Spoken, Written, or Drawn: Comparing Ways to Deliver Scene Information at the Table

Game masters often face a small, recurring decision during a session: how should this piece of information reach the table? A game master can say it, hand it over on paper, or draw it. Each channel has different strengths, and the choice is rarely neutral. A name spoken once is gone as soon as the next sentence begins. A name on a card stays until someone moves the card. A sketch holds spatial relationships that would take a paragraph to narrate and still might not stick.

This comparison looks at the three channels side by side: spoken narration, written material, and drawn or visual aids. It draws on a well-known idea from learning research, cognitive load theory, to explain why the channels behave differently, and it keeps the advice grounded in table-craft rather than certainty. Much of this is a matter of preference and group makeup, and the article says so where that applies.

The goal is not to crown a winner. It is to help you pick the right channel for the right piece of information, and to combine them so that no single channel carries more than it comfortably can.

Why the Channel Matters: A Brief Frame

Cognitive load theory comes from education research and is commonly traced to John Sweller's 1988 paper on cognitive load during problem solving (Sweller, 1988). It does not come from gaming, so applying it to a tabletop session is an analogy rather than a proven rule. It is still a useful lens, because the central observation is easy to recognize at a game table: people can only hold so much in mind at once, and how information is presented changes how much of that capacity it consumes.

A few practical consequences follow from that lens, and they shape the comparison below:

• Information that vanishes the moment it is spoken must be held in memory by every listener. Information that persists in the room does not.
• Information that is spatial is often cheaper to show than to describe.
• Information that players need to look up repeatedly belongs somewhere they can look.
• Information that is only atmosphere can safely stay fleeting, because losing it costs little.

These are working heuristics, not laws. Different people process spoken, written, and visual material differently, and no source cited here claims a universal ranking of the three channels.

Channel One: Spoken Narration

Spoken narration is the default tool of the game master. It costs nothing to prepare beyond what is already in your head, it can respond instantly to what players do, and it carries tone, pacing, and character voice in ways that paper cannot.

What Speech Does Well

Atmosphere and emotion. A pause, a shift in voice, or a slowed pace can make a moment land. Justin Alexander's advice on evocative description, published on The Alexandrian, recommends a handful of evocative images and a couple of cool details rather than lengthy descriptions (Alexander, 2017). That kind of selective detail is naturally suited to speech, where a single vivid image can do the work of a paragraph.

Responsiveness. Speech can adapt mid-sentence. If a player's face shows confusion, the game master can rephrase. A handout cannot do that.

Flow and momentum. Pacing is partly a matter of how smoothly information moves, and spoken narration keeps a scene fluid. The Angry GM (2025) has written about when to stop narrating, giving a way to adjudicate and narrate long, time-consuming actions so the scene keeps moving.

Where Speech Struggles

Persistence. Speech is gone as soon as it is said. Listeners must hold names, numbers, and instructions in memory or lose them. This can be one reason a scene stalls: a stream of new names delivered by voice alone asks every player to take in and remember each one in real time.

Order and sequence. Speech is linear. If five things happen in a room, they arrive one after another, and a player who lost the first while processing the second cannot easily go back.

Spatial detail. Describing the layout of a room, the positions of combatants, or the geometry of a chase in words alone tends to produce different mental pictures at different seats. In many groups' experience, disagreements about where things are, and about whether an action is possible, can trace back to this.

Players who process better by reading or seeing. Some players may find written or visual material easier to take in than spoken material; table preference varies, and it is worth asking rather than assuming.

When Speech Is the Right Choice

• Opening a scene with mood and a few striking details.
• Voicing a character in the moment.
• Delivering consequences quickly after a roll.
• Anything players do not need to recall in ten minutes.

Channel Two: Written Material

Written material includes handouts, name cards, clue notes, letters, a visible shared document, or a running list on a whiteboard or screen. Its defining feature is that it persists. Once a player has written a name on a card, they no longer have to remember it.

What Writing Does Well

Persistence and lookup. Players can return to the information at their own pace. For names, titles, numbers, and instructions, this is often the cheapest way to lighten the load.

Precision. Written text can state exactly what a letter says, what an inscription reads, or what a contract promises, without drift between retellings.

Parallel access. Several people can read a handout at once, or one person can read it while another asks questions. Speech is a single stream; paper is not.

Reference for recurring information. Alexander's discussion of dungeon keys describes the value of keyed, scannable notes for the game master (Alexander, 2014). The same logic applies in reverse: when information is organized so it can be scanned, it is easier to use than information that must be recalled.

