Let's talk about fluff.
Fluff and crunch are terms used when talking about TableTop Role Playing Games (TTRPGs). Put simply, crunch is the rules of the game and fluff is everything else. Consider the following example. Say you're levelling up your TTRPG character and I give you the following text as an ability you could pick on level up:
Deal 1d4+1 points of damage to target creature. For every two levels above first, deal an additional 1d4+1 to target creature.
That's pretty mechanically clear, right? I left off some boring stuff about range, and who can use this ability, and resistances, and so on, but that's pretty mechanically clear. This is an ability for doing 1d4+1 points of damage to someone, at a distance. That's the crunch. It tells you exactly what something does, inside the framework of the game.
Something a little more subtle is that crunch relies on the existence of other mechanics very intimately. For example, this description only actually makes sense if you understand the concept of damage, which further only makes sense if you have defined hp and injury and death (and implies a lot about what the game thinks is important to track). Sure, damage makes intuitive sense because games have been using that term for many years. But this ability also depends on knowing about levels, what "1d4+1" means, and heck there's a whole bunch of other rules for using abilities that are just assumed (full of complex phrases like "standard action", "somatic component", "attack of opportunity", and "concentration check").
All of those rules exist. But you have to either know them or go looking for them to fully understand this one ability. While you can make a pretty good guess as to what this specific ability does, the full context may be lost.
So that's crunch. Alright, here's a challenge question for those of you who didn't immediately recognize this ability: what does this ability actually do?
Yes, obviously it does 1d4+1 damage to a single target. But what actually is it? Is it a gun? A magic spell? A psionic blast? Do you kinda just rewrite reality and your opponent is suddenly injured? Do you tell a joke so bad it causes brains to bleed? Remember that TTRPGs exist on two layers always: the hard mechanical reality that this ability lives in and the pictured imaginary reality that exists in the heads of the players. That's the very definition of fluff: everything that helps bridge the gap between the two realities. Let's zoom out a little and look at the full spell description:
A missile of magical energy darts forth from your fingertip and strikes its target, dealing 1d4+1 points of force damage.
That's right, I left some parts out! I'm the worst!
The phrase "A missile of magical energy darts forth from your fingertip and strikes its target" adds nothing to the mechanics of the spell. If you were to reduce DnD to a coldly numeric version, that text would be meaningless. Wasted space. It arguably adds nothing to the simulation of the game world. And yet the designers of DnD 3.5 still felt it was necessary to add that description. But if the descriptive text adds nothing, then why does the 5e version STILL have descriptive text?'
Because you can now picture it. You now know it's a spell and have some description to work with. It's easy to imagine a wizard, witch, or sorcerer casting a missile of magical energy (perhaps a magic missile even, if you will). It has focus now.
Fluff matters a lot for TTRPGs and game designers know this. The power of tabletop roleplaying is in the roleplaying. It's all about what you see in your head, that image of your character standing tall and being awesome (or being sad and dramatic and gay, whatever you're into) that drives you to keep playing. And given just a cold mechanical description, it can be very hard to translate that into a good mental image of your character.
Fluff vs crunch is an important concept for game design, and it's common to measure tabletop systems by how "fluffy" and "crunchy" they are. But rather than pontificate about the benefits of either being better, I want to reveal that I've really buried the lede on this one.
Yup, that's right, I made a whole-ass video game to prove a point for a blog post. I even rewrote it from scratch in Godot 3.5 after finishing it in Godot 4.0, to avoid a web export bug that would've prevented you from playing it in browser. That right there is the value you pay me for (actually no one is paying me).
So, go give it a quick play and tell me what you think. It'll take you a minute, tops.
Arrows keys are input.
Seriously, try it before you go any further.
No, really.
Well.
It's just a bunch of numbers, right?
What if I told you that you just played (approximately) level 1-1 from Mario? Really! Type "reveal" into the top text bar and hit reset if you want the proof. The top pair of numbers are your current position, the second row is the flag position, the third row is all pit traps that will kill you, the fourth is the boundaries of the game world. Everything after that is a list of all the in-level blocks.
The numbers are a perfect encapsulation of everything in a video game that could be considered "crunch": the actual hard mechanics. Everything else (the graphics, the sounds, and the music) is the fluff. Crunch is what happens, fluff is what tells you how to understand what happens. The game is perfectly playable in this numerical state. Completely and utterly playable, same win condition, same rules. But it's impossible for a human to understand and process the information presented this way.
Ah, but didn't I cheat a little by not defining the numbers? Well, does Mario? Does Mario give you a detailed list of each and every property of the game, perhaps in some form of flowchart? No, Mario just kinda kicks you into this level and expects you to figure it out. And you can partially because it's a well designed level, but partially because of excellent fluff usage.
