make it seamless for them. I could also start fleshing out anything that they might side-track off to in the journey. Due to having all of the important, quest related, locations mapped, I could play with the rest of the map with my spare time.
TL;DR: Start with where the players are and detail it as much as you want. Map where the players are headed, and where you anticipate where they're headed. Finish the rest of the map in your free time. You do not have to map what they don't see, just make sure you have a few details you can tell them about the location.
Have the starting town made. A handful of NPC's present for a questhook or two, one NPC in conflict with another for some reason, and 1 or 2 shop/information NPC's. Also have a rough idea about possible conflicts in the direct area. A rough idea about a faction or two.
With that you should have enough information to fall back on in order to improvise. It always helps if you know that there is a possible faction that could oppose or aid the players. All you do is make up a name, race and put a faction representative on their path. It is often also a method you can employ to nudge the players in the direction you want. Making sure they don't go too far off track.
Knowing roughly the information/quests in that area will aid in throwing another NPC on their path in order to get the players on track. As example. Group 1 had a firbolg talking to farm animals to get basic information. talking to some key NPC's they eventually gather enough info to head south into a barren wasteland where there is an old ruin. Group 2 also had a firbolg which did a ritual among some trees to commune with them. gained a vision of nightly attacks. he convinces his friends to find out which farm is on that route and lays an ambush. they wait and end up fighting a group of creatures. the tabaxi takes a piece of those creatures back to town where a gnome priestess does a biopsy. she informs the group that this creature is similar to those of the nightly attacks. the creatures were also moving from the northern woods to the south. ---different group play, different improvisations, but the same outcome. all using the same basic rough understanding of some NPC's and information available in the area. which NPC provides which information, and in what way, is less important and can be improvised.
Keep in mind that the job of a DM often involves smoke and mirrors. The illusion of free will and that their decision matter. Does it really matter if they go left or right upon reaching a fork in the road? Your players won't know what lays ahead of them in any given direction. Use that to your advantage when it comes to improvising and possible even stalling the players while you come up with stuff for them to experience.
Example how I "flesh out" a simple village
Terrelton
Dusty town. 2100 population, mostly human and minority mix of other races. Gets on by livestock and leather trades. Several large foul-smelling tanneries in the hills east of town. Run by Merchant’s Council with connections to various criminal elements.
Locations Town Hall used mainly by the representatives of the Merchant’s Council Shrine to Pelor has its own building where people can congregate to hear Thea Redstream. General Store is adjacent to a large storage where caravans can load and unload their goods. Pawn Shop run by brigands to sell their stolen goods if the Council doesn’t want it. Jail to keep up the appearance of some sort of law. The 20 guards are arrogant former brigands working for the council Broken Antler Inn is a place for travelers to get some basic food and drinks as well as a place to sleep. Guard your belongings when resting.
NPC’s Holme Barrowes – halfling, male, merchant, 47, short straight red hair, brown eyes, pockmarked skin, square glamorous face, goes straight to the point, doesn’t like parting with his money, pet rabbit named Noodle, lead of the Merchant Council. Urdur Metalbreaker – dwarf, male, merchant, 212 years, short straight black dyed hair and golden eyes, rugged and scarred skin, beefy build with a round revolting face and a very long beard, slightly deaf from right ear, perfectionist, considers everyone else as idiots, sporadically lies. Second of the Council. Mirabelle Wheatflow – human, female, merchant, 22 years, long curled blonde hair and blue eyes, silky chocolate skin, 162cm muscular build, edgy pretty face. Rather acts then talks, mostly talks in whispers, misquotes proverbs. Up and coming ambitious Council member. Leille – human, female, cleric lvl6 of st. Cuthbert, 23 yeards, straight gray hair and green eyes, soft brown skin, unremarkable face, very self-confident. Thea Redstream – human, female, cleric, 41, long curled black hair and cyan eyes, soft brown skin, slim build, edgy forgettable face, clumsy, can’t stand laziness, goes out at night looking for branches. Gilla Hornblower – halfling, female, trader, 90years, long braided brown hair and eyes, smooth golden skin, triangular shaped face and a piercing on her lip, she’s very patient and runs the Pawn Shop. Cassius Hilless – Half-orc, male, 25, short gray mohawk and red yes, soft gray skin, light allergy to gnomes, very greedy and open for bribes. He runs the jail and “guards”. Contact for the Crimson Tigers and Lady Dagger from Drellin’s Ferry. Thomas Chaucer – half-elf, male, 32 years, is a trapper, cropped brown hair and cyan eyes, rough pockmarked white skin, a lean build, oval bland face, is a pacifist, always shares his wisdom. Leofrick – human, male, innkeep, 32 years, bald head and golden eyes, dark skin, sharp face features, straight to the point, head of a local drug dealing circle.
