Looking for the previous Early Access Feature? It’s here.

Looking for the Proofreadaloud? It’s here.

---

This is gonna get weird. I have no idea where it’s going. I’m also sorry for doing two bullshits in one month. This is what I’ve got.

I got on this kick a few months ago, trying to figure out why certain things were just so hard to do in tabletop roleplaying games like Dungeons & Dragons. And why they seemed to be getting harder with every new edition. Why are certain things so hard to build, run, and play? I even ranted a bit about some of them. At the top of my list were such things as traps, social interactions, wilderness travel and exploration, and town-based downtime, but the list has probably a dozen things at this point. Those four really bunch my panties though because they’re things that belong in fantasy adventure roleplaying games and so they should be really easy to do.

But they’re not.

This really bugs the crap out of me because, get right down to it, all I want is to run one great game. I want to run a perfect fantasy adventure campaign, and I know I’m not alone in that. I’ve got tens of thousands of readers, a thousand active followers, and hundreds of paying subscribers who all want the same thing. We all want to run one perfect campaign. Some of us don’t want to run fantasy adventure, but that ain’t my problem. Some of us are just wrong. Some of us like the wrong things. What can you do? You wrong people are still welcome here, of course, and I’m glad my work helps you.

So my dream is to run one perfect fantasy adventure campaign and to empower everyone else to do the same. There are two ways I can pull that off. First, I can give myself and everyone else the tools to run a perfect game in whatever system they’ve chosen, as long as everyone else accepts I mostly focus on the one system I’ve chosen and that they’ll have to adapt my work to their crappy, wrong choices and put up with me constantly digging at their wrong opinions. Second, I can build my own system that empowers me and everyone else to run perfect adventure campaigns as the primary design goal.

As for that second thing, I’m on it. It’s long-term though. The working title is Project Slapdash, and my most ardent supporters are already part of the process. We’re going to be doing some fun prototype testing before the summer’s done. As for the first thing, well, that’s what I’m doing, and I plan to keep doing it for a long while yet. Especially because so many of you have chosen to support me in that endeavor.

Thus, I’ve become obsessed with the question of why certain things are so hard to do in pretty much every system. Even the systems that claim to do them well. Especially, ironically, those systems. Why are some things so hard to build, run, and play?

It’s funny because roleplaying games are just action resolution systems at their core. Players do things, and the rules determine whether they work and, sometimes, how well they work. All based on a very simple premise that players choose actions and the character’s skills and abilities set the odds. It’s a simple, elegant, perfect way to play a game based on answering the question, “If you were your character in this situation, what would you do?”

The point is, anything you can break down into a string of actions or attempts should be easy to handle.

Take, for example, social interaction challenges. They shouldn’t be hard. I know people piss and moan about various mechanical problems with social interaction in various games. In D&D, for example, people complain that the dice system is too binary, there are too few mechanical social options, the player participation is inherently limited, player social skills aren’t accounted for, and Charisma exists, but none of those complaints actually lasts long in the face of arguing with me because they’re not real problems and the few that are real are also so insignificant they’re really easy to solve or ignore.

In the end, a social interaction is just a back-and-forth. Imagine you’re trying to convince a scared witness to give you information. You, the player, just say things, one after another, and dice get rolled for each until you either win or lose. Either the witness tells you the information, or else he stops listening, flees, kicks you out, calls the guards, or just has a breakdown.

And yet, Game Masters struggle with social interaction challenges. Most don’t know how to even build one beyond setting different DCs for different social checks. Most can’t run them at the table properly, but that’s okay, because most players don’t know how to play them. That’s why the standard social interaction challenge boils down to players saying the same things over and over and making the same die rolls until the Game Master finally arbitrarily decides enough successes have happened to call it a win. Most Game Masters don’t even let the players lose. They don’t know how losing even works, so they just let the players keep talking, and the players never stop trying to say the one right next thing that’ll win the encounter.