Fluff explains the rules of the game far better than a tutorial could. Looking at the sea of numbers, it is unclear how to interact with the world and even if there is a world at all. But given a single dude (Mario), it becomes obvious that you are the dude. From there, it's very natural to assume that you can move around. Every human has expirience with gravity, so if you jump, you know that you will come down.
Good fluff can serve as a replacement for a tutorial, by letting you know that the video game reality resembles the real world in key ways. For example, Tears of the Kingdom is a very crunchy game in that it has a lot of different mechanics to learn and master. But because most of the mechanics are physics related, it is easy to have a very intuitive understanding of many of them (wheels spin, blocks fall, fans push, gliders glide, rockets zoom, etc).
Crunch is important in that crunch is what makes it a game. A crunch-less game is somewhat difficult to present because a pure fluff game exists purely in the mind of the viewer. Any attempt I would make to define it would fail because that would be crunch. Perhaps a novel would count? Or a conversation? Or something more esoteric and ill defined. And while those things are worthy and fun, so is existing within and slowly mastering a framework of rules. A framework of rules can be intellectually stimulating in ways that pure fluff lacks and encourage different ways of thinking about problems. Also games are fun to design. Game design should be fun for the designer! That's the first rule!
As with all aspects of game design, the real challenge is to nail the balance of fluff and crunch. The exact balance you want depends on what kind of game you're trying to make. A more crunchy game is more 'hardcore', in that mastery of the rules is often the goal. This is very common in strategy or tactics games, where the fluff only ever serves to be an iconographic representation, quickly and easily taken in by human senses. In Civilization, no one complains that the armies appear to be 1-10 guys, when they really should be thousands of guys big. It represents the troops in a visual manner. Seeing a unit with a bow gives you a very good idea of what it's is capable of, far better than a numerical description would. Scanning a battlefield and spotting that 10 tiles have troops is a lot quicker than comparing detailed troop lists to figure out if you're outnumbered.
Good fluff is wonderful and captures the imagination and senses. Bad fluff limits imagination. Remember that magic missile example? It could just as easily be blasts of lightning or fire or shadow or pure energy or a gun or a bow or a magical bow made of pure light that shoots lightning and fire and pure energy. By specifying what the attack looks like, descriptive power is taken away from the player. And that hems in possible character options. This gets even worse when the crunch leads some abilities to be "optimal", forcing a decision between being true to the character and being mechanically powerful. While a lot of people have strong opinions about which of those two they prefer, I argue that a well designed system should prevent that choice from ever needing to occur.
Did you notice I played one other trick with the magic missile description? My first version of it left off a crunch word, which I then included in the second version: "force damage". Yup, "force" is actually a crunch word because it's a name for a rule that directly interacts with other rules. Had it been "fire" damage, then spells like Resist Fire or a fire dragon (that is immune to fire damage) would have direct mechanical rules interactions with it. Magic missile being force damage is actually why magic missile is one of the most reliable spells in the game; almost no core monster manual enemy resists force damage and force damage works on ethereal creatures (another crunch mechanic with an evocative name).
Evocative names are a clever trick fluff plays on you. See, because crunch is invariant to fluff, crunch is also invariant to naming. I could just as easily rename "force" damage to "type 4" damage, without changing anything about the game. But now large parts of the game stop making sense. If "type 1" damage is physical and "type 2" is fire, why is Resist Type 2 Damage a much lower level spell than Resist Type 1 Damage? From a game design perspective, it's because Type 1 damage is much more common, so that's how the balance works out. But if they're descriptive, it makes sense that Resist Fire is easier to cast than Stoneskin and is very clear why they're separate spells.
Back in 2011, Pathfinder released a book called Ultimate Combat which contained a playable class called "Samurai". And certain corners of certain forums I was on at the time lost their minds, because "Asian things don't belong in a fantasy world". And while there's several interesting and worthy things to say about the use, abuse, fetishization, and rejection of Asian stereotypes in gaming (particularly given the traditional attempt to recreate an idealized white European medieval aesthetic that never really actually occurred), I am not the right person to write that analysis by any metric. I can call it out though. That's fucking dumb, get over yourselves, people from a decade ago.
But note that people were objecting to fluff. The crunch of the class was fine. Good mechanics, decent niche in design space, and perfectly in line with everything else in the system. As someone pointed out, you could rename the class to "knight" and no one would've taken issue (which just highlights the stupidity on display). So why didn't players just think of it as a "knight", if it bothered them so much?
Fluff can make you lazy. Fluff can be restrictive. People learn to be bound by the fluff.
One of the best and weirdest tricks for a GM to learn is reflavouring. That's when you take some crunch and make it into something new, purely by changing the fluff. It's a lot quicker than designing a new monster's worth of crunch! And no one will notice if your "short, green, and very ugly goblins" have the same statblocks as the "stout and hearty human guards". The players never see the statblocks!