Quest Hooks 1. Leile is a crusading cleric looking into the Merchant’s Council. She has many leads to follow, but can delegate one task to the players. Apparently they’re mining not far from town and she wants to know more about it. Find whatever evidence possible at the mine. 2. Thea Redstream is pressured by the local drug dealers (Leofrick) into selling drugs to her congregation
I personally have learned to love the "blank space," compared to having the entire world built out on a map I can roll out. Players are interesting and crafty - they sometimes walk into session 1 with backstory, their own lore, and thoughts. For example, someone rolled up a ranger with "Mountains" as their terrain of choice. So I shifted from "sparse grassland" that I was originally thinking of to "cold mountain town," and began to build off of that. I've had someone say they wanted to play someone in the kingdom's assassins guild - suddenly there was a assassins guild in the city.
Letting your players assist you in world building gives them stake and agency in the narrative (especially if they enjoy / want that.) Don't be afraid when they walk into a new town to ask a player to build or set something. For example: "Oh, PlayerX, you have the background of Entertainer right? Well, let's say you played at the Inn of the town you're coming up to. What happened last time you saw the innkeeper?" They may say nothing. Or they may say PlayerX robbed the innkeep / seduced their kid / caused a giant party that wrecked the inside.
Suddenly, your world has more depth to it and you didn't have to create it.
That blank space also works when entering ancient tombs. The rogue checks for traps. Dice "says" there is some so you make one up. Another player then describes what they want to do. The dice once again says "yes" so you improv add something. That way an entire memorable tomb gets created for you in an organic fluid way. Only when the dice says nothing happens you as a DM need to add something to keep the event going.
Blank slate building is a lot of fun, and can really make for some interesting sessions. The only thing I'd caution about it is that this approach not suitable for newer DMs who don't have a firm grasp of the game an it's mechanics.
Some of us have been DMing since Gygax created this game so, when it comes to knowing the mechanics, we can whip up something faster than a Bard can seduce an NPC. A player rolls a 23 to find a trap on a door, DM hasn't planned for a trap but it sounds like a good idea, we go with a classic needle point trap, set the DC and go with it. This type of thing is easy since you know that the player has already found the trap, so there's no real danger and the player "wins" because they found it and can disarm it. The trick is when that same player rolls a 14 to find a trap and you think a trap is a good idea. This time you have to make a choice, is a 14 enough to find the trap or not. Let's say the DC for the trap is just missed, you like the number 15 for DCs. Now you have to figure out what type of trap makes sense for the location you're in, a poison needle, a falling blade, magic mist, dropped stones, what would work best. Then you have to come up with whether there's a second save involved for not taking damage, and all the rest of the mechanics of the trap. Not easy to do if you don't have an arsenal of past experiences and numbers rolling around in your DM toolbox.
If you're not confident in your improv skills, put some work into your mapping. If you want to test your chops, keep the details to a skeleton, only jotting down the important notes and leave out the fluff. However, having a blank page will let the players take a hand in world building and will make for some interesting additions to your game you may never have thought of.
I think Improvisation is just another word for stealing and re-skinning material from your favorite books or elements of pre-made modules. As a new DM invest in reading books and such. They give ideas. Unless you run into creative blocks like I have lately.
One of my group's longest and most favored campaigns literally started with "you're seated together around your campfire in the forest when...". That was all we knew: our characters were familiar with each other, we were in a forest and it was probably late in the day. What followed was year after year of gradually improvised and growing world.
Rollback Post to RevisionRollBack
I am one with the Force. The Force is with me.
To post a comment, please login or register a new account.
Hi all,
I'm a new DM who's trying to create a world for my first campaign. I was wondering, from a prep standpoint, do I need to have the entire world completely fleshed out before session 1, or can I just start the game with a basic framework (names of cities, major powers, general locations) and only have in depth info (npcs, plots, etc.) for the first city/area that my players start in?