You and I and all of us Game Masters know there’s something deeply, fundamentally wrong there, but we don’t know what it is. We just know it doesn’t feel good. It doesn’t work. So we all keep looking for the broken mechanic. We all keep having dumbass arguments about Charisma as a stat and player skill vs. avatar power. If there’s something wrong, it’s got to be in the mechanics, right? Where else would it be?

But what I’ve noticed is that, for all the talk about broken social mechanics in the D&D space, very few Game Masters actually know the published mechanics. No one ever brings up the Attitude system from D&D 2014 or the more complex Attitude vs. Willingness system in D&D 2024. No one seems to have noticed there’s actually a formula for setting the DCs for social checks. Or, as D&D 2024 stupidly calls it, the Influence action.

Maybe that’s down to presentation, or maybe it’s down to how every Game Master assumes they already know the rules and so don’t bother actually reading them. I don’t know. I don’t care.

This same crap happens with wilderness travel. It’s not like D&D doesn’t provide rules for travel, pace, ambush, foraging, navigating, tracking, and getting lost. All the elements are there. Everything you need to determine how a trek across the wilderness goes, but Game Masters struggle terribly to design wilderness adventures, they struggle to run them at the table, and most players don’t know what to do beyond saying, “We go to the Lost Temple of Forohfor.” If the Game Master says, “Okay, how,” the players just stare the way a cow looks at an oncoming train.

Traps? “I search for traps.” “You rolled goodly. There is a trap.” “I Thievery the trap.” “The trap is defeated.” Hoody frigging hoo, what a fun time that was.

Don’t even get me started on the whole, “What the hell are we to do with town,” thing. That’s way harder to discuss in one paragraph because, even though every fantasy adventure Game Master worth his dice knows something’s wrong, no one even knows what town is in game design terms. Especially because I ain’t talking about urban adventures. I’m talking about the stuff that happens before, after, and between adventures that should be gameplay but somehow has become not gameplay and relegated to bookkeeping between sessions. Or just excised altogether.

I don’t want to get into all that crap today. Let’s just look at social interaction again.

A few years ago, I shared my basic, rudimentary way of handling social interaction challenges. I don’t remember if I ever made an official article out of it. Maybe. Probably several. Maybe it was in a live chat. Maybe it was in my book. It doesn’t matter. It was kind of half-assed. Just a simple set of extra mechanics to fill in whatever the hell was actually the gap in the system. Despite using it myself to this day and refining it multiple times, I’ve struggled to turn it into an actual system or a hack or a publishable pile of instructions. It’s really just a handful of stats and some basic guidelines. Nothing especially weird or complicated. I try to keep everything simple and approachable because I’m actually good at making game mechanics. But, for some reason, I can’t seem to put them in a form that works for everyone else consistently.

The same thing happened with Town Mode. Remember Town Mode? A few years ago, I tried to build a coherent system around turning the stuff that should be gameplay and happen in town but isn’t back into gameplay. But, after a promising start, I found myself flailing around and I couldn’t get the pieces to snap together. I blamed the systems I was working with for not having enough mechanical hooks and levers and then gave up hope, and then my life fell apart. But people are still asking me for Town Mode.

The issue wasn’t the hooks and levers. If there were missing mechanics, I could have built them. Something else was missing. Something big and weird.

Meanwhile, though, I mostly can get Town Mode to work at my table. Kind of. There are pacing problems, but I always struggle with my pacing. Especially online.

But back to the social interaction thing. What’s weird is some people raved about the half-assed, slapdash, confused crap I’d shat out. It was all they needed to get it working. Others are still asking me to explain how to run social interaction challenges to this day because they just can’t make what I wrote work.

