On the Definitive Inadequacy of Boxed Text

I made a statement last week I believed to be a truism. Other people did not and made some comments and asked some questions.

Let's address some of those comments.

Is boxed text bad?

Yes. Yes it is.

Why? Can you prove it?

"You feel afraid and grab your weapon hilt"
"Uh, I'm immune to fear and as a warlock, I don't have a weapon?"
We are not talking about that 90%. What we are going to look at is why "Good" boxed text is bad. Let's look at some facts, facts beyond the numerous, public, repeated, comments about the response to boxed text.

Human beings have limited attention. They can only spend so much.

Listening to boxed text requires FOCUSED ATTENTION: the cognitive process of selectively concentrating on one aspect of the environment while ignoring other things.

The average length of human focused attention is between 6-10 seconds.
More data from the Associated Press.
A textbook explaining these terms, and various theories about information processing.

Focused attention is different then sustained attention. Sustained attention is our ability to focus on tasks. This type of attention allows us to produce consistent results. Sustained attention is what allows us to play D&D for four hours, watch a movie, play a video game or read a book. These types of activities involve participation of our brain which is why we are able to maintain our focus.

Video games and movies? Aren't those passive activities? 

No. Cognitively, they provide multiple sensory inputs and can cause the brain to respond to stimuli. They are specific artistic creations to present only articles of central interest. That is literally the skill of film-making.

But surely I can have sustained attention to boxed text?

Sure, you can force yourself to focus on the dripping water in the sink without thinking of other things, or perhaps meditate by clearing your mind. It is difficult. It is a skill.

It is easier to do, if provided with things that are engaging. But even with videos and other multi-media presentation, you lose 1/2 of all people at 60 seconds. And that's with video!

Can't boxed text have information that is processed as "central interest" and therefore might draw our attention better?

In general, no. "Central Interest" to our auditory system is going to be things like exceptionally loud noises, predator noises, a child's cry, or perhaps information that affects our status quo ("I take damage?!") most boxed text, by definition, is not going to contain those things.

Isn't boxed text the best way to present the information?

Sure, if you only want an absolute maximum of 4 things (cambridge link) to communicate in under 8 seconds, you're golden!

HOLD UP! When a module eschews boxed text, then you will have a module that doesn't have a clear separation between information the players have and the Dungeon Master knows! You need that. 

Well, yes. 

What I'm saying is boxed text, even well written boxed text is the absolutely worst way to accomplish that. Just because a rocket powered sled is faster than walking doesn't mean its an ideal way to travel to work.

The primary purpose of "good" boxed text is that it provides all the information available to the characters, separate from the information the Dungeon Master has about the contents of the room. 

Putting this information inside the interior of a block of text is bad information design. It is decent design for presentation but terrible design for reference. Room information must serve both purposes.

See, the boxed text clearly differentiates between what the players are immediately told, from what they can discover. But it buries the list of things that the players are told in a paragraph sized box of text and then separates those items, from what happens when the players actual begin manipulating those items. So first you have to find the pedestal in the boxed text above, and then you need to find what manipulating it does (or what's on it, etc.) in the block of texts below. That's bad design. 

Brendan suggests using a highlighter to overcome this problem. Well, that's how my dad did it, it should be good enough for me, right?

We shouldn't need to fix the design by marking up the books, the design in the module should serve its purpose!

Well, What am I paying for then!? I don't want a skeletal outline, I want to experience this module! All Set Design is, is an outline.

Well no. A single example of my room, from a personal notebook is not representative of what set design looks like in a product, or a module. That is partially what I'm trying to communicate. I'm working to get some examples of what I mean when I talk about set design out in the wild, rather than a simple one-off example of how a single room is keyed.

I am not suggesting "Bullet points and Bold text" to make the distinction. What I am suggesting is designing the information so that setting and game information can both be presented and referenced in an efficient, pleasant manner.

But how will you communicate the creative vision of your product without boxed text and background information?

To be absolutely clear, to hell with the novelist aspirations of the module writer.
Regardless of their answers, the Hermit raises his hands, awakening the 10 skeletons hidden beneath the earth of the hut, and launches into an attack, attempting to "recruit" the party into his undead army. -Legacy of Savage Kings, Pg 7. Harley Stroh
Not boxed text. Does a fine job of communicating something awesome, besides "monsters attack".

