How Do You Improvise as a DM?

A Practical Guide to Staying Flexible When Players Do the Unexpected

by Jackson Dean Chase

Founder and Lead Designer, MageGate Games

Why trust this guide?

MageGate Games founder and lead designer Jackson Dean Chase is a USA Today bestselling author, veteran Game Master with more than 40 years of gaming experience, and the author of dozens of 5e-compatible sourcebooks and GM resources.

MageGate Games was founded to create practical tools that help Game Masters prepare faster, make stronger decisions at the table, and build more memorable campaigns.

Our approach to improvisation is simple:

Do not try to invent everything from scratch.

Recombine what you already know, follow the logic of the situation, and let player choices create the next problem.

If you are still learning how much to prepare before a session, see How Much Should a DM Prep for Each Session?

If you are running your first game, start with First-Time DM Tips: What Actually Matters?

Prep Is Good, Improv Is Better

Improvisation is one of the most important DM skills—and one of the most intimidating when you are new.

Players will ignore obvious hooks, question NPCs you never expected them to care about, travel somewhere you did not prepare, solve problems in strange ways, and occasionally derail an entire session with one decision. That's normal.

Good improvisation is not about inventing brilliant material from nothing.

It is about making reasonable decisions quickly using the information you already have.

The better you understand the situation, the people involved, and what happens if nobody interferes, the easier improvisation becomes.

Improvisation Is Not the Opposite of Preparation

Many new DMs assume improvisation means preparing nothing.

That usually makes improvisation harder.

Useful preparation gives you raw material.

If you know:

• what the important NPCs want
• what threats are active
• what locations exist nearby
• what factions are doing
• what clues are available
• what happens if the players do nothing

then you can improvise around almost any player decision.

Improvisation works best when preparation gives you pieces, not scripts.

1. Ask What the Players Are Actually Trying to Do

When a player proposes something unexpected, do not immediately decide whether it works.

First ask:

“What are you trying to accomplish?”

A player might say:

“I want to climb onto the roof, jump across the alley, and drop behind the guard.”

The important part is not the exact sequence.

Their actual goal might be:

“I want to get behind the guard without being seen.”

Once you understand the goal, adjudicating the action becomes much easier.

You can decide:

• whether a roll is needed
• which ability or skill applies
• what success looks like
• what failure changes
• what consequences are reasonable

Improvisation becomes much simpler when you resolve intent, not just literal actions.

2. Follow the Logic of the World

When you do not know what should happen next, ask:

What would make sense here?

If the players insult a powerful noble in public, perhaps:

• the noble demands an apology
• the noble demands a duel to first blood or to the death and may employ a master duelist to fight in his place
• guards intervene, perhaps to injure or humiliate the party for their insolence
• vile, outrageous rumors are spread by bards or others being paid by the noble
• the party loses access to someone important
• the party loses access to the kingdom's political elite (whether nobles or appointed/elected officials)
• a rival or enemy of the noble secretly approves and may aid the party if they continue to act against the noble they insulted
• tax collectors conduct a surprise audit and "discover" an outrageous amount is owed to the crown
• one or more player characters are arrested on serious trumped-up charges until the "proper apology" is offered to the noble
• a faction, guild, temple, or other organization that one or more party members belong to decides to terminate their membership
• the party is banished from the area the noble controls on pain of death if they re-enter; the more powerful the noble, the worse this restriction becomes

Alternatively, the insulted noble could take action against the party's friends, family, lovers, or allies instead... especially if they make for easier targets, or the noble wants the party free to perform a quest or other dangerous action instead of an apology. Hot-headed nobles will seek swift revenge, while cold-blooded ones will wait as long as necessary to exact a more fitting and elaborate vengeance.

If they free a bandit instead of killing them:

• the bandit may owe them
• the gang may view the act as betrayal, not trusting why the party would let the bandit go
• the bandit may know useful information
• the bandit may become a recurring NPC
• the law may hear of it and accuse the party of being in league with the gang
• the bandit may reunite with the gang to stalk the party to rob them

You do not need a prewritten answer.

You need a consequence that follows naturally from what just happened.

3. Use NPC Motivations as Your Compass

NPCs are much easier to improvise when you know what they want.

Suppose the players unexpectedly negotiate with an enemy captain.

You may have prepared the captain only as a combat opponent.

If you know the captain:

• wants promotion
• fears disgrace
• dislikes the local commander
• values the lives of their soldiers

then you already know how they might respond.

They could bargain.

They could surrender under favorable terms.

They could betray their commander.

They could stall for time.

They could agree publicly while planning revenge.

You do not need prepared dialogue.

Motivation generates dialogue.

