If there’s one thing that I can proudly say I’ve done in my ~10 years of near constant DMing, it’s horribly fail in almost every conceivable way. I’ve run terrible, unsatisfying sessions. I’ve created utterly uninspired settings, devoid of any character or pull. I’ve written massive character backstories for NPCs that players talk to for three minutes, then forget immediately. I’ve written less than a sentence of notes for locations that then became frequent stomping grounds. I’ve written mysteries with massive holes that needed patching on-the-fly, monsters where I’ve missed key statistics, and homebrew magic so ridiculous that it wasn’t worth the bytes it was stored in. I like to think, however, that through this process of constant failure, I’ve slowly honed my DM skills to the point that I avoid a lot of the common pitfalls — instead, plummeting into new and hitherto unexplored ones. This post, I’m going to talk about what obscure problems I’ve run into, and how to avoid them.
Perfect Failure, Perfect Success
There is a very common meme in the RPG space, the meme of “wacky anecdotes from when people roll 20 or 1”. The story of the dwarf who rolls a natural 20 on a pit and convinces it to close up, that moment when your fighter rolls a natural one and buries his broadsword in an adjacent friendly wizard. These are funny anecdotes, but they’re also a terrible way to play a game. Unless the system you’re playing specifies extreme events from critical successes/failures, you probably shouldn’t add them. The reason for this is fairly simple — sure, it was highly entertaining when someone rolled a natural 20 on their acrobatics check, and gained the power of flight as a result, but what do you do when it happens again ten minutes later? The same again? A natural 20 occurs 5% of the time, which is pretty damned common in a dice-heavy game like D&D, Pathfinder, etc. If you’re having absolutely catastrophic things happen as a result of natural ones, then you need to be ready for this to occur several times a session.
Most systems dial back on the importance of rolling critical results. In contrast with common house rules, natural 20s on skill checks in D&D5e do not automatically succeed. If your bard is comically weak, then it doesn’t matter what they roll, they cannot lift an object that has a DC greater than 20 + their skill mod. In Pathfinder 2e, if you roll a natural 20 but your result is 10 below the DC, you’re still going to fail1. This might feel quite cruel, because rolling a maximum result on a dice is typically a momentous event, but it’s necessary to maintain the structure of the game. If a natural 20 guaranteed success on skill checks, then the dumbest moron in the land would have a 5% chance of being as knowledgeable than the smartest being on the planet. While this might be funny, and the vibe that you want for your game, you shouldn’t do it without heavy consideration. Games that have extensive house-ruled mechanics for criticals tend to degenerate into absolute nonsense — even when using the ‘official’ critical hit optional rules like the decks for Pathfinder 2e.
Even when it comes to regular failure, you should be careful to never have the outcome be immediately extreme (unless the system is balanced around this notion). If a failure on a Diplomacy roll means that the NPC will never speak to the player again, you’re going to be running out of NPCs by the end of the second session. I think the Climb rules in Pathfinder 2e illustrate this point very well.
Wait a second, where’s the line for when you fail? There isn’t one, because failure in this instance simply means that the action is wasted. The character attempts to move across the incline, loses their footing for a moment, and stops moving. Only when you critically fail, do you actually begin to fall, and even then, Pathfinder 2e has additional rules to alleviate that. This doesn’t mean there’s no consequence for failing — the character wastes an action doing nothing in a game where action economy is extremely important.
For me, I’d summarise this whole point as being “don’t exaggerate success and failure from the result of a single dice roll”. Sure, if a character rolls consecutive critical failures when they’re attempting to move across a sheer drop, you can describe them slowly and painfully plummeting to their death. That’s the mechanical punishment for being incredibly unlucky. Singular results, however, shouldn’t mean a great deal in the grand scheme of things. A failure to Persuade an NPC means a joke that doesn’t land, or an argument they don’t fully agree with. A critical failure means a joke that offends them, or a nonsensical anecdote causing discomfort and difficulty with the conversation moving forwards: it doesn’t mean they immediately draw a crossbow and attempt to murder you in the street. Similarly, a critical success doesn’t mean they immediately abandon their life, swear themselves to your cause, and ask for your hand in marriage. Pathfinder 2e has additional caveats in conversation with the Request action: “Some requests are unsavory or impossible, and even a helpful NPC would never agree to them.“
If you want a world where it’s possible to convince the queen to give the players all of her worldly possessions after one minute of conversation, that’s fine. With a campaign where absolute nonsense has a 5% chance of occurring every dice roll, just be ready for that game to have absolutely no structure after even a handful of sessions.
