No Foolproof Illusions

This post has some required reading: a blog post from Swedish designer (?) Sandra Snan called Blorb Principles.

[By the way, Sandra's blog is absolutely bonkers. She's written about blorb a lot but you'll have to hunt for it. Good luck]

Additionally, you may also wish to read this post from Rise Up Comus and/or this post from Technoskald's Forge. They are both good, although not necessary to follow my line of thinking here.

I like Sandra's blorb principles. When I first read her post, I felt... relief. For many years, I've felt a sharp and uncomfortable distance between my own playstyle and the philosophies described by my colleagues and other popular designers. Sandra's post was the first time I saw someone clearly articulate a set of preferences I've long held but which I couldn't effectively advocate for on my own. It feels nice to see your own philosophy given a name, and to finally have a way to easily connect with like-minded GMs.

But part of why Sandra's post instantly clicked for me is because she was describing things I already believed, techniques that I already rely on. By far the most common response I've seen to the Blorb Principles is still outright confusion. So just like many others before me, this post is my own effort to explain why I prefer a Blorby approach to the alternatives other people offer up. Maybe this will help it make a little more sense to some people.

The Burden of Prep

Most people currently bearing the GM title first got that job from a modern WotC edition of Dungeons & Dragons. I know I did. So let's put ourselves in those shoes for a moment.

When you're new to the job, you probably put a lot of work into it. It's normal for rookie GMs to pour themselves into worldbuilding, crafting NPCs, drawing maps, making stat blocks, plotting out some epic quest, curating playlists, setting up fancy terrain or a detailed VTT backdrop, and especially, writing out fluffy descriptions of everything. This takes a lot of time and energy.

Committing that level of prep before you kick off the campaign might seem feasible to the amateur GM. "I'll start the campaign once I have enough material to be ready." But once the ball gets rolling and you're meeting your friends to play every week, it quickly becomes clear that this level of prep is unsustainable. You can't commit to that same standard every single week. And so comes the burnout.

In many ways, improving as a GM is the process of learning how to optimize your prep. You have a finite amount of time between sessions to spend on prep, and so you must develop a prep strategy if you want to keep a campaign going. Hell, I would argue that "coping with the burden of prep" has become one of the biggest drivers of game design itself in the history of this hobby.

Let the following be postulated: it is impossible to exhaustively prep 100% of the GM-originating elements of the game. Therefore, all GMs must do some amount of improvisation no matter what.

This is important because it helps us to see through a common flawed assumption: that you're either a prepper or an improviser. Obviously, no GM is fully one or the other. Because we acknowledge that all GMs must employ improvisation, the real questions are then "how much do you leave to improv?" and "which elements do you leave to improv?"

Quantumania

There's a million tools, techniques, and pieces of advice people have for this. But there's a specific strain of practices and game design elements that I think fall under a shared ethos. It's the main thing which I'm seeking to contrast myself against. I'm not really big on semantic perfection so I'm just gunna loosely lump together a bunch of things which I think you can agree are broadly related:

• "Quantum" states for game information, whether it's the location of a creature/object or the answer to a question or whatever
• Procedurally-generated content (e.g. rolling on random tables to populate the world or to generate encounters)
• Non-binary results of die rolls (e.g. "succeed, succeed with a complication, fail, fail with a complication" that sort of thing)
• The oracle die / luck die / the die of fate
• Anti-canon worlds
• Shared narrative authority (e.g. "let the players add to the fiction")

Of course, not every GM who favors some of these ideas favors all of them. Quite the contrary. I know some folks who swear by their reliance on random tables but who wouldn't be caught dead letting their players contribute an improvised piece of set dressing in the dungeon.

Furthermore, when I say that I'm contrasting myself against this general approach to running games, that's not to say that I reject it entirely. Trust me, I make frequent use of improv, procedural-generation, and the glorious die of fate.

