In Defense of Flavor Text

Flavor text, or boxed text, is a staple of classic adventure modules. The bit of copy set aside to read to the players. The part they can know, minus things like where the secret door is or the solution to the riddle carved on the wall.

In recent years, many (certainly not all) writers and reviewers of OSR-style products have voiced criticisms of boxed text. They find it distracting and disruptive to setting the pace of a game, even to the point of being jarring. The text can make also assumptions; such as the characters taking certain actions or that events having occurred in a certain sequence. 

These are valid complaints. While the GM can certainly amend text on the fly to reflect the current status of his individual game, part of the point to a published product like a module is to do some of that lifting for him.

When I have written flavor text for adventures, I have learned (imperfectly) to only present what can be immediately perceived upon entering a room or area, regardless of what may have happened before (within reason) and leaving further revelations to be determined by the players’ choices. Not all published products follow this model, so I can understand some people’s frustration. But before we throw the text out with the bathwater, allow me to offer my take on the value of flavor text in general.

As stated previously, boxed text takes some of the work off the GM’s hands. It gives him something to tell the players about their surroundings in an easily presented format. Granted, some boxed text can go on for far too long. A few sentences should be enough. Alternately, you can have a short initial description that breaks and allows for player agency. Then, if the players “trigger” the next part, there can be a followup box of relevant text. 

Of course, the flavor in flavor text is often the mood-setting, or atmospheric, descriptions to help people feel more immersed in the game. While this is a laudable goal, I often feel it is over-emphasized in boxed text. I mean, that clever imagery about the moon through the dead tree branches is cool and all, but not if I have to listen to a minute and a half of it before the GM tells me there’s a werewolf under the tree.

Another reason for prepared text is to help the players. Not just in setting the mood or presenting their initial impressions of their surroundings, but in filling in what I call the “perception gap.”

There is an old-school GM-ing conceit of “The player never expressly said they checked X, so they don’t know Y.” While this is fair as far as it goes toward allowing for player agency and their characters to stand or fall by their choices, there is a range between what one overtly sees or hears, and what one might realistically perceive if one were actually there. This is the perception gap.

Articles and books about adventure design often address this with advice like using all five senses when giving descriptions, not just sight or sound: What do they smell? Is it cold or hot? Dry or humid? and so forth. The idea is it helps the players visualize and become immersed more in the game. This is good advice, but has two potential pitfalls.

First, this can be a lot more work for the GM. Hence the previous point about using a published module. If an author has already incorporated some of these descriptors into the text, then the GM doesn’t have to. I am a proponent of not overusing this technique, though. Every room shouldn’t have a laundry list of sensory stimuli. Use the technique sparingly, and usually only when it will have an impact. Which brings me to my second point regarding the perception gap.

Providing more detail to players via prepared text not only for atmosphere, but to provide hints relevant to gameplay. A rotten smell might indicate zombies in the cellar, or a drafty room might hint at a secret passage. These are things that a person would probably notice just by being in a place, but may or may not consciously register. Some may be less obvious than others, but still definitely perceptible. Unless you want to train your players to repeatedly stop you and ask for input for each sense, scattering some of these details in subdued ways gives players a chance to follow up on these cues without necessarily smacking them over the head with them. The tricky part is to make such cues normal enough in one’s “GM patter” that metagaming players don’t pounce on every adjective, but distinct enough that they were given a fair shot before saving throws are called for. This sort of lexical balancing act can be tricky, which is all the more reason to use text that was prepared ahead of time either by the GM or by a published module’s author. 

As with any aspect of game design, boxed flavor text can be done well or poorly or anywhere in between. When it’s done well, I feel it can be a real asset to the players’ experience, both for setting the mood and filling the perception gap.