Also, for those who have created their own worlds and run campaigns through them, what are the most important pieces of information to have when describing an area and what are some things that seem important but could really be left out or improvised?
for me, all I really prepare in advance is the starting town, and a few quests and just have a random generator for names on my phone. The PC's will generally tell you what they want to do you can always set up dungeons, ruins, forests between seasons. if you plan too much they will never follow everything you want them to
Build "just enough". You don't have to keep everything in your head, just some broad ideas.
First, start with a session 0, where you talk about your world and give the players info as to what they can be and what rules you are using.
Then, create their first area and a bit more. In your head (or on paper) you can have a lot more fleshed out ideas-wise but slowly grow that organically as the players explore. Keep fleshing out things a bit more as you get used to your players and things solidify in your head. At least for me as DM I feel I provide a broad outline of a world, but the character interactions are what turn the world from an idea into actuality.
To answer your question, for me I like the following:
Hope that helps!
"An' things ha' come to a pretty pass, ye ken, if people are going to leave stuff like that aroound where innocent people could accidentally smash the door doon and lever the bars aside and take the big chain off'f the cupboard and pick the lock and drink it!"
Hi there, welcome to DMhood!
First, no, I don't think you need to have your entire world developed before starting. Like you said, a framework of places, people, and events are the main parts you need for running the game. My advice is to always just prepare content that the group will interact with in the game--don't spend an hour writing a page about the cyclopian frost kraken goddess of the northern sea if the first 10 sessions are meant to take place in the far-southern desert.
Second, talking about areas and locations. The most important thing to note about an area is its problems. Anytime the group shows up somewhere, it should become apparent quickly that something is wrong (doesn't have to be clear what exactly). In light of that, the group should know where they can go to find out more information--is there a tavern, a temple, a townhall, or a keep? Next, who's in charge and what do the people look like? Is the area patrolled by professional uniformed soldiers, or defended by local militia with farm tools? Is everyone a poor peasant, or do they seem well-off? Lastly, is environment. You don't have to turn into Tolkien and give in depth descriptions of the shades of green seen in the forest and the grass, but just let everyone know what environment surrounds them.
tl;dr: When describing an area point out clues that show where the action is, note what types of buildings can be visited, describe how people look or what station they seem to hold, and explain what the surrounding environment is like. Doesn't have to happen in this exact order, but I think these are the most vital pieces of information to be given.
Oh, one more thing and this is very important. Do not go into this thinking it will be like Critical role, Dice Camera Action, or any of the D&D Streams. If you put that kind of pressure on you or your players and it doesn't happen that way you might think something is wrong.
When you first start your game, have the map of the local area figured out, anything beyond that is just a description. From there, as you get more settled in, flesh out what you might need in the very near future. Once you're feeling good about all of that, finish the rest of the map in your spare time.
I run a game in a home brew world called Loromir. The group started out in the city of Talmond, I had the entire town mapped out, down to which NPCs lived where, which dieties were worshiped at the local hall, what the major NPCs wanted politically, and which farms and ranches had what crops or animals. From there I mapped out the region for what would be about 2 days travel from Talmond. The coastal islands they could reasonably reach, the forests, mountains, caves, and other geographical items were all plotted. I made sure that any structures that would be important to the main quest, and side quests, were present. The reason for this is because I knew the players were going to spend a lot of time in the town, it's the location where the players were going to get their feet wet and start to learn more about the game. The other reason is, I kind of go overboard with details once I start...some of that information wasn't necessary.
Any near by cities were noted, not fleshed out, just notes. A few lines that would be common knowledge for travelers, as well as some basic information about how the town functioned. This information could be trickled down to the players if they asked about it, if they found a map, or if they needed the information due to a quest. I didn't need a map, I didn't need to know anything more than the description of the town and any key NPCs that might be there. However, when the party decided it was time to leave Talmond, that's when I sat down and mapped out the town they were headed to. I didn't need to map out any of the other near by towns because they weren't important yet, just their basic details.
Once the players gave me an idea of where they were taking their party, I could anticipate their destinations and map accordingly. I could map out days or weeks of journey ahead of them, giving me the ability to flesh out details and make it seamless for them. I could also start fleshing out anything that they might side-track off to in the journey. Due to having all of the important, quest related, locations mapped, I could play with the rest of the map with my spare time.