However,

Taken together, these ideas and techniques solve the burden of prep chiefly by avoiding it. I've known many a GM and designer who advocates an approach to gaming that can basically be described as "learn how to improvise anything and everything." Which, I'll admit, is an incredible skill to develop. I've seen it in action and it's really impressive. By honing the creative skill of spontaneously generating new ideas from your noggin' and combining it with a robust set of tools, procedures, and mechanics that fill in the rest of the gaps, it may just be possible to never prep anything again!

…But that doesn't mean the results would be equally satisfying in play.

Intelligence-Based Strategy

My co-writer referred to this as a "house of cards" approach. The GM has a deck they draw from as they go, adding each new piece to a structure they're building one card at a time. It might make something amazing, but it is fragile.

This is where I think you lose some folks. Their quantum-centric playstyle is working just fine for them, so what is this alleged fragility? I think that Sandra's writings (and those of many other Blorb-enthusiasts) have focused a bit too much on the importance of the imaginary world's integrity. Sandra herself originally wanted to rely on the word "simulation" for her theory, and she still invokes that idea often. But to be honest, an internally-consistent and impartial "simulation" is not really what I like about blorb.

After all, a sufficiently detailed and consistent procedure for generating content is also a tried-and-true method for creating an impartial simulation. Think of the famous tables in Traveller, for example. You can literally run an entire campaign off of those. So while I do value referee impartiality and I don't want my game to be compromised by the inherent biases of my human brain, I also know that I can accomplish that by just outsourcing decisions to dice and mechanics. That's what Sandra refers to as the "Tier 2 and Tier 3 truths" in her original essay. So clearly, maintaining the simulation's integrity alone is not the only value in blorb. What is it that makes Tier 1 truths so important?

For me, the far more compelling strength of blorb is how it supports challenge. A fair challenge requires that players have agency. That the outcome of their victory or defeat can be directly tied back to the decisions they made, but also that those decisions were meaningful expressions of their ability to think critically. It's that last part that a lot of people miss. Just because you've given the players multiple choices to pick from doesn't mean you've succeed in presenting a fair and meaningful challenge. 

Let's say you offer me two doors to pick. The one on the left conceals treasure and cake and victory, while the one on the right conceals a hungry dragon who'll kill me. If they're completely identical from all outward appearances, with no differences in sight, sound, smell, etc. then what's supposed to inform my choice? It's true that if I were to pick the left door then I would be victorious, and you would be technically correct in saying that "your victory is the result of you making the best decision!" But it's not like that decision had anything to do with critical thinking. It was arbitrary. That's not true agency, that's just luck.

Players can't make meaningful decisions without sufficient access to information. And the more game elements that you leave up to chance or improv, the less information the players have to rely on in order to inform their plans.

I'm not trying to invalidate your desire to avoid prep. You can and should still cut down on prep in lots of places. Just not everything. What Sandra calls "things with points and sharp ends" is in reference to all the elements that are salient to challenge. These are the ingredients of the game which benefit from being prepared ahead of time. The location and description of spiderwebs in the vampire's castle? Probably not. The location of the vampire's coffin? Oh yeah, let's get that one nailed down first thing.

Let me try to paint a detailed picture of what happens if you substitute prep for some other solution when it comes to challenge-salient elements.

Why I'm Not Really Into Blades in the Dark

One time I saw someone online claim that there really shouldn't be so many questions about how to run heists in RPGs. (Paraphrasing from memory) "They're just dungeons, aren't they? That's all dungeoncrawling is: a heist. An enclosed building with treasures inside and lots of security measures to work through. If you know how to run a dungeon, you know how to run a heist."

I disagree. I think that the techniques we've developed for dungeoncrawling are inadequate for simulating the fantasy of pulling off a caper. But I understand that it may not be obvious why.

Let's pick on random encounters. I like them. They're a good technique. I use them frequently and I advocate their strengths to rookie GMs all the time. For one thing, they certainly help you reduce prep! No need to plan out who the players run into or when or how strong they are or whatever. Just let the dice introduce the encounters at random and run with it.

But I think that they're bad for capers. Why? I'll give you three reasons.