Where Writing Struggles

Preparation cost. Handouts take time to make. Michael E. Shea's work on prep emphasizes spending effort where it pays off rather than preparing everything (Shea, 2024a). A handout for a throwaway detail is poor value.

Pacing risk. Handing someone a page can pull them out of the scene. A long handout slid across the table asks players to stop playing and start reading, which can drain momentum, particularly for players who read slowly or dislike reading under observation.

Accessibility. Written material is not equally accessible to everyone. Dyslexia, vision differences, or simply a preference for talking things out can make dense handouts a barrier rather than a help. If you use written material, short and scannable beats long and literary.

Overuse. A table covered in paper can create its own load. If players must decide which of twelve sheets matters, the channel has become the problem.

When Writing Is the Right Choice

• Names and one-line descriptions of recurring characters.
• A short recap of the party's current goal.
• Documents that exist in the fiction: letters, notices, contracts.
• Numbers or lists players will need to consult more than once.

For related physical and digital aids, The Dungeon Trainer has also published articles on dungeon crawl props and tools (The Dungeon Trainer, 2026a) and on essential Dungeon Master tools (The Dungeon Trainer, 2026b).

Channel Three: Drawn and Visual Aids

Drawn aids include rough sketch maps, diagrams, tokens, miniatures on a grid, a virtual tabletop map, or a quick doodle on scrap paper. Their special strength is that they carry spatial and relational information without requiring anyone to translate it into words and back again.

What Drawing Does Well

Spatial relationships. Where the door is, how far the stable yard is from the hearth, which side of the bridge the enemy stands on: a sketch settles these at a glance. Mapping has deep roots in the hobby: the original 1974 edition of D&D included a keyed sample dungeon map (Alexander, 2014), and many groups still treat mapping as part of the fun.

Shared understanding. When everyone looks at the same picture, the chance of seven slightly different mental versions of a room shrinks. This does not eliminate disagreement, but it gives the table a common reference.

Low preparation, if kept rough. A sketch map does not need to be a polished piece. A rough rectangle with a few labeled features is often enough. Digital mapping tools such as Dungeon Scrawl exist for those who prefer digital maps (Dungeon Scrawl, n.d.). Printable graph-paper generators are also available online (Incompetech, n.d.). A pen and scrap paper work as well.

Relationships and connections. Diagrams can show how factions, locations, or clues connect. This echoes Alexander's critique of the linear 'plotted approach' to scenario design, in which every arrow is a chokepoint that can halt the adventure if players miss it (Alexander, 2010).

Where Drawing Struggles

Time at the table. Drawing while players wait can stall a scene. Preparing the sketch beforehand, or drawing only the few lines that matter, avoids this.

False precision. A neat map can imply that everything on it is exact and everything off it does not exist. A loose sketch invites questions; a polished one can discourage them. Which is better is a matter of table preference.

Not everything is spatial. A sketch cannot carry a character's motive or a rumor's source. It is the wrong tool for information that is not about space or connection.

Tool friction. For groups playing online, virtual tabletop features such as dynamic lighting and walls may add atmosphere and clarity (Roll20, 2026; Foundry Virtual Tabletop, 2026a, 2026b), but they also add setup and learning cost. Use them where they pay for themselves.

When Drawing Is the Right Choice

• Any scene where position, distance, or layout matters to decisions.
• Chases, standoffs, and anything involving movement through space.
• Showing how locations or people connect.
• Settling a recurring argument about where something is.

Side-by-Side Comparison

The table below summarizes how the three channels tend to behave. Treat it as a guide to tendencies, not a verdict.

| Question | Spoken | Written | Drawn |
|---|---|---|---|
| Does it persist after delivery? | No | Yes | Yes |
| Preparation cost | Low | Low to moderate | Low if rough, higher if polished |
| Best for mood and emotion | Strong | Moderate | Weak |
| Best for exact wording | Weak | Strong | Weak |
| Best for space and layout | Weak | Weak | Strong |
| Best for names and lists | Weak | Strong | Moderate |
| Risk of stalling the scene | Moderate if long | Moderate to high if long | Moderate if drawn live |
| Adapts mid-sentence to reactions | Strong | None | Limited |
| Accessibility concerns | Hearing, attention | Reading, vision | Vision, spatial processing |

Two patterns stand out. First, persistence is the dividing line between speech and the other two channels. Second, each channel is strongest at a different kind of information, which suggests that matching information type to channel is more productive than favoring one channel across the board.