Making people listen to prose, (or read fiction in a rule-book) is possibly the worst way to transmit information about a creative vision or setting. I've often found the only defenders are those who, well, write prose or fiction in a rule-book to communicate a creative vision or setting.

It's not 'the worst way' because of some abstract dislike of fiction. I love fiction. It's 'the worst way' because it is very bad at accomplishing the goal of transmitting that information. 

Short fiction is often skipped, only read by the Dungeon Master, or is unmemorable and not very good at it's stated purpose. (Quick! What's the plot of any book you've read? Now look at your RPG shelf and see how many examples from the fiction you remember reading!)

But beyond those reasons, it literally isn't relevant at the table. What you can buy on the equipment list is. And it is those things, the literal game objects that create the creative vision and flavor. People who play Dungeon Crawl Classics remember that their character started out with a chicken or a shovel. That becomes the take-away of the creative vision.

I want my monies worth!

I agree. I want that same thing from a module you want. I want it to be creative. I want it to be evocative. I want it to provide setting, environment and theme. 

Boxed text is a rip-off that doesn't do those things very well. I want my monies worth from module also. I want something better. 

If you like extensive, detailed, theory posts like this, then take the opportunity to become my patron and make it more feasible to spend the time writing!  Hack & Slash Patreon Campaign.


  1. I absolutely hate boxed text that rambles for paragraphs and closes with "And a Bholder flaots in the air above the glowing altar" there is no reason at all to mention anything else ... the beholder is what folks are going to notice first, they aren't counting doors, they aren't looking at the mosaics on the wall they are noticing that big ball of death.

  2. Great Post , I had never looked at boxed text in such an analytic way. Very useful info

  3. Courtney wrote: Brendan suggests using a highlighter to overcome this problem.

    Well, sort of. The takeaway I hoped to communicate with that post was:

    the organizing principle should not be the nature of game entity (monster versus treasure, for example). Rather, features with higher referee immediacy should be emphasized.

    Highlighting is just a convenient want to retroactively apply this approach to products that already exist but have suboptimal information design.

    When creating new material, I use either underlining (if handwritten) or bolding (if word processed) to emphasize the immediately relevant features.

    I do think that the "set design" approach (especially if sprinkled with descriptive phrases for atmosphere) is superior to boxed text, and accomplishes, practically speaking, what is needed to run an area (which boxed text does not).

    I remain to be convinced that this outline approach is aesthetically superior to prose with emphasis though. We need some full implementations to see, I think, beyond statements of principle.

    1. I kind of think that aesthetic superiority comes from the stroh method of adventure writing.

      Not descriptive text, but evocative play environment.

      "The axe is frozen in the barbarians hand. If the players remove it, they have to break off several frozen fingers. "

  4. In my experience, Text box = Lazy DM. The box o' text gives the impression of a canned encounter, ready to eat, no further preparation necessary for the DM. So rather than preparing for the module by reading up on the encounter, digesting it, and figuring out how to portray it to my players in a meaningful fashion, I'd just read the box o' text, on the assumption that everything would flow naturally from there, no matter who my payers were or what their intentions for the encounter.

    In more simplistic encounters, this worked out just fine. But, often enough, it led to me scrambling on the spot to figure out how things are actually going to go down, which means bored players twiddling dice while the DM does work that they might have done beforehand had they not been lulled into a false sense of preparedness by the box o' text.

  5. This is the whole truth and nothing but the truth.

  6. I've been reading - or really, reading past - a lot of boxed text lately. While I think your rule is correct, Robert Wiese's "Wreck Ashore" is an exception and therefore a good test. The boxes therein are very small, no more than a short paragraph, so the attention span problem doesn't arise, and they're only used in one abandoned "static" area of the module so no immediate information is left out.

    Having chewed on that for a while, I suspect that Wiese's good use of boxed text wouldn't work if it became a convention. ("Aha! He's reading aloud, this is a safe room!") Although that in turn would breed anti-anti-boxed text scenarios...

  7. I still think there are some chinks in your argument. Here's my response:


    1. This came up in the other conversation on G+.

      Boxed text, traditionally, requires focused attention. Much like if I were to read ten names out of the phone book, it lists items in the room, creatures, and various features. Attempting to assimilate auditory stimulus in this way requires focused attention. It is exactly what focused attention means - paying attention to parse data we are being given. It is exceptionally hard with transitory media (audio/video) for us to maintain that focus.