1. Imagine that the players are planning their caper, and they're figuring out how to deal with all the guards on patrol throughout the bank. They have this idea: neutralize every guard. If there's a manageable number of them, like 5-7 maybe, then it should be feasible to just take them out one by one. Even better, maybe the players can coordinate taking out several guards simultaneously to reduce their risk even further. If they can just knock out every guard, then they're free to proceed without much issue.
   
   I think that's a great idea. It might be a bit difficult, but it's a sensible solution that's also perfectly in line with the conventions of the heist genre. I think we ought to validate it.
   
   Buuut if you're running the patrol guards as just random encounters generated by a die roll, then this plan is impossible. There's no "set number" of random encounters. They spawn into existence as the dice indicate.
   
   Okay okay but let's say you compromise a little. You say, "I can make that work with my random encounters. I'll just set a number, let's say 6, and only ever roll that number of encounters. After the 6th time I roll up a guard, that's it." But you haven't accounted for...
   
   
2. Imagine that the players are planning their caper, and they're figuring out how to deal with all the guards on patrol throughout the bank. They have this idea: research the security staff ahead of time. Running a background check on the people who'll be present during a heist is a classic caper move. Get your hands on some personnel files, ask around for rumors and dig up some dirt, maybe even make contact with them under false pretenses out at the bar or club. Soon, you can pick out a guard who's susceptible to bribes, or who you can blackmail, or who's easily distracted, or who might owe you a favor, etc.
   
   I think that's awesome. It's probably the most realistic strategy the players could employ, actually. Any security expert will tell you that the human element is almost always the biggest vulnerability in any operation. I think we ought to validate it.
   
   Buuut if you're running the patrol guards as just random encounters generated by a die roll, then this plan is impossible. Each guard can't be treated as a unique obstacle integrated into the wider campaign world if they don't exist as anything beyond a generic stat block.
   
   Okay okay but let's say you compromise a little. You say, "I can make this work with my random encounters. I'll just make a table of 6 unique guard NPCs, and I'll come up with a detail or two for each one. They're still encountered at random when the dice roll says so, but at least I can confirm that the players randomly rolled into Bob rather than Steve." But you still haven't accounted for...
   
   
3. Imagine that the players are planning their caper, and they're figuring out how to deal with all the guards on patrol throughout the bank. They have this idea: maneuver through the blind spots in their routine. Taking some time to scope out the joint, the players note the guards' standard operating procedure. Who keeps watch from which positions at which times? When do they change shifts? If they patrol across a wide area, what's their route? How long does it take them to complete a full cycle?
   
   I think that's groovy. It's probably the most basic strategy you see in stealth video games, right? By timing their moves right, the players can exploit gaps in the system and dart in between the spotlights. I think we ought to validate it.
   
   Buuut if you're running the patrol guards as just random encounters generated by a die roll, then this plan is impossible. By definition, the players can't rely on any given stretch of time being free of guards if the presence of guards follows no predictable pattern.

Random encounters are great for a lot of reasons, among which include how much you're able to cut down on prep. But this is an instance where I think the game benefits immensely from the GM taking the time to nail down some specifics in advance instead of just allowing a procedure to generate the challenge for them.

This is why Blades in the Dark doesn't satisfy what I want out of a heist game. Not that it uses random encounters, necessarily. But that it abstracts and removes the parts that would empower a player to think critically about the tangible scenario and apply real strategy to it. I'm interested in challenge-oriented games, which BitD doesn't really fit into. The flashback mechanic is a cute gimmick and all, but its basic function is to undermine the single most appealing part of the whole exercise for me. I've heard some people say that BitD lets you "skip the boring parts" which I cannot possibly fathom. Planning the caper is the game. Gathering information, assessing risks, and coming up with solutions to problems using the resources at your disposal is the entire point.

To me, achieving a great heist game is not done with a game design solution. It takes a level design solution. One with at least as much thought put into its creation as the amount of thought you want the players to put into its completion.

Living in a House of Illusion

It may be tempting to continue trying to weasel out of that prep. To find a way to still preserve the desirable properties of good prep in a game run entirely on procedures and improv-assisting tools. "If it's simply built-in on a structural level, the mechanics I design today can substitute for years of prep time I'd be doing throughout the rest of the campaign!" But I can keep poking more and more holes in this.

