# RPG A DAY 2025: Day 15 - Deceive tags: #thoughts #thoughts/RPGaDay/2025 #game/apocalypse-world #game/rpg/dungeons-and-dragons ![[RPG a Day 2025 (illo).png]] Deception is a tricky subject in RPGs, though not when you're talking about player characters deceiving NPCs. That's pat. Everyone accepts that is not only possible but desirable. There are vast cages around how that is done. How player agency plays into that. How long it'll last and how much it'll do. That is the player acting on the game world through the proxy of his character, and for most RPGs, that is literally what they do. No, it's only tricky when you're talking about deception of the player characters. It's mainly murky because there has been such an effort in mainstream RPG design to mute any kind of meta-thinking about the character experience. That is, traditional RPGs are very focused on trying to unify player and character knowledge. They denounce any separation of those things as "destroying immersion." Any action based on that knowledge is downright anathema. This is why deceiving the player character is such a delicate operation. To mechanically resolve the fact that the player is being acted upon, there needs to be a mechanical engagement. But the fact that there is an engagement can give away the fact that the character is being manipulated in a way that lets the player know something that the character does not. Usually, this is done via hidden knowledge. That is, the GM rolls the die behind the screen. The player knows that the mechanics have been engaged, but they don't know how or why. This is still information leakage across that anathema barrier, but it's considered reasonably acceptable, or at least unavoidable. One of the side effects of this is when the character is deceived by an NPC. Players often feel it's because the GM has lied to them, which is, under the strictest sense of the term, absolutely true. The GM has to create a deceptive atmosphere in order to deceive the player because player and character knowledge are assumed to be desirably unified. This is fine if your GM position is inherently oppositional and not simply acting as a referee. Games which tout themselves as highly simulationist often stumble on the fact that what they are *"simulating"* is ultimately the desire of the GM to steer the game in a certain way because the NPCs are directly under the GM's control—what they want and how they go about achieving it are not mechanically driven. I don't actually believe this is resolvable from a strictly simulationist point of view without player versus player engagement. There must be something to referee for GM neutrality as merely the agent of simulation to be believable. Someone must willfully engage in committing the deception in order for the simulation to simulate the impingement of the deception on the character. (Yes, I know this is extremely abstract and philosophical, and I promise you it doesn't get better for the rest of this article.) What then to do about the question of deceptive agency? After all, the GM holds all the cards in a very literal sense. Their will drives the world. They are the prime mover without question. Ways to handle this within the context of the Simulationist game have included things like random oracles, which provide a reasonable facsimile of external motivation. Consider the reaction table in **[[Dungeons and Dragons|D&D]]**. > Someone with much more motivation than myself went through and made a [distillation and comparison of the reaction table throughout the history of D&D across multiple editions](https://osrsimulacrum.blogspot.com/2020/09/across-editions-reaction-table.html). This is brilliant and insightful, and I encourage you, if you have any interest in this subject at all, to go and check out the article. > > Amusingly, [they weren't the only ones delving in these things from that perspective](https://thealexandrian.net/wordpress/50021/roleplaying-games/random-dd-tip-relationships-reactions). > > [Or even the second](https://yumdm.com/reaction-rolls/). Let's take the one from **D&D 2nd Edition** in part because it's got enough complexity to be interesting and not so complicated that it requires multiple recursive rolls. ![[DnD 2e - Encounter Reaction Chart (illo).webp]] Now the first thing you'll notice is that this is for monster reactions and not for general NPC reactions, but there really isn't a table for general NPC reactions because that's one of those filthy roleplayer things. However, it has been used over the years as a general NPC reaction table. Why? Because it displaces the responsibility for making that decision from the GM onto the system, which is theoretically a far more neutral arbiter than even the referee of the game. Having displaced the responsibility, if the player becomes unruly as a result of a character deceiving them, the GM can simply point at the rules and say, *"The devil made me do it."* There is some measure of value to be gained by being able to put the responsibility on an agent outside of the GM himself. For those who are rule-bound, being able to say that the rules are responsible for the setback is part and parcel with the rest of the experience. That's not necessarily a criticism. It's simply an observation of the way that the rules are architected and the game is played. In the late 80s and early 90s, we saw a move in the RPG hobby toward greater Narrativism as a design principle.