4. Reuse Material Instead of Throwing It Away

Players skipped the ruined watchtower you prepared?

Do not assume that work is wasted.

Move or reshape the useful parts.

The watchtower might become:

• a border fort farther down the road
• an abandoned monastery
• a ruined customs station
• a cave complex beneath another location
• a stronghold occupied by a different enemy

The same principle applies to:

• NPCs
• traps
• treasure
• encounters
• clues
• maps
• villains
• environmental hazards

Reuse the function of the material even if the surface details change.

The players do not know what was originally planned.

5. Keep a Small Bank of Flexible Material

Improvisation gets much easier when you have a few reusable pieces ready.

Keep short lists of:

• names
• NPC personalities
• rumors
• complications
• minor encounters
• environmental hazards
• secrets
• treasures
• faction goals

You do not need dozens of entries in every category.

Even five or ten flexible ideas can carry a surprising amount of play.

This is one reason simple prep tools are so useful: they reduce the number of things you must invent under pressure.

6. Use “Yes, And,” “Yes, But,” and “No, But”

You do not need to accept every player idea exactly as proposed.

Three useful responses are:

Yes, and...

The idea works and creates another opportunity.

Yes, but...

The idea works, but there is a cost, complication, or tradeoff.

No, but...

The exact idea cannot work, but another meaningful option remains.

For example:

Player: “Can I convince the guards that we are inspectors?”

Yes, and: They believe you and escort you inside.

Yes, but: They believe you, but insist on accompanying you.

No, but: They do not believe you, but one guard hints that a bribe might change things.

No, but: They pretend to and escort you inside, then lead you to reinforcements and attempt to arrest you.

This keeps the game moving without making every attempt automatically succeed.

7. Turn Failed Rolls Into New Problems

A failed roll does not have to mean nothing happens.

Instead, failure can create:

• delay
• noise
• attention
• resource loss
• new danger
• incomplete information
• a worse position
• a difficult choice

If the party fails to sneak into a manor, perhaps they still get inside—but now the guards know someone is there.

If they fail to persuade a witness, perhaps the witness refuses to talk unless the party solves another problem first.

If they fail to cross a dangerous river, perhaps they make it across but lose equipment.

Failure should often change the situation rather than freeze it.

8. Ask a Player to Help Define the World

You do not always have to invent every detail yourself.

If a player asks:

“What is the tavern called?”

you can sometimes respond:

“You have been here before. What is it called?”

Or:

“What does your character know about this neighborhood?”

“What rumor have you heard about this place?”

“What makes this guard memorable?”

Used selectively, this does three useful things:

• reduces your cognitive load
• gives players more ownership
• creates details the table is more likely to remember

You still decide what becomes canon, but you do not need to supply every decorative detail personally.

9. Slow Down When You Need Five Seconds

Improvisation feels harder when you assume you must answer instantly.

You do not.

It is completely reasonable to say:

• “Give me a second.”
• “Let me think about how they would react.”
• “I need to check one thing.”
• “Interesting. Okay…”

A short pause is much less disruptive than making a bad decision because you felt pressured to answer immediately.

Most players will barely notice.

10. When You Truly Do Not Know, Make a Temporary Ruling

Sometimes you will encounter a rules question or situation you cannot resolve quickly.

Do not let one uncertainty stop the session.

Make a reasonable temporary ruling.

Then say:

“We will run it this way for now, and I will check the exact rule afterward.”

That preserves momentum.

You can correct the ruling next session if necessary.

A game that keeps moving is usually better than a technically perfect game interrupted by ten minutes of rule searching.

11. Let Consequences Create the Next Scene

One of the easiest ways to improvise is to stop asking:

“What should happen next?”

and ask instead:

“What happens because of what just happened?”

The players:

• stole from the guild
• spared an enemy
• broke a promise
• destroyed evidence
• made an ally
• embarrassed a noble
• released a monster

Each action creates consequences.

Those consequences can become:

• the next encounter
• the next NPC problem
• the next rumor
• the next faction move
• the next session

Improvisation becomes much easier when the campaign grows out of accumulated cause and effect.

12. Know When to End the Scene

Improvised scenes can drag because nobody knows when they are finished.

End the scene when:

• the players got what they came for
• the NPC has nothing new to offer
• the decision has been made
• the danger has passed
• the consequence is clear
• the situation changes

Then move forward.

You do not need to play every conversation, journey, or minor interaction to exhaustion.

What If the Players Completely Derail the Session?

Sometimes players choose something so unexpected that your prepared material becomes almost irrelevant.

That is not a disaster.