      It is apparently super-popular to try to redefine what focused attention is, to try and make it resemble sustained attention! Anyone can listen to a story, right? A podcast? That requires sustained attention!

      But those things are conversation or narratives. Those are exactly the kinds of things sustained attention can follow. Yes focused attention is paying attention to threats, noises and dangers in the wild. It is also what's requires to remember a phone number you're given, doing addition, and parsing information you are given. This is what boxed text is forcing you to do

      But the point was made that the boxed text being referred to wasn't providing a list of features and objects of interest in the room. It was "Setting a scene."

      Now, I have never met anyone in thirty years and thousands of people I've played with that sat around and read descriptive paragraphs to 'set scenes' thirty or forty times a session. But that isn't the whole of gamers, and perhaps there's someone out there that plays role-playing games because the game is frequently interrupted with being read to out of a storybook.

      That isn't why I play the game, and I've never met anyone who's played the game that way, but clearly it's possible that dude exists, so whatever. Boxed text is probably just swell for that guy.

      But as a Dungeon Master that requires content it has all the drawbacks I listed above. And since I, nor any player I've ever had the pleasure of gaming with has ever request more sitting around and being read to by adults, the boxed text has all the problems listed in the article.

    2. Apparently my response is too long to post here, but since you've also made this same reply over at my blog, I've placed my response to it there:


    3. Tad's Comment: "I’ve done a bit more research on attention and have discovered that your use of “focused” attention was correct; I had been mistaking the “focused/sustained” distinction with a different distinction, that of “exogenous” vs “endogenous”.

      Exogenous (exo- outer gene- origin/cause) attention is a stimulus driven form of attention orienting. We are hard-wired by evolution to attend to changes in our environment, such as motion and loud noises. This is not something we are generally in control of, our attention is just naturally attracted to certain stimuli. Endogenous (endo- inner gene- origin/cause) attention on the other hand is an intentional, goal directed form of attention. This occurs whenever we choose to orient our attention on a particular item or task. It is executive function. Focused and sustained attention both reply upon the executive capacity of endogenous orienting. This article has more information on this.

      Anyway that all aside let’s get back to the argument at hand. So let’s now grant the first premise of your argument, viz,, that listening to boxed text requires focused attention. What about the other premise, that focused attention is not sufficient to allow us to pay attention to box text? I think this is an exaggeration. One way to demonstrate this is to show that if true your argument’s conclusion would prove to much. For it would turn out to be the case not only that we can’t pay attention to boxed descriptions, but to any descriptions at all, including those room descriptions that you want to give your players “conversationally”.

      The reason for this is simple. Anytime one is required to pay attention to a list or set of details they will need to use the same array of executive functions (e.g. endogenous orienting and memory). This is true regardless of what format the details are given in (in a book, a play, a conversation, a speech). Likewise there will always be the threat of “attention capture” by other stimuli as well as memory load limits. So the upshot is that if our executive capacities are not fine tuned enough to allow us to listen to a description read aloud then neither will they suffice to allow us to listen to the same description given in a different form.

      “Ah”, but you say,”that’s just where you’re wrong, conversations and narratives involve sustained attention, not focused attention, so it’s all good!” I don’t know whether or not in general this is true (on what basis do you make these claims?). But regardless when offering an itemized list of information the problem will be the same.

      So say you are describing a room to your players without box text. Either you do this all in one go, listing all the information they have immediate access to, or you break this up into bite-size chunks, allowing them to interject with questions and comments. If the first, how is what you are doing really any different from reading a box text? If the second, admittedly you gain the advantage of reducing information overload but at the expense of allowing even more opportunities for the attention of your players to wander. These interjections creates space for all manner of rabbit trails. Players might get fixated on one thing mentioned and by the time you move on they are no longer listening but thinking about that object. This is a familiar experience. It’s easy for our attention to drift in a conversation. Frequently I find myself focusing on a point someone has made and a minute later find I’ve not been paying attention to anything they’ve just been saying. Or that instead of listening to what they are saying I’m busy trying to think up a response to something else that they’ve said. This “selective” focus means that in a conversation we don’t always attend to everything being said. Presenting all of the information up front gets around this problem but has the disadvantage just mentioned. So whichever way you go I don’t see how a conversational model offers any gains in player attention over that of a boxed text model."