Old Man Henderson
In a lot of RPGs, when it comes to character creation there is an understood ‘tiering’ of skills and attributes. Usually, every system has some sort of ‘God Stat’ — a statistic that either demands having some points in for a functional character, or one that can have points piled into for good effect. When talking about the canonical fantasy attributes (STR, DEX, CON, INT, WIS, CHA), the go-to example tends to be dexterity/agility, which has enjoyed a comfortable spot in almost every fantasy RPG. With skills, there’s a bit more variance, even within the traditional dungeoneering RPGs. I’d categorise skills roughly into the following categories of ‘god stats’.
- Ability to perceive (perception, spot hidden, listen)
- Ability to sneak (stealth, disguise, hide)
- Ability to move objects and yourself (athletics, acrobatics, dodge)
- Ability to heal (medicine, first aid, surgery)
- Ability to persuade (diplomacy, persuade, charm)
- Ability to deceive (deception, lie, fast talk)
- Ability to discern (insight, psychology, sense motive)
- Rarely applies, see below.
Now, you might completely disagree with my selection here. “Where are survival skills!” I hear you cry. “What about library use!” Yes, yes. I’ve missed out quite a few, more system specific skills/attributes to allow for it to be a more general list, but bear with me here. I think I would quite comfortably say that you need to have a party with points in these things for almost any campaign. If you try to play a D&D5e campaign with a party lacking a good Perception score, you’re largely going to have a bad time. If you try to play a Pathfinder 2e campaign with a party lacking a good medicine score, you’re largely going to have a bad time2. If you try to play a Call of Cthulhu campaign with a party lacking a good Persuade/Charm/Fast Talk score, you’re largely going to have a bad time. Obviously, all of these things have conceivable exceptions, but I think these are reasonable assumptions in the main.
So the question is: why are these skills so valuable, or even required? My answer to this would be: because these skills are the ones that emerge ‘thoughtlessly’ in the course of play. If you have a world with societies and NPCs of note, then the character’s ability to interact with those NPCs is going to arise. If you have a world with dungeons (in the broadest sense of the word), then the character’s ability to scrutinise, navigate and manipulate that dungeon is going to arise. With ‘intellectual’ skills, RPGs tend to rely entirely on the player using their wits to solve puzzles or mysteries. Mostly the closest you’ll get to a ‘solve the thing’ skill is something like the investigation skill in D&D5e, or the Idea Roll in Call of Cthulhu 7e. As such, the ‘intellectual’ category lacks the god skills that the other categories have.
Now you’re wondering, “where’s the common DM mistake?”. The existence of ‘god-skills’ implies the existence of…normal skills. Skills that aren’t guaranteed to be useful for a campaign, things like Performance, Religion, or History. These are skills that are mostly taken for flavour purposes, rather than them serving a definite mechanical purpose. If you’re playing a cleric, you’re going to have a (relatively) high Religion skill. The common DM mistake is to think that it’s solely the player’s responsibility to introduce ways for these obscure skills to come up in play. It is not. It is the DM’s as well. The DM should be using the character sheets as inspiration for content — almost like a food menu at a restaurant. If a player has an extremely high Lore (Circus) score (yes, that is a suggested lore in PF2e), it’s utterly absurd to believe that the player can reasonably bring that about by themselves. Instead, as the DM, you should be looking for an opportunity to introduce that check. Maybe the party is assaulted by an aggressor, who moves in a way distinctive to a travelling show? With Lore (Alcohol), perhaps the party stumbles across a disguised noble, who’s drinking a beverage far too rich for their supposed station. If a player’s investigator has a hefty score in Accounting, put them in a situation where reading important financial documents gives them a minor clue for moving forwards.
This does not mean you need to have progression locked behind an iron gate marked “Knowledge of 15th Century Architecture”. It means that you should read your player’s character sheets, and identify opportunities to introduce content that you would not otherwise have thought of. Perception will come up on its own, social skills will come up on their own, but you need to work for the rest. I guarantee however, that when you introduce the obscure lore check that a player put points into, they will love you for it. It’ll be even funnier when they fail that check. In addition, it leads to a considerably richer experience when the players are doing something other than rolling the same five extremely frequent checks, and get to think about the (usually) more minor parts of their character.