[^1] One of the key changes between more conservative, classical manifestations of RPG play and the new Narrativist games was the greater acceptance of a bifurcation in the understanding of the world and character experience between the player and the character themselves. The player was allowed to know things that the character did not know. Specifically, the player was allowed more information about the world and the state of affairs than the character, and sometimes that difference was the whole point of the game. One of the frontrunners in the new fiction-forward, narrativist game design space was **[[Powered by the Apocalypse|Apocalypse World]]**, though it certainly wasn't the first. What **AW** did was reify a lot of the ideas that were being kicked around in the story game side of the hobby, put them down, and sold them in a product. For a lot of people, this was their first introduction to thinking about characters in a way which was explicitly disjointed from the player. While in classic **D&D**-derived play, there is an understanding that the character is a pawn within the context of the game, being moved around by the player, their level of knowledge about the world should be equalized. In **AW** and other early Narrativist games, the character is recognized as a pawn, but more so as a character being written/acted by the player who is directing them into a series of experiences for which the audience—that is, the other people at the table and themselves—will find interesting and entertaining. For some people, this was a transition they were never going to accept, nor were they comfortable with other people playing with it. For others, it was a bright ray of new sunshine, which allowed them to find new ways to explore storytelling with their friends. It is an essential and extreme change in stance within the context of the game space. The player changes from being a player of a game, imagining themselves to be a character, to a more distant authorial stance, treating the character as an element to be manipulated and the story itself, in part sharing primacy. As discussed at the time, it was specifically not simulationist, because the idea of the mechanics was not to model some underlying physical/social/psychological processes, but instead to reflect the story first and allow the players to interact at that level.[^2] What does this mean for deception within the context of play? Well, one of the ideas that **AW** introduced was that the GM should never roll the dice. Instead, the GM introduces fictional elements in a story-first manner, which then channel the players into reacting within the fiction. Those actions trigger mechanical "moves," which are the traditional elements of mechanical gameplay and where randomness may be injected into outcomes. In practice, this means that the deception of the players and the deception of the characters no longer have to be funneled through the same channel. Negative results to the character can be immediately communicated to the player because they are the result of that player's actions or choices. Moreover, the player can be told that the character is being deceived. There is no expectation that the player can't know that, and almost every expectation that the player must know that, so that their play can continue with the fiction taking primacy. This is a massive change to the underlying philosophy and has repercussions well beyond the question of whether the character is fooled by someone or not. That's a story for another time. Compare this to the way that the **D&D** reaction rule plays out. It may and probably will be made in secret. The GM then plays out the reaction that the NPC has to the character with the knowledge of what their inclination is. Then, in order to deceive the character, the GM must deliver false information to the player and determine whether or not the character believes it in secret. This is a lot of hidden knowledge, and it is very difficult to juggle along with everything else, to the point where many GMs just avoid situations in which there is active deception beyond momentary lies, though some of the best and most discussed character experiences in traditionally architected games revolve around a well-pulled-off deception by a GM of the players themselves. The focus, however, is specifically on how the GM deceived the _players_, not the characters. This is the continuation of the idea that character and player understanding of the world should be unified and that deception thus must be played out on the players themselves. When it works, it works really well, but when it doesn't, it often leads to resentment, frustration, and irritation, that the GM has deliberately provided false information that the GM is not acting as a fair arbiter of the experience but instead deliberately providing false information. In a strange way, this is most heavily contrasted not with the story game with the traditional GM/player dichotomy, but instead with what became the GM-less game, still played with multiple people at the table. In a GM-less game, no one is responsible for being the sole arbiter of truth. Architecturally, it is literally impossible. Thus, when a character is deceived, it is as a direct result of another player's intervention, mechanically mediated by the extant mechanisms for distributing power throughout the group. That power is expected to be distributed relatively equally, blunting the emotional response of betrayal. This gets a bit esoteric by necessity. Consider **[[Microscope]]**, which is a role-playing game by abstraction, in a sense. The players adopt characters as a temporary measure in order to participate in scenes. Every scene has a driving question which will be answered, and when that question is answered, the scene is over. It may be returned to at any point if someone else wishes to pick it up, but the same players may not be playing the same characters. How, then, in this very abstracted context, can intercharacter deception actually occur? Often, it is literally driven by the opening question. *"Why did Captain Hunter believe Commander Hosea when he lied about the readiness of the troops for the third battle?"