Why are some things so hard? Why is it so hard to get certain mechanics right? Why is it so hard for Game Masters to use the toolsets in their systems to pull off certain things and have it feel right? It ain’t just D&D, by the way. I know that from my own experience and from talking to lots of others. Y’all think that just because I primarily choose D&D as the best option for the sort of real, honest-to-goodness quest-based fantasy action adventure I want to run, I don’t read, study, or play anything else, and I don’t talk to or listen to anyone else ever. If you paid actual attention, you’d notice a lot of the mechanical hacks and systems I’ve slapped together over the years are inspired by or even ripped off from older editions and entirely different systems. The original Tension Pool was a Frankensteining of old-school D&D’s turns-and-rounds timekeeping with the chaos pool from Margaret Weis Studio’s Marvel Heroic Roleplaying Game with a few other herbs, spices, and original ideas thrown in. My current Time Pool isn’t, however, stolen from Harper’s progress clocks. Neither is my Whatever Stat. You’d shit yourself if I told you where I stole parts of that idea from.

Story for another time. Moving on.

To run social interaction challenges, I give the NPC a few specific stats. First, there’s Disposition. It describes how receptive the NPC is to PC’s. How much does the NPC like, respect, or fear the characters. It sets the base DC for all interactions. Every NPC also has modifiers that describe how receptive they are to Persuasion, Deception, and Intimidation. Some NPCs won’t be bullied, but they’re gullible. Others aren’t empathetic, but they respect strength. Those modify the DCs for different social approaches. Every NPC also has a Patience score that determines how long they’ll tolerate the characters’ bullshit. Once they’re out of Patience, the NPC shuts down. That might mean turning hostile, running away, having an emotional breakdown, screaming for the cops, or whatever. Doesn’t matter. Every actual die roll the players make against the NPC chews up Patience. Every NPC also has at least one Objection. They might have several, and they might have some Motivations too, but they always have at least one Objection. It’s the reason why they won’t just do what the players want right off the bat. Motivations are reasons why they might help the characters. Every Objection and Motivation has a numerical score. Once an NPC’s Objections have been reduced to zero or once there exists no Objection higher than the highest Motivation, the NPC gives in.

It’s that simple. Mostly. There’s some extra elements and nuances and guidelines to the whole thing, but lots of Game Masters have been able to take that one paragraph explanation and run social interaction challenges successfully ever since. Others can’t. And I can’t seem to get a good, solid, complete write-up out of it. I don’t know what the disconnect is. Or rather, I didn’t. But spoiler alert: the reason is the punchline of this whole long-ass screed.

I do know one issue is that some Game Masters struggle to resolve social actions, but even that’s not hard to fix. Just because a character is talking doesn’t mean they’re taking an action. As a Game Master, you play out the conversation, in either the first- or third-person or both depending on the players’ comfort levels and your personal preferences, and listen carefully to everything the players say. You’re looking out for statements that could affect the NPC’s Disposition, Objections, or Motivations; reasonably inspire a Motivation; or reveal information about the NPC. Once a player says something that might pull one of those levers, you stop the action and ask for an Ability Check. Based on the result, you adjust the proper stat and then respond appropriately as the NPC. Of course, you always want to provide feedback and telegraph other levers.

It ain’t complicated; it works great. For some Game Masters, it’s been a literal game changer, but others just can’t seem to pull it off, and I couldn’t find a better way to present the system than just saying, “This is just how it works.”

The problem is Mental Models. Or rather, the problem is a lack of mental models. Actually, the problem is not even knowing mental models exist. The problem is that many of us are stuck in a procedural mindset.

A few weeks ago, I was doing one of my Mostly Monthly Live Chats for my generous supporter community. One attendee asked me how I’d build a good swarm encounter because swarm monsters suck as written. For a half-hour, I spitballed my way through the issue and ended up with a random pile of mechanical doodads and gewgaws. I did not build a coherent system by any stretch. But the supporter took what I did spit out to the table or to his own personal project or whatever and had great success building an actual encounter out of it. I, myself, was so pleased with how the whole discussion went, I threw it on the anemic YouTube channel I do nothing with except for half-assed, terrible video game streams no one wants to watch on nights when no one’s available to watch them.

I’m many things, but a marketing genius I am not.

Anyway, check out the swarm thing if you want to.