Let's examine the language rules in Lamentations of the Flame Princess. I've seen this suggested as a basic houserule in the blogosphere for ages. Instead of players picking out their PC's known languages at character creation, they instead just keep them undecided. When the moment first arises that you encounter another language, roll some dice based on your Intelligence. Run into some French bandits? Oh look, I happen to speak French! Wash up on the shore of Spain? Uh oh, turns out I don't speak Spanish.

Clever, right? Take a challenge-salient element (languages known) and move it from prep over to procedure. And I'll admit that this probably works fine for most groups. But it's not an equivalent substitute for just prepping the answer ahead of time. Why not?

Imagine if the player had chosen at the outset of the game to know French or Spanish. If that answer was there from the beginning, they might have made different decisions leading up to this point. If the player knows that their character can't speak Spanish ahead of time, then maybe they wouldn't have risked sailing in Spanish waters. If they already knew that they could speak German, maybe they'd have chosen to go in the opposite direction two or three sessions prior.

You might think you can get away with waiting to answer questions until they're "first asked," but you can't actually judge when a piece of information first becomes relevant. If the information was already in the players' hands before anyone thought to ask a question about it, they would act differently. I like having very concrete scenarios prepped because I can hand my players a ton of info at the very beginning, giving them lots of fuel to inform their ambitions from the get-go.

Here's another example. Let's example the famous hazard die. It seems like a really elegant way of incorporating that challenge-salient info into the procedure itself. "You wanna be able to track the location of monster? Well, one of the results on the random encounter die says "spoor." Boom, the procedure now generates the info that players need to in order to strategize. Right?

Again, I'll admit that's pretty clever. But I also insist that it's not an equivalent substitute. Yes, you're including the monster's tracks and scent and all that. But it's still also randomly distributed, not conforming to any kind of in-universe logic. The players can't make sense of it as naturally lining up with the environment in ways that reveal meaningful things about the monster's behavior. That only comes from you putting some time and thought into the monster's lifestyle and lair.

Facts and Logic

(hurr hurr Ben Shapiro joke)

"Dwiz, why do you care so much about logic? Loosen up, you square. Can't you just go with the flow, embrace the zaniness of D&D, and have fun?"

When we sit down together to play, I'm asking my players to believe in the imaginary world that I'm offering to them. "Even though you know this isn't real, act like it is." So yes, almost every player I've ever had at my table has automatically (and without much thought) simply expected a degree of logical cohesion in my games. Not because I specifically told them, "I put a lot of thought into making this scenario make sense! Everything is rationally explicable! I have mastered the art of outmaneuvering Cinema Sins!" Of course not. The reason they have that expectation is because that's what choosing to believe means. By asking them to participate in good faith, that implies that you're also asking them to expect logic, whether you realize it or not.

And a logical foundation to your strategy is part of that definition I gave for "true" agency. Not long ago on Discord I claimed that I don't like relying on the die of fate too often. Not that it's a bad tool, but that if you're using it a lot then maybe that's a bad sign. Let's look at an example of the perfect situation to apply the die of fate.

There's a crisis. Giant veggie monsters are attacking the suburbs. The players are chased by a big hungry pumpkin and go running into a neighbor's house. They try to gather supplies and hatch a plan. When they reach the kitchen, they wanna raid it for equipment. Are you ready to answer questions about it? Don't worry, I don't think you need to have prepped the answers. It's just a simple kitchen. Common sense will be sufficient here. 

"Can I find knives or a rolling pin or a big iron pan?" Yeah, of course. It's a kitchen.

"Can I find a grenade launcher?" No, of course not. It's a kitchen.

"Can I find chopsticks?" Hmmmmm... maybe? Some kitchens have them, some don't. That's certainly where you would find chopsticks, but that doesn't mean there's a greater than 50% chance that any given kitchen would have a set. I don't have any in my own kitchen! Sounds like it could genuinely go either way. 