Hootsby, the Level 17 Pigeon
Videogame RPGs like World of Warcraft or Divinity: Original Sin 2 have set a fairly ugly precedent when it comes to NPC/creature design in a levelled environment. I’ll give you a specific example of this problem: both games have settlements/civilisation that they expect the player to frequently visit for things like services, repairs, buying magic items, etc. However, in DOS2’s case, they don’t want you to be able to just murder the whole town and steal all the items, otherwise it’d lead to a horrible imbalance in favour of the players. In WoW’s case, they don’t want players from opposing factions to be able to easily rampage through the town, murdering all the NPCs and preventing other players from interacting with them. So there’s two immediate approaches here: you could flat-out prevent players from interacting with NPCs in that way, removing their agency and preventing them from imbalancing the game3. The second option, is that you make the town guard or merchant NPCs high enough level/sufficiently powerful that the players are in serious danger if they attempt to attack them. In DOS2’s case, this means that most of the merchants are weirdly capable in things like magic and swordsmanship, despite seemingly spending their lives yelling in a village square. In WoW’s case, this means that all of the town guard sit at the current max level of the players, and will usually cream the shit out of any individual starting trouble.
For a videogame, which lacks the moderating influence of a player – DM relationship, and with a computationally limited ability to adjust the world, these decisions make mechanical sense. The challenge of combat in DOS2 would be somewhat trivialised if you were able to gear yourself with the best available gear by easily slaughtering merchants at the very start4. The ‘story’ of WoW would be heavily disrupted if you could plunge a sword into the king’s head at any moment. The decisions, do not, however, make even a lick of narrative sense. In WoW’s case, a player character sat at the max level has overcome colossal challenges. They’ve fought and won against beasts from other dimensions, titans built as guardians of the world, outer gods who have come to consume the whole planet. Wielding armour and artifacts embued with holy power and arcane magics, they find themselves equally matched by a member of the town guard, wielding a mass-manufactured sword and uniform. It’s nonsense, clear baloney, but it’s a fact that the game needs you to accept (or at least, not think about) to not cause the wheels to fall off the rollercoaster.
In levelled, fantasy roleplaying games like D&D5e and Pathfinder 2e, levels do not just reflect a power change. Levels reflect a context change. There is a frequently quoted set of contexts with associated level bounds within these systems:
|1 – 5||Local problems, villages, towns.|
|6 – 10||National problems, cities, countries.|
|11 – 15||Planetary problems, continents, hemispheres|
|16 – 20||Planar problems, material plane, universes|
The idea is, as the players increase in level, the context of the campaign shifts to accomodate their new abilities. Importantly, this is nothing to do with mechanical balance. This is purely to do with the narrative implications. Shifting the context at level sixteen says to the players that the mundane world can no longer offer a problem they can’t easily overcome. Their accomplishments have grown them to the point that they need to engage with transcendetal beings, or other dimensions, to be given any sort of challenge. By simply adjusting the levels and stats of otherwise mundane NPCs, you rob your players the opportunity to look back and see how far they’ve come. If the difference between a level 1 bandit, and a level 17 bandit in your world purely amounts to greater damage numbers and stats on a page, then something has gone wrong and an opportunity has been lost.
To codify this into a mistake then, the problem is merely upscaling the challenges that were faced in previous levels, instead of introducing new challenges that better fit the power the players now wield. This is something I have stumbled across in some of the PF2e prewritten content. I won’t have any spoilers, but the players go to a city that has a level 17 city guard in garrison. As far as I can see, they’re not blessed by the gods, they’re not forged out of clay and given sentience through magic — they’re just normal human humanoids.
Now admittedly, the city isn’t an entirely mundane one, but my point still stands that there hasn’t been an established narrative reason for the guard to be walking gods. If you need your players to visit a city where you don’t want them to immediately tear up the place with their immense power, you probably want to set it somewhere that isn’t extremely mundane. Put the city in a pocket dimension, with divine otherworldly guards — not just ‘Todd the Guardsman’ who, for some reason, is capable of cleaving a hill giant in half. Ask yourself, if your town guard are so powerful, why aren’t they the ones solving the problems that the adventurers are? There could be a myriad of reasons why this is the case, political, religious, not just a power disparity. There just has to be a reason, otherwise you end up with WoW-world.
1 Your roll fails to hit the DC, which means it’s a failure. It’s 10 below the DC, which means it becomes a critical failure . The natural 20 only adds one level of success, which returns the result to being merely a failure, but a failure nonetheless.
2 I’m increasingly convinced that the whole system is balanced on someone having the ability to effectively Treat Wounds.
3 The Elder Scrolls games are a mix of these approaches. Sure, you can murder a shopkeeper in Skyrim, but they won’t have the purchasable catalogue of their store on the body — making the murder relatively pointless. With some NPCs, you’re prevented from killing them at all, simply rendering them unconscious despite the fact you’re burying a battleaxe into their skull.
4 Worth mentioning that DOS2 does indeed give you the ability to outright butcher NPCs. If you go back to a town having gotten several levels over it, you’re able to absolutely blow it to smithereens. Though at that point, the gear they drop is likely to be useless. I wanted to have DOS2 as an example of a non-MMO RPG, but in a lot of respects, it bucks the normal trend here.