Matching Information Type to Channel

A useful habit is to sort what you are about to deliver into rough categories and pick a channel for each.

Mood and sensory detail. Speak it. Keep it short and specific. This information is meant to be felt in the moment, not retained.

Names, titles, and wants. Write it, even if only a word or two on a card. Pair it with a short spoken introduction so the name arrives with a face.

Layout and positions. Draw it, or show it on a map. Mention the key points aloud so the sketch and the speech reinforce each other.

Goals and stakes. Say it briefly, and keep a one-line written version visible. A recap line is the classic example, and short written reminders help players return to the point of the scene.

Choices. Speak a short shortlist, and make clear that players may do something else. Shea's treatment of scenes as a flexible catch-all step in prep fits this approach, where a scene is a handful of possibilities rather than a script (Shea, 2024b).

Secrets and clues. Think about whether players will need to recall them. Alexander's Three Clue Rule recommends providing multiple routes to a conclusion so that a missed clue does not stall the game (Alexander, 2008). Some game masters find that a clue important enough to include three times is worth writing down once for the table.

Combining Channels: The Principle of Reinforcement

The strongest approach is rarely a single channel. When two channels carry the same idea in complementary ways, each supports the other, and the load is shared.

• Say it and show it. Introduce a character aloud while a player jots the name on a card. The voice gives the name life; the card gives it permanence.
• Draw it and name it. Point to the sketch while narrating. The words explain the picture; the picture anchors the words.
• Write the question, speak the options. Keep the party's goal visible in writing while offering choices aloud.

A caution: reinforcement is not the same as repetition across every channel. Delivering the same information three ways in a row can itself become noise. Use a second channel where the first is weak, not by default.

Considering the Players

The best channel depends partly on who is at the table. Some practical points:

Ask, do not assume. Short questions during a session zero or early in a campaign about how people like to receive information can save a great deal of friction. Session zero is a natural place to ask; Shea (2021) describes using it to get everyone on the same page about the game and to set expectations, and the same conversation can include how players like to receive information.

Watch for signals. A player who keeps asking for names to be repeated may benefit from written cards. A player who asks where things are may benefit from a sketch. A player who goes quiet when handed a long page may prefer a spoken summary.

Offer, do not impose. Making name cards available to everyone, without requiring their use, lets each person choose what helps them. This also avoids singling out anyone.

Remember new players. New players may be carrying the most, since they are learning the rules and the fiction at the same time, so extra persistent aids may help them, though tables differ.

Common Mistakes When Choosing a Channel

• Speaking everything. Treating narration as the only tool leaves the table to hold every name and detail in memory.
• Handing over walls of text. A long handout can substitute reading for play and pull people out of the scene.
• Drawing too much, too soon. A fully detailed map can overwhelm and can imply that unmarked features do not exist.
• Using the same channel for every kind of information. Names, layouts, and moods are different problems.
• Adding tools for their own sake. A new aid should earn its place by lightening a real load.
• Ignoring the people at the table. A technique that works for one group may not suit another, and no source cited here claims otherwise.

A Quick Decision Guide

When unsure, run through these questions in order:

1. Will players need this in ten minutes? If yes, put it somewhere persistent. If no, speak it and let it go.
2. Is it about where things are? Draw it.
3. Is it a name, number, or exact wording? Write it.
4. Is it mood or consequence? Speak it, briefly.
5. Will adding a channel help, or just add clutter? Add only what lightens a real load.

Example: The Lantern Road Company

This is an illustrative scenario, not a report of an actual game. It shows how the three channels might be combined for one scene with a fictional group.

The Lantern Road Company is a four-player party escorting a merchant caravan through a mountain pass. The game master is preparing the opening scene at a crowded waystation, and wants to decide how each piece of information should reach the table.

Sorting the information. The game master lists what the scene contains. There is atmosphere: smoke, wet wool, the noise of a busy common room, the sense that the pass may close soon. There are three characters: a warm, hurried innkeeper who wants help with a broken cart, a smug rival caravan master who wants to leave first, and a nervous tax collector who wants an escort. There is a layout: a common room with a hearth, a door, and a stable yard. And there is a goal: reach the next town before the pass closes.

Choosing the channels.