* Everyone in the scene, and beyond that, everyone at the table, knows that Commander Hosea is about to lie to Captain Hunter. There is no doubt. The players know what will happen within that scene. It is inevitable and, in fact, must happen. The scene continues until the question is answered. The deception is known up front. The question is literally to answer why. Just as equally, the question could be, depending on what initiates the scene, how the lie was told, and not whether it was believed or not. The question could be: what happened after the lie was told? In every case, the players know up front that there is a lie and that it will be told here, and what will happen in the scene by being revealed. This depends entirely on player and character knowledge being very far from unity, and yet it still works as a satisfying game in a GM-less architecture. The power is distributed around all participants at the table, and because it is, they all must have relatively equal knowledge of the state of the world at any given time. The deception happens to the characters explicitly and not to the players. That's not to say you can't have deception between players in a GM-less game, but it comes via the mechanism of surprise. The deception must be about surprising the player with something they hadn't considered before it's revealed in the process of play. That is still very possible in a GM-less architecture and is often even desirable. This brings us around to perhaps the strangest case within the design space. What do you do with deception in the context of a solo game? In a solo game, the one player is also the creator of any complication. All the opposition comes from the imagination of the one player. How is that deception perpetrated? In many, if not most cases, the answer takes us back to that original **D&D** example that we talked about: the random generation chart. Not only that, it's true for a similar reason: to displace the responsibility for making the decision about the initiation of the deception from, in this case, the player to the machine of the mechanics. Once that responsibility is displaced, then it can be brought in as a surprise from that external trigger. Since solo games absolutely depend on the fact that character and player knowledge are very much entirely separate, this doesn't introduce as much philosophical friction within play as it does when the assumption is that the character and the player minds are close to being one. When the expectation is that the player and the character know different things, then the player can introduce deception to the character while knowing there is deception going on. This can happen as a result of an external trigger, such as an oracle, or it can happen simply because the player believes that the fiction should involve an NPC deceiving the character at this time because of other elements of the fiction itself. That is, if a character is introduced who is the character's angry ex-girlfriend, and an element of the established fiction is that they broke up because she cheated on him. It is fictively compelling for, in a future confrontation, her to lie to him. While it's not fictionally inevitable, it is coherent and consistent. The player can freely introduce this idea, along with interaction with that NPC, as logical freighting. At that point, the mechanisms of play kick in, and what remains to be resolved is whether or not the character believes those fictively introduced deceptions. As **AW** itself introduced as one of the axioms of its philosophy: *play to find out*. Deception within the contextual framework of a role-playing game is a complex and ultimately difficult, perhaps even thorny, philosophical pit. Deception requires a responsible party to initiate. Someone must be being deceived, and someone must be doing the deceiving. It's not something that can be a passive mechanism. How that responsibility is organized within the context of play is something that has been addressed many ways, and it's fascinating to contemplate the points of convergence and the points of divergence. Hopefully you haven't felt deceived in the course of this article. Hopefully it brought you to a wider understanding of the possibilities of play and the diversity of mechanisms of play that's out there. Now, I encourage you: go forth and deceive. [^1]: No, this isn't the post where I go on at great length about a deconstruction of the [GNS Theory](https://en.wikipedia.org/wiki/GNS_theory) and why it's terrible, but still a useful tool. I'm holding that one in reserve for another day. You're welcome. [^2]: Okay, maybe I lied a little bit when I said I wasn't going to get into GNS Theory. This is one of the places that I think that the overall theory breaks down significantly, especially in light of more modern game design developments. This is *not* non-Simulationist play. It simply reproduces a simulation of *storytelling*. The process of modeling how story elements interact with each other in an abstract way rather than trying to model physical, social, and psychological processes. Again, an article for another day and probably good for a PhD thesis at some point.