So, what’s the secret sauce? What keeps screwing me up? What keeps screwing everyone up? I’ll tell you what it is. I finally figured it out. It has a name. It’s mental models.

A mental model is a kind of simulated thing you keep in your head. It’s actually just you understanding how something works so that you know what it’ll do when you do things with it or at it or to it. My social interaction crap is not a procedure for resolving social interaction challenges. It’s not a set of rules. I just built a model of how a social NPC works and explained the relationships between the elements so that you, a Game Master, can figure out how a given social NPC responds to different actions.

To some of you, that probably seems like the same thing. It’s really hard to explain the difference here.

Maybe let’s think about hit points. Hit points don’t really involve a lot of step-by-step mechanics and procedures. They’re just a stat you track. They measure damage. Or capacity for damage. You know when damage happens, you have to reduce a creature’s hit points, but that’s not because it’s a rule or a procedure. Instead, you do it because that’s what hit points are. You also know that anything with half its hit points or less is Bloodied or Staggered or whatever, and when something loses its last hit point, it’s either dead or dying, depending on whether it has a name and plot protection or not. You also know that, in the fictional world, hit points measure how beat to crap something physically is because hit points are meat.

Fight me, bro.

Do you get what I’m saying? Please tell me you do. This is so hard to explain. Resolving an action is a procedure. You memorize the steps, and you follow them. Hit points are a concept. You don’t have to memorize all the things to do with them because you know what they are and what they represent.

Maybe compare how traps are presented in modern D&D against traps in the D&Dv3 days. At the turn of the millennium, which is now too many years ago, a poison dart trap was a mechanical trap comprising a location trigger element such as a tripwire or pressure plate, a device capable of making a ranged attack, like a crossbow or a spring-loaded device, a poisoned dart payload, and it had to be reloaded and reset manually. Someone had to take it apart, reset the mechanism, and put a new poisoned dart in the barrel. Each of those elements was explained individually elsewhere in the rules. Now, in 2026, a poisoned dart is a triggered action. When a creature moves onto the pressure plate, they must succeed on a DC 13 Dexterity saving throw or take 3 poison damage. With a successful Search action, the creature can see the holes in the wall containing the dart launchers and can then plug them with wax, cloth, or debris, or else they can detect the pressure plate and wedge it with an iron spike or similar object.

They’re basically the same, but one is presented as a set of steps and procedures while the other is a device built of components, each of which the Game Master understands conceptually. It’s a small difference in this case because a trap is a trap, though one takes up a lot more page space and you have to read through two paragraphs to find the bits you need at the table. Also, one entails a much more comprehensive procedure for building, pricing, and setting the difficulty for traps.

More insidiously, though, one of them pushes procedural thinking that can inadvertently limit open-endedness. There’s a lot of specificity in the sabotage methods in the 2024 rules I don’t think is doing anyone any favors. But I’ll leave that alone for now.

My social interaction thing also isn’t rules and procedures. It isn’t a bunch of “If the player does this, do that…” crap. Actually, I leave the procedures up to you, the Game Master, based on guidelines, judgment, and your understanding of the model.

“Here’s how NPCs think and react. Here’s how that’s represented mechanically. Adjust them as the players take actions.”

That’s kind of how you have to do social interaction, isn’t it? It’s too open-ended for anything else.

Some of you might remember my many rants against proposed social combat systems that boil interactions down to lists of broad actions the players can take and the results of each. Some of you accused me of basically building social combat with my interaction challenge system. Hopefully, now you get the difference. One is procedural. The other is based on giving the Game Master a mental model to evaluate, adjust, and manipulate.

But lots of you won’t see the difference, and that’s the problem. There’s a knack to using mental models instead of following procedures, and it’s mostly down to how you were trained to think about the game. You don’t even have to know mental models are a thing to use them.

Video game designers, especially the ones that make open-ended games, are really savvy about teaching players mental models. I could mention redstone in Minecraft, electricity in Breath of the Wild, or Portal, like, all of Portal, as examples of mental modeling done right, and you’d all nod along. Those ones are super obvious. Less obvious is the fact that most levers and buttons in video games that control doors and puzzles also make players build and experiment with mental models. They’re just simple models.