TL;DR: Start with where the players are and detail it as much as you want. Map where the players are headed, and where you anticipate where they're headed. Finish the rest of the map in your free time. You do not have to map what they don't see, just make sure you have a few details you can tell them about the location.
Previous advice says it mostly.
Have the starting town made. A handful of NPC's present for a questhook or two, one NPC in conflict with another for some reason, and 1 or 2 shop/information NPC's. Also have a rough idea about possible conflicts in the direct area. A rough idea about a faction or two.
With that you should have enough information to fall back on in order to improvise. It always helps if you know that there is a possible faction that could oppose or aid the players. All you do is make up a name, race and put a faction representative on their path. It is often also a method you can employ to nudge the players in the direction you want. Making sure they don't go too far off track.
Knowing roughly the information/quests in that area will aid in throwing another NPC on their path in order to get the players on track. As example. Group 1 had a firbolg talking to farm animals to get basic information. talking to some key NPC's they eventually gather enough info to head south into a barren wasteland where there is an old ruin.
Group 2 also had a firbolg which did a ritual among some trees to commune with them. gained a vision of nightly attacks. he convinces his friends to find out which farm is on that route and lays an ambush. they wait and end up fighting a group of creatures. the tabaxi takes a piece of those creatures back to town where a gnome priestess does a biopsy. she informs the group that this creature is similar to those of the nightly attacks. the creatures were also moving from the northern woods to the south.
---different group play, different improvisations, but the same outcome. all using the same basic rough understanding of some NPC's and information available in the area. which NPC provides which information, and in what way, is less important and can be improvised.
Keep in mind that the job of a DM often involves smoke and mirrors. The illusion of free will and that their decision matter. Does it really matter if they go left or right upon reaching a fork in the road? Your players won't know what lays ahead of them in any given direction. Use that to your advantage when it comes to improvising and possible even stalling the players while you come up with stuff for them to experience.
Example how I "flesh out" a simple village
Terrelton
Dusty town. 2100 population, mostly human and minority mix of other races. Gets on by livestock and leather trades. Several large foul-smelling tanneries in the hills east of town. Run by Merchant’s Council with connections to various criminal elements.
Locations
Town Hall used mainly by the representatives of the Merchant’s Council
Shrine to Pelor has its own building where people can congregate to hear Thea Redstream.
General Store is adjacent to a large storage where caravans can load and unload their goods.
Pawn Shop run by brigands to sell their stolen goods if the Council doesn’t want it.
Jail to keep up the appearance of some sort of law. The 20 guards are arrogant former brigands working for the council
Broken Antler Inn is a place for travelers to get some basic food and drinks as well as a place to sleep. Guard your belongings when resting.
NPC’s
Holme Barrowes – halfling, male, merchant, 47, short straight red hair, brown eyes, pockmarked skin, square glamorous face, goes straight to the point, doesn’t like parting with his money, pet rabbit named Noodle, lead of the Merchant Council.
Urdur Metalbreaker – dwarf, male, merchant, 212 years, short straight black dyed hair and golden eyes, rugged and scarred skin, beefy build with a round revolting face and a very long beard, slightly deaf from right ear, perfectionist, considers everyone else as idiots, sporadically lies. Second of the Council.
Mirabelle Wheatflow – human, female, merchant, 22 years, long curled blonde hair and blue eyes, silky chocolate skin, 162cm muscular build, edgy pretty face. Rather acts then talks, mostly talks in whispers, misquotes proverbs. Up and coming ambitious Council member.
Leille – human, female, cleric lvl6 of st. Cuthbert, 23 yeards, straight gray hair and green eyes, soft brown skin, unremarkable face, very self-confident.
Thea Redstream – human, female, cleric, 41, long curled black hair and cyan eyes, soft brown skin, slim build, edgy forgettable face, clumsy, can’t stand laziness, goes out at night looking for branches.
Gilla Hornblower – halfling, female, trader, 90years, long braided brown hair and eyes, smooth golden skin, triangular shaped face and a piercing on her lip, she’s very patient and runs the Pawn Shop.