So if it could go either way, do you say yes or no? If you say "yeah sure" then you're handing the party a free win for no real reason, yet if you say "nope, sorry" then you're stifling their progress for no real reason. By the time the question is asked, it's become impossible to decide an impartial answer.

This, readers at home, is exactly why we have the die of fate. Roll a d6, on a 4-6 the answer is yes. On a 1-3, the answer is no.

So far, all of that is fine by me... but only because I don't really estimate those chopsticks to be highly salient to the challenge. I could be proven wrong, and that would be really cool, but they're probably not going to save the day. Using the die of fate for situations like this is a good technique that I recommend frequently. But as I said earlier, it can go too far. Here's another situation that seems really similar, but where I fear the die of fate may be undermining the sporting integrity of the game.

Imagine instead that the players ran into the garage for equipment. Just like the kitchen, you can probably answer most questions. But some of them are trickier.

"Can I find a heavy-duty chainsaw? Like, for cutting tree limbs?" Hmmmmm... maybe? I'm pretty sure only a minority of suburbanites own chainsaws, but the chance that at least one person in this neighborhood owns one is pretty high. And a tool like that could make a pretty huge difference when used against big veggie monsters...

Outsourcing the answer to the die of fate would indeed be impartial. But impartiality is not the only relevant factor here! I know players. I know how they think. If they had the idea to look for a chainsaw somewhere in the neighborhood, I just know that they'd intuitively prioritize Randy the Handyman's house over Dennis the Conspiracy Theorist's house. They expect logic. It would feel pretty shitty if they were denied a great solution just because they falsely assumed that the imaginary world behaved like the real world in a believable way, when really it's actually built on randomness. Correspondingly, it would feel phony if they simply lucked into a chainsaw in the first house they ran into simply because the die favored them that time. An asset that good, that much of a game changer, probably shouldn't come down to mere luck, right?

If I was unprepared and I had no choice but to resolve that using random chance, it wouldn't ruin the game. It would be okay. But if there were a logical placement for where that chainsaw is found, foreshadowed earlier in the game by NPC dialogue or rumors or environmental clues, then that's way better.

You can Blorb, too!

Blorb is about selective prep. Think critically about your scenario, identify the elements most salient to challenge-based gameplay, focus on prepping just that stuff. This is a skill you have to learn and develop. But it is possible. For as long as people have been playing RPGs, there's been a persistent meme that players are so chaotic, totally unpredictable, no plans survive contact with the party, trying to anticipate the trajectory of the game is completely futile, etc. With all due respect, I think this attitude is annoying and unhelpful. It's massively overstated and often becomes a self-fulfilling prophecy.

I won't lie to you and say that you can become a perfect oracle of the game, either. You can improve at this skill, but it'll never be perfect. Maybe you come up with something that you merely thought of as surface-level flavor, but then a particularly clever player finds a way to exploit that detail as gameable information. What good is your blorb then?

My answer for you is to just do your best, anticipating the challenge-salient elements as well as you can, and then be pleasantly surprised at the rare occasion a player finds a way to make the color of the wallpaper or the location of spooky cobwebs actually matter. Just because you can be surprised doesn't mean that it's completely pointless to try using your foresight. Trust me, learning to identify the most challenge-salient ingredients in any given scenario and putting some effort into making them really solid and well-realized is probably the single best level design technique there is.

But if you're one of my quantum illusionist friends and you've grown used to doing almost no prep, yet you're also convinced by my argument in favor of blorbier results, then there's an answer: run more pre-made material.

Seriously, stop making all your own adventures, you nerd. There's a massive industry with thousands of new scenarios published every year, many of which are free, which you could be taking advantage of. Someone else has already done your prep for you! By spending even just a fraction of your campaign running published adventures, you get all the benefits of hours and hours of solid prepwork without having to spend that time or energy yourself. You'll need to develop an eye for a good adventure scenario versus a bad one, but once you get a hold of a solid module with several sessions' worth of content inside, you're all set. By our efforts combined, the greater canon of high-quality RPG scenarios out there provide enough blorby goodness to last forever.

-Dwiz