Really, anything that asks the player to experiment or find a strategic or creative solution necessarily involves some mental modeling somewhere, and games that lean heavily on such things rely on clever tutorialization. They don’t teach players to execute actions, but rather how the game’s systems work.

Which is where I done fucked up with my whole social interaction thing.

Most Game Masters did pick up what I was putting down. Even the ones over-trained to think procedurally and to execute instructions by modern roleplaying games with their legalistic writing and exceptions-based design. The problem was that I didn’t give the players a mental model for social interaction, and I didn’t tell Game Masters they’d have to do it either. That’s on me. In my defense, I thought I was writing rules. This is kind of a big epiphany.

Players need mental models, or else they can’t predict how their actions will affect anything in the world. I’m not saying players need a rundown of every social stat on the NPC statblock, but players do need to understand, conceptually, how social NPCs work. They need to know that the more NPCs like you or the better their mood, the more receptive they are to influence, but also that NPCs vary in their receptiveness to different approaches. They need to understand NPCs have internal objections that can be broken or overridden. They need to understand they can get clues about what’s in the NPC’s head through conversation, attentiveness, and active intuitive feeling-out.

With that mental model, players can say things like, “This dude’s got a bug up his ass. Let’s buy him a few drinks to butter him up before we start asking for things.” Or they can say, “Let’s find out what this dude’s into so we know what to offer him as payment.” Or they can say, “This dude is not buying our sob story; let’s see if we can cow him into submission.”

In other words, once the players know how NPCs actually work, conceptually, they can approach their interactions strategically. Isn’t that exactly what you want? Likewise, isn’t that what you want when the players are planning a wilderness trek, trying to bypass a trap, or figuring out how to spend their days in town before some incident gets all incisive and drags them off on another adventure?

In normal games, mental models are something between the game designer and the player. The game designer trains the players to understand them through instruction and play. Tabletop roleplaying games, though, are complicated. They’re created by system designers who pass them to scenario designers, who may or may not be Game Masters, who design and execute games for the players. Who builds the mental models? Who trains who? Who trains the players?

Ideally, the system designer should be doing the mental modeling. They’re the only one talking to everyone. They make player-facing rules, they provide Game Mastering procedures, and they provide scenario design instructions. They’re also the only guaranteed professional in the batch. See, players and Game Masters don’t need to understand what mental models are or how they work to use them. If they’re baked into the system, if the system is designed around them, everyone will just use them. But that means doing more than just writing rules and spells.

Unfortunately, tabletop roleplaying game system designers are very stuck in procedural thinking themselves, especially nowadays. Worse yet, tabletop roleplaying game design is kind of unsophisticated and haphazard, and that’s treated like a feature rather than a bug. Because tabletop roleplaying games are supposed to be free and open-ended and let everyone do anything they want. Meanwhile, they can lean on the Game Master to bring the reasonable, rational mental modeling.

But that’s just… wrong. Mental modeling is infinitely more important in open-ended games than in simple, procedural games, and someone needs to make sure the Game Master, scenario designer, and players are on the same page. At least for the big stuff. The stuff that comes up over and over in every game. Things like traps, social interaction, wilderness exploration, and town.

I don’t want to be too unkind, though. I respect many roleplaying game designers and, frankly, until I’ve actually published a system and proven my chops, I’m just an asshole blogger and occasional hacker talking a big game. But, again, that’s the longer-term plan. It’s in motion, but it won’t help you today.

What will help you today is, hopefully, my ability to now provide some better solutions to these ongoing problems. It’s time to re-explore the wilderness and town mode. I still want to fix traps. I want to revisit social interaction. Those will be nice projects for the back half of the year as I continue to rebuild and recover.

I’m just gonna need you to use some good judgment and accept you’re gonna have to teach your players a few things because the system designers sure didn’t.