Cassius Hilless – Half-orc, male, 25, short gray mohawk and red yes, soft gray skin, light allergy to gnomes, very greedy and open for bribes. He runs the jail and “guards”. Contact for the Crimson Tigers and Lady Dagger from Drellin’s Ferry.
Thomas Chaucer – half-elf, male, 32 years, is a trapper, cropped brown hair and cyan eyes, rough pockmarked white skin, a lean build, oval bland face, is a pacifist, always shares his wisdom.
Leofrick – human, male, innkeep, 32 years, bald head and golden eyes, dark skin, sharp face features, straight to the point, head of a local drug dealing circle.
Quest Hooks
1. Leile is a crusading cleric looking into the Merchant’s Council. She has many leads to follow, but can delegate one task to the players. Apparently they’re mining not far from town and she wants to know more about it. Find whatever evidence possible at the mine.
2. Thea Redstream is pressured by the local drug dealers (Leofrick) into selling drugs to her congregation
Lots of solid advice given out so far.
I personally have learned to love the "blank space," compared to having the entire world built out on a map I can roll out. Players are interesting and crafty - they sometimes walk into session 1 with backstory, their own lore, and thoughts. For example, someone rolled up a ranger with "Mountains" as their terrain of choice. So I shifted from "sparse grassland" that I was originally thinking of to "cold mountain town," and began to build off of that. I've had someone say they wanted to play someone in the kingdom's assassins guild - suddenly there was a assassins guild in the city.
Letting your players assist you in world building gives them stake and agency in the narrative (especially if they enjoy / want that.) Don't be afraid when they walk into a new town to ask a player to build or set something. For example: "Oh, PlayerX, you have the background of Entertainer right? Well, let's say you played at the Inn of the town you're coming up to. What happened last time you saw the innkeeper?" They may say nothing. Or they may say PlayerX robbed the innkeep / seduced their kid / caused a giant party that wrecked the inside.
Suddenly, your world has more depth to it and you didn't have to create it.
DM PbP Series (Completed) : The Blackmire Company, Voyage of the Windhook ,Shanty of Soulripper, The Streets of Dunnour
That blank space also works when entering ancient tombs. The rogue checks for traps. Dice "says" there is some so you make one up. Another player then describes what they want to do. The dice once again says "yes" so you improv add something. That way an entire memorable tomb gets created for you in an organic fluid way. Only when the dice says nothing happens you as a DM need to add something to keep the event going.
Blank slate building is a lot of fun, and can really make for some interesting sessions. The only thing I'd caution about it is that this approach not suitable for newer DMs who don't have a firm grasp of the game an it's mechanics.
Some of us have been DMing since Gygax created this game so, when it comes to knowing the mechanics, we can whip up something faster than a Bard can seduce an NPC. A player rolls a 23 to find a trap on a door, DM hasn't planned for a trap but it sounds like a good idea, we go with a classic needle point trap, set the DC and go with it. This type of thing is easy since you know that the player has already found the trap, so there's no real danger and the player "wins" because they found it and can disarm it. The trick is when that same player rolls a 14 to find a trap and you think a trap is a good idea. This time you have to make a choice, is a 14 enough to find the trap or not. Let's say the DC for the trap is just missed, you like the number 15 for DCs. Now you have to figure out what type of trap makes sense for the location you're in, a poison needle, a falling blade, magic mist, dropped stones, what would work best. Then you have to come up with whether there's a second save involved for not taking damage, and all the rest of the mechanics of the trap. Not easy to do if you don't have an arsenal of past experiences and numbers rolling around in your DM toolbox.
If you're not confident in your improv skills, put some work into your mapping. If you want to test your chops, keep the details to a skeleton, only jotting down the important notes and leave out the fluff. However, having a blank page will let the players take a hand in world building and will make for some interesting additions to your game you may never have thought of.
I think Improvisation is just another word for stealing and re-skinning material from your favorite books or elements of pre-made modules. As a new DM invest in reading books and such. They give ideas. Unless you run into creative blocks like I have lately.
You only need to build what the characters will experience in the upcoming session. That being said, build as much as you feel like.
"Not all those who wander are lost"
One of my group's longest and most favored campaigns literally started with "you're seated together around your campfire in the forest when...". That was all we knew: our characters were familiar with each other, we were in a forest and it was probably late in the day. What followed was year after year of gradually improvised and growing world.
I am one with the Force. The Force is with me.