I think that as 5e seems to be most people's introduction to tabletop roleplaying, it needs to stress this as a skill more. I distinctly recall coming to a GM of an upcoming game and laying out my orphan bard's character concept. I ended it by mentioned my planned subclass (college of whispers) and was immediately asked about my time in the "college of whispers", something that I had assumed was just an evocative subclass naming scheme but apparently is an actual thing that exists in the game world. I was a little put out. Going to college would've wrecked my planned backstory (don't worry, we sorted it out eventually). Moral of the story, names may not change mechanics, but they sure do have power to shape how people think about your game.
All of the best characters I've ever seen involved reflavouring, from the minor "pretend this oboe has the same weapon stats as a greatclub" to the downright major "throwing out all the established fluff about this class and doing something new". And they fucking owned. Learn reflavouring as a skill. If you're reading this right now, your homework for your next TTRPG character is to identify some fluff in the rules and then do the exact opposite in your character description. Fuck it, there are no gods, rename your spells! You don't use "magic missile", you use "Gandalf's powerful doomsday blast".
Remember when I mentioned a couple posts ago that I once wrote a tabletop roleplaying game? Well apparently I'm just going to keep making blog posts where I talk about what I was thinking about while I wrote it, until I guess I just post the damn thing or something. Rule 2 of that project was no fluff (remember, rule 1 was wealth is a part of levelling). As a deliberate counterpoint to the issues of DnD, I went out of my way to try to avoid fluffing anything. The only fluff in there is the names of the abilities because they have to be called something (and they're as lackluster names as I could make them). I was hoping that the added upfront challenge of describing your character would reap rewards down the line in terms of personalization. Jury's out on if that worked.
Reflavouring is a good skill for video game designers, too. You should never write the same code twice, and reusing code is exactly like reusing crunch: if you change the sprites and descriptive text carefully, no one will notice.
As with the tabletop space, too much fluff causes problems in the digital space. One of my criticisms of Civ 6 is that the tile graphics get too busy and it can often be hard to tell what specific type of tile each tile is and what improvements or buildings have been built. This is a failure of fluff by overuse of fluff. When you start making fluff for its own sake, pretty graphics for the sake of pretty graphics, fluff stops communicating what it needs to.
You can also have poorly designed fluff. People have often joked about the "insurmountable" ledges and tiny trees in Pokemon that look like they could be easily bypassed by a child. Sure that's minor. But it still causes confusion in interpretation of game mechanics (why do I need a special move for my Pokemon to chop down a tree when it has giant scythes for hands).
I would argue that a huge issue here is the quest for realism in gaming graphics. Because making the graphics better is easier than making every decorative object intractable, we've ended up in a place where games have to launch with "detective vision" that highlights interactive objects, because otherwise it's impossible to tell what is merely decorative and what is part of the game. The entire time I'm playing Divinity: Original Sin, I have the 'tab' key held down because that highlights interactables. This negatively impacts the game. It hurts my finger and causes me to just zoom from highlight to highlight without taking anything else in, missing out on appreciating the worldbuilding of the fluff. And poor design can blur the line between detective vision and reality.
I'll always argue for stylized graphics over hyper realistic ones, and one reason is that stylized graphics have a lot more control over what draws your eye. It makes it easier to use fluff effectively to communicate options to the player. The purpose of fluff is to render the crunch in such a way that humans can comprehend it easily. If you want to write a novel or draw art for its own sake, do that. In a game, everything has to bend towards being understandable.
Balancing crunch is definitely a skill and you can read any essay on game design or game balance to get a feel for it. Balancing fluff and crunch together is a lot harder, but worth it. By making your games easier to understand, you save yourself effort on handholding through the tutorial and increase the player satisfaction of figuring things out on their own.
I think this also gives me direction as a designer. To be honest, I love doing crunch (designing mechanics is fun!) and I dislike doing fluff (there's just so much of it). My dream would be to partner with someone who is the opposite (any maybe actually good at drawing or music). Maybe someday! But for now, I'll continue focusing on doing minimum fluff games, and only putting fluff in where it really counts.
I'm going to start a new bit. One of the things that annoys me about the new web is that discovery is so weird. I hate algorithms recommendations and really just want personalized recommendations from a friend. So I'm going to practice what I preach and link one thing I find cool at the end of every blog post. No particular connection or theme! Just one thing I find cool!
Today's does have a connection though, because it is The Alexandrian! A TTRPG blog full of GMing advice, all of it built off years of expirience and well thought out. Even if you're just a player, skimming some GM articles might help you give better feedback to your GM or even improve your ability to play along.