• Atmosphere is spoken. The game master opens with two or three striking details and stops. Nothing about the smoke or the weather needs to be remembered later, so it stays fleeting.
• Names and wants are written. The game master prepares a short note card with the three characters, each with one trait and one want. As each character is introduced aloud in a short burst, the player running the cautious scout jots the name on a card. The voice gives each person a face; the card means no one has to hold three names in memory.
• Layout is drawn. A rough rectangle with the hearth, the door, and the stable yard marked sits in the middle of the table. When a player asks whether the scout can slip out to the yard unseen, the sketch answers most of the question at a glance, and the game master adjudicates the rest.
• The goal is a one-liner. A single written line, reach the next town before the pass closes, stays visible, and the game master repeats it aloud at the start of the scene.
• Choices are spoken, loosely. When the party reaches a decision, the game master offers a shortlist: help the innkeeper, question the tax collector, or watch the rival master. Players remain free to do something else.

What changed. The same characters, the same tension, and the same pass exist as in the earlier telling of this scene. What changed is where each piece of information lives. The voice carries mood. Paper carries names. The sketch carries space. No single channel was asked to do all three jobs, which is the point of the comparison.

A note on restraint: the game master does not add a handout for the weather, or a diagram of the innkeeper's family, or a second map of the stable. Each channel was added only where it lightened a real load.

A Note on Tools

Some groups prefer to keep their notes, maps, and shared references in one place so everyone can see the same material. The Dungeon Trainer is one option for groups who want a shared space to organize and play together (The Dungeon Trainer, 2026c). Whether you use a tool like that, a whiteboard, or a stack of index cards is a matter of table preference. The principle stays the same: match the channel to the information.

Takeaways

• Spoken narration is strong for mood, emotion, and responsiveness, and weak for persistence.
• Written material is strong for names, exact wording, and lookup, and carries preparation and pacing costs.
• Drawn aids are strong for space and connection, and are weak for information that is not spatial.
• Persistence is the main dividing line: what must be remembered belongs somewhere durable.
• Combine channels where one is weak, and avoid repeating the same information in every channel.
• Ask your players what helps, offer aids rather than imposing them, and add a tool only when it lightens a real load.
• All of this is a matter of tendency and table preference, not fixed rule.

Sources and Further Reading

• Alexander, J. (2008, May 8). The Three Clue Rule. The Alexandrian. https://thealexandrian.net/wordpress/1118/roleplaying-games/three-clue-rule
• Alexander, J. (2010, May 27). Node-based scenario design, part 1: The plotted approach. The Alexandrian. https://thealexandrian.net/wordpress/7949/roleplaying-games/node-based-scenario-design-part-1-the-plotted-approach
• Alexander, J. (2014, May 19). The Art of the Key. The Alexandrian. https://thealexandrian.net/wordpress/35180/roleplaying-games/the-art-of-the-key
• Alexander, J. (2017, September 4). Random GM Tip: Evocative Descriptions. The Alexandrian. https://thealexandrian.net/wordpress/39227/roleplaying-games/random-gm-tips-evocative-descriptions
• Dungeon Scrawl. (n.d.). Dungeon Scrawl. https://www.dungeonscrawl.com/
• Foundry Virtual Tabletop. (2026a). Lighting. https://foundryvtt.com/article/lighting/
• Foundry Virtual Tabletop. (2026b). Walls. https://foundryvtt.com/article/walls/
• Incompetech. (n.d.). Graph Paper Generators.
• Roll20. (2026). Dynamic Lighting. https://pages.roll20.net/dynamic-lighting
• Shea, M. E. (2021, April 5). Running session zeros. Sly Flourish. https://slyflourish.com/runningsessionzeros.html
• Shea, M. E. (2024a, June 3). High Value Prep. Sly Flourish.
• Shea, M. E. (2024b, September 2). Scenes: The catch-all step of the Lazy Dungeon Master. Sly Flourish. https://slyflourish.com/scenescatchall_step.html
• Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257-285.
• The Angry GM. (2025, July 31). When to stop narrating. The Angry GM.
• The Dungeon Trainer. (2026a). Essential Dungeon Crawl Props and Tools: A Game Master's Toolkit. https://thedungeontrainer.com/blog/essential-dungeon-crawl-props-and-tools-a-game-master-s-toolkit
• The Dungeon Trainer. (2026b). Essential Dungeon Master Tools: What You Actually Need at the Table. https://thedungeontrainer.com/blog/essential-dungeon-master-tools-what-you-actually-need-at-the-table
• The Dungeon Trainer. (2026c). The Dungeon Trainer.