You have four main options:

Stall for Time

Perhaps a severe storm or natural disaster occurs. It slows or prevents travel and gives you a chance to redirect the players into some alternate version of whatever you prepared. Weather and hazardous terrain can create cinematic and emotional complications that affect both the party and their enemies. That makes combat more exciting and exploration or travel more interesting.

Maybe instead of the party going to a villains' lair you haven't prepped for, the villain (or its allies and minions) ambushes the party before they ever get there. The ambush could either be en route or in their current location or home base instead.

Having the villain ambush the players reads completely different to the players than a random encounter. Random encounters rarely push the plot forward, but intentional encounters with the enemy they know (and love to hate) keeps things moving. It makes the players scared and excited to be caught off guard instead of bored by fighting pointless bandits, beasts, or wandering monsters.

Improvise Forward

If you understand the world well enough, follow the players' decision and see where it leads.

If you're unclear about their intentions, ask them one simple question:

"What do you intend to gain by this?"

Knowing their reasons and objectives can often shake something loose in your head:

• a complication, delay, or detour arises (could be natural, magical, legal, or otherwise)
• an ambush or other trap unfolds
• a side quest opportunity presents itself
• an unexpected ally, rival, or enemy learns of what the party intends to do and seeks to aid or hinder them
• friends, family, lovers, or allies are either put in danger by the party's actions, or have a sudden emergency that demands immediate attention

Recycle Prepared Material

Move useful NPCs, encounters, clues, or locations into the new direction.

We provide fun examples for how to do this in First-Time DM Tips: What Actually Matters?

End Early if Necessary

If the players make a major decision that requires preparation you genuinely do not have, it is acceptable to say:

“That changes things enough that I want to prep the next part properly. Let’s stop here for tonight.”

That is better than forcing weak material because you feel obligated to fill time.

What New DMs Get Wrong About Improvisation

New DMs often believe good improvisers:

• invent brilliant ideas instantly
• never pause
• never consult notes
• never reuse material
• never make mistakes
• always know the rules
• can run an entire campaign with no preparation

That is not how strong improvisation usually works.

Good improvisation is mostly:

understanding the situation + knowing what people want + making reasonable consequences + reusing flexible material

That is a skill you can practice.

How MageGate Games Approaches Improvisation

MageGate Games treats improvisation as structured flexibility.

The DM should prepare enough that the world has:

• motives
• pressures
• threats
• resources
• locations
• consequences

Then the players provide the direction. That is why reusable campaign material is so valuable.

A flexible villain, lair, trap, encounter, magic item, faction, or environmental complication can be adapted to many different situations.

The goal is not to prepare every possible future.

It is to build enough useful pieces that you can assemble the future when the players reveal what they actually care about.

A Simple DM Improvisation Checklist

When the players surprise you, ask:

• What are they actually trying to accomplish?
• What would logically happen here?
• What does the NPC want?
• What existing material can I reuse?
• Does success create another opportunity?
• Does failure create another problem?
• What consequence follows from this choice?
• Can a player help define a minor detail?
• Do I need a moment to think?
• Is the scene finished?

If you can answer those questions, you can usually keep the game moving.

Frequently Asked Questions

How do I get better at improvising as a DM?

Practice making small decisions quickly. Focus on NPC motivations, logical consequences, and reusable material rather than trying to invent elaborate scenes from nothing.

Do good DMs improvise everything?

No. Strong improvisation usually depends on useful preparation. Knowing the situation, NPC motivations, locations, threats, and consequences gives you material to improvise with.

What if I freeze while DMing?

Pause. Ask what the players are trying to accomplish, review what you already know, and choose the most logical consequence. You do not need to answer instantly.

Should I let players invent details?

Sometimes. Letting players define small setting details can increase engagement and reduce GM workload, as long as you retain final control over what fits the campaign.

What if my improvisation contradicts something I established earlier?

Correct it if the contradiction matters. If it is minor, choose the interpretation that keeps the game coherent and continue.

Is it okay to reuse encounters or NPCs the players skipped?

Yes. Reusing unseen material is one of the easiest ways to reduce wasted preparation.

What if the players go completely off script?

Follow the logic of their choice, reuse prepared material where possible, and stop the session early if the new direction genuinely requires substantial preparation.

Explore More Resources for Serious Game Masters

MageGate Games publishes practical 5e resources designed to help Game Masters create stronger encounters, memorable villains, distinctive magic items, dangerous environments, deeper fantasy religions, and more useful campaign material without unnecessary prep.

Explore MageGate Games' 5e GM Resources

Explore MageGate Games’ growing library of encounter design tools, boss lairs, monster customization systems, and Game Master resources built for serious 5e campaigns.

• Grab Free GM Resources from the MageGate Games Store