    4. My reply: "Except, your reasoning is all backwards.

      We aren’t saying “here is boxed text, what will happen when it is read?” What we are doing is looking at people listening to boxed text, seeing that their eyes glaze over, that they aren’t processing or listening to the information, and generally waiting for it to end. Then taking that data, we are looking for why.

      The why is the problem with focused attention.

      The reason “expense of allowing even more opportunities for the attention of your players to wander.” isn’t a problem, because them investigating a room, asking questions, and following the various rabbit trails is literally playing the game. They like it! It’s fun to investigate. Their attention drifts to whatever they are interested in, when they investigate it.

      Look at it this way – They may choose to not look around the room. Then if they wander into some shenanigans, it is their fault. It gives them agency,

      instead of the often mentioned situation where they are ignoring the boxed text because they are eager to play and not really processing what’s being said for biological reasons anyway, and you either have to re-explain it to them after, or deal with their hurt feelings because they got burned because they wouldn’t jump through the hoop of being quiet and attentive in a room full of their friends, instead of, you know, playing D&D"

    5. "My reasoning" just follows the structure of your own argument. Your claim, recall, is that 90% of box text is poorly written, but that even the 10% that isn't is bad. The reason being that it requires focused attention to listen to box text and we haven't got this in sufficient enough quantity. The argument structure is deductive, not inductive.

      On the other hand if you are attempting to support the second premise of your argument based on the inductive inference that whenever you've read box text to your players in the past their eyes have glazed over you've got a whole new set of problems. The first of which is that by your own admission 90 percent of the examples of box text out there are bad and this fact by itself might account for your players's inattention, as well as the widespread antipathy toward box text within the OSR community. A second problem is that there are quite a few people out there who have not encountered problems when reading box text to their players. In those forums you linked to earlier there were plenty of folks who came to the defense of box text. So as it stands, any inductive inference from the evidence you sight would just amount to a hasty generalization.

      As the other point (about player agency and descriptions), fine, but you still need some some device that answers the question "what do your players see when they open the door?" Unless you decide to not tell them anything about what they see until they ask you about it presumably you will at the very least need to list off the general features of the room. My point was simply that your argument about the limitations of focused attention would just as easily be a problem for doing this as it would for reading a box description.

    6. A) It’s Cite
      B) that reply doesn’t contain the fabled example of “good” boxed text.

      The reason my argument on the blog starts out with the example of “We aren’t talking about bad boxed text” is to preemptively address the argument, “We’ll, everything you are saying is true, but only about bad boxed text.” Personally, I’d love to see an example of good boxed text. It’s amazing how hard it must be to locate, being the trouble people have producing it.

      C) Many people will defend things, because they are deluded or unable to see the negative effects of their choices. Ever been in a drunk tank? Met a drug addict? How about know anyone that is a bore at parties. Their support of it doesn’t indicate any inherent value, compared to observational complaints about attention wandering, a real, physical, drawback noted again and again.

      D) It is clear that player led questioning is a simple and engaging technique for the players. You have a list of what’s in the room. Then they dynamically engage you in play. (DM:you open the door, a chamber lies beyond. P1: how big is it? DM: narrow, perhaps fifteen feet across, and about fifty feet long. P2:how many exits? Etc.) As opposed to making them sit still, be quiet, and read to them, making them wait till they play again.

      Surely you can see how a Socratic method is effective at transmitting information? The players are asking what they want to know. The answer to their question, is presumably, short, containing the information they are seeking. As opposed to a long difficult to reference block of text containing all information front-loaded into something that is going to be read by someone who’s not a professional orator.

  8. I don't like descriptive boxed text, but I'm happy to have box outs that contain specific information that's very pertinent. Of perhaps setting it out so that you have bolded titles of each paragraph like: Pressing Details (creatures, NPCs monsters, traps), Description (what is the room like), Physical Clues.

    A quick rundown and refresher above a more organic, longer and more evocative set of information which helps in retention / digestion by the Storyteller during the preparation phrase. After all, quick info is needed at the game table. Informative and evocative info (though still not 'narrative') is needed for absorption prior to it.

    At least, that's how it works for me.


Related Posts Plugin for WordPress, Blogger...