That is, the game doesn’t require NPCs as you may have seen them in 5e adventures, streamed games, narrative-based or storygames, etc. An adventure can have no monsters, no people to meet, and still be very much a GRAVEROBBERS adventure. And when they are there, the conventional wisdom of making your NPCs seem complex, motivated or “real” doesn’t necessarily hold.
Characters with depth can be a part of your game - my players enjoy interacting with black market dealers or co-conspirators during a Prep phase, so I’ll at the very least give them names or quirks - but from my perspective as a Judge even the most fleshed out character in a game run using GRAVEROBBERS is mainly there as a function of game design. The game’s primary antagonist, diegetically a system of oppression enforced by the House, is non-diegetically expressed through an automated game mechanic. Traders and conspirators, as mentioned, are for getting items, resources and info in players’ hands, and these Prep interactions are also driven by the underlying card system. These mechanics function with or without the window dressing of verisimilitude.
When characters do exist in adventures, it’s for gameable reasons before dramatic ones. The line can be fine, and it’s kind of a chicken-egg thing if you’re a good enough designer/writer, but I think the distinction is important.
This all sounds very “game-y”, and it is. That’s by design. You can flesh out NPCs with backstories and motivations, and your game will likely be better for it, but the text and games you run using it should function just fine without them. Learn that part of the game first, put it first in your design, and I reckon GRAVEROBBERS will work better for you.
(Then, if you want, you add in three dimensional NPCs with complex motivations and elaborate histories, and really knock their socks off.)
Preamble over, let’s look at the most common kind of NPC-type “character” your players will be encountering during your games, the common or garden guard, through a lens not of running characters as most of us have come to understand the term, but as a function of game design. I’ll detail three things that guards are in GRAVEROBBERS, but first let’s touch on how and why they’re not - again, merely not necessarily - characters.
First up, detail. Your guards don’t need to have names. They don’t need backstories or motivations, and you don’t need to flesh them out or think of them as human. This kinda works diegetically if you look at it right, as they’ve given up their humanity by buying into a system of oppression, wearing the uniform of the oppressor and enacting the House’s orders. As far as any deeper fictional reasoning for this lack of detail, GRAVEROBBERS only offers what it functionally needs - this won’t satisfy all players, or all Judges, but you know your table better than me so I’ll leave that with you.
And, y’know, if you want to get into morally grey, both-sides storytelling and try making your players find out that guard had a family or liked to bake bread or some other #notallcops bullshit, play a different game. The rules as written require a faceless, oppressive, antagonistic force. Moral ambiguity is incompatible with the core gameplay loop. ACAB.
So then what is a guard in your game if not a person?
Well, like a wall or a key or a door behind a bookcase, the guard in GRAVEROBBERS is an element of your level design.
We can’t go that way because there’s a wall; we can’t get through because the door is locked; we can’t pass because the guards will see us. At its base level, a guard is something you put into your dungeon to stop the players moving forward in a particular way, at least temporarily.
And crucially, a guard is not a passive obstacle but a hazard - this is a stealth game, and with the failure state being spelled out in the text as “being caught or hurt”, a guard represents ample opportunity for both. Because of how the ruleset works this likely isn’t something you’ll need to heavily enforce as the Luck mechanic will represent the danger for you, but don’t forget standard play. Consider where guards take prisoners in your level, whether they attack on site, etc.
Telegraph the presence of guards. The Prep phase is ideal for this - players could case the joint, get inside info, find schedules or scope out a guard’s beat. During the Job, don’t spring guards on your players either. As the opposite of the failstate “caught is “not caught”, ie as-yet-undetected, they should be aware of the obstacle before it’s aware of them. Stealth is a given.
Of course, as with most obstacles in a world that runs on fictional positioning or what GRAVEROBBERS calls standard play, a guard as an obstacle is not impassable - which leads to the main thrust, challenge.
Once an obstacle is identified, it can be overcome. Part of the joy of ttrpgs is in the freedom for players to set goals themselves and then devise and execute their own, invented plans in order to overcome obstacles. (It’s most of the joy in an adventure game. This is where the meat is.)
Be aware when designing your dungeon that a guard - unlike, say, a solid tunnel wall - is an obstacle that begs to be overcome. Assume your guards will be dealt with. Set up opportunities to do so. It’s fun.
As far as overcoming that challenge... during play, the options presented by the players’ Odds might be the most obvious solutions, and don’t dismiss the impact of a successful Violence roll, but standard play is where the medium shines. So accept your players’ plans! Talk them through with them as a Judge, not helping per se, but encouraging creative ideas and identifying dead ends. Be generous, especially when the players can back up their schemes with concrete fictional detail - their items, clothing, or truths they’ve uncovered about your world.
(This is one area in which a guard being “fleshed out” can be useful in a sense, but remember the goal is not to create complex human beings but provide gameable detail. An allergy or a fear of rats is immediately useful, adventure game-style info, but don’t get bogged down in specifics. Setting information in GRAVEROBBERS exists to be put to use in gameplay first; worldbuilding is not a goal but flows naturally thereafter.)
If you’re unsure about a plan, talk it out - rolls are not catch-all oracles but pivotal moments of risk in this system, so don’t fall back on them too readily to divine outcomes, especially if the actions being undertaken by the player characters don’t fall under any of the Odds. The gaps in those four categories are there for a reason - let ideas from outside that box play out.
And remember, the game is not linear, and a challenge overcome is not the end of the story. A guard you sneak past is still on duty, a guard you seduce is still your enemy - a guard you kill is a corpse that may need hiding.
Your players can, and will, put your guards to use for their own ends.
Guards have info, and info is a weapon. Guards are info - their existence and behaviours tell you something about the world. They might not be “people” but they’re a design element that can perform all the useful, gameable functions that people can.
And if the guard is dead or tricked or trapped or tied up, etc etc, you have not only cleared the obstacle but opened new paths through the fiction. They can be stolen from too. Either way, you can have their stuff.
This is crucial to consider during dungeon design. Any weapon you give a guard is a weapon that may end up in a player’s hand. Don’t forget this, but even more importantly don’t ignore its potential - offering interactable detail and open ended tools to your players within the sandbox is the best and easiest way to facilitate play. A sword is fine and good, but I tend to default to a nice pike or crossbow. Give the guards torches if there are dark areas or flammable things, metal items if there’s electricity about. Uniforms are fun (also, put spares lying around, or offer them in Prep).
Don’t shy away from creative uses of this stuff from your players. They’re going to think up things you didn’t, this is a feature not a bug, it’s the point and the fun of GMing - your share of the discovery that adventure games offer (while the players’ is in seeing how their creativity expands and changes your world and their understanding of it).
Putting a guard into your level design, like any good gameable element, is going to explode with possibilities during play. Acknowledge this during design, lean into it, and encourage it during play. By looking at guards as functioning pieces of the game first, and characters or people second, you can facilitate the kind of creative problem-solving gameplay that makes OSR-genre adventure games like GRAVEROBBERS shine.
(And don’t forget, if they do kill a guard, Heretics can see ghosts!)