Originally, I wrote this long, rambling introduction about picking a role-playing system to run modern-era mystery games and about arguments with people about binary skill systems and why I personally prefer the freedom binary systems afford over things with narrative dice pools and hippie-dippie drama point bulls$&%. But I realized it was just a bunch of garbage meant to forestall arguments about which game systems were superior and justify all of the great advice I am about to selflessly bestow on all of you. I got to thinking: if the advice is truly great advice, won’t that be self-evident? Shouldn’t it stand on its own?
Well, it does. And I don’t need to justify myself to any of you. I’m f$&%ing awesome and I know it. I don’t have anything to prove. But you do deserve an explanation as to what this series of feature articles is going to be about. I have a feeling none of you will buy it if I just say “they are about awesome!” So, cue the rambling introduction. I hope its not as long.
I tend to focus on mysteries, investigations, and conspiracies in my games (interspersed with a kick-ass dungeon crawl now and again) which focus on the PCs using their skills and knowledge to overcome obstacles, gather information, and figure out what is really going on so they can fix it. In discussing those games on Twitter, a few folks have brought up some common issues they’ve run into and some great discussions got started.
When you start looking at mystery gaming, most of the issues (apart from the big one about how to structure a mystery story) are really about using the game’s skill system to its fullest potential. And the same techniques you use to run a great investigation apply broadly to just about any skill-based encounter or adventure in just about any RPG system. When I decided to write about the topic of skill-based gaming, I started by listing a few topics to touch on and the list just kept growing. You could write a book.
Well, I’m not going to write a book. But I’ve never been above milking a topic until there is nothing but chalky, white dust issuing from a shriveled… this metaphor turned disgusting. I’ve always been willing to exhaustively explore the full scope and scale of a topic, splitting infinitives with reckless abandon as I go.
So, what you are reading is the first feature article about Getting the Most Out of Your Skill System. And its a great starting point. And unlike some of my previous articles, its more broadly applicable. So, whatever your genre, whatever your game system, you should be able to use this advice. My own D&D/Pathfinder/d20 roots will be on display, of course, but I am using the same skills in my Hackmaster 5E game. It’s awesome! check out the free basic game.
Wow. Now that’s an introduction.
Rule #1: Players Can Only Declare Actions or Ask Questions
When the DM asks a player: “what do you do,” there are only two valid responses. And neither one involves the name of a skill.
First, the player can ask the DM a question about the world or the situation. “Do I know anything about the strange rune?” “Do I recognize the name ‘The Clan of the Pointed Stick?” “Do I see anything hiding on the ceiling?” Notice, none of these things require the player to mention skills. The DM can respond with an answer or ask for a specific roll. “Make an Arcane Lore check, but only if you’re trained.” “Yes. The strange old man in the mask mentioned it last week. It is apparently a clan of martial artists.” “Make an Observation roll with a -5 penalty because its dark.”
Second, the player can describe what action his PC is taking. And he should do so as if the D&D adventure were a book and his PC was a character. It doesn’t matter what skill or ability score the player thinks his PC should roll; what matters is what the PC is actually doing in the world and what the PC is hoping to accomplish. “I’ll give the door a solid, standing kick.” “I get a running start and jump over the chasm.” “I subtly offer the guard a bribe to let us pass.” The DM will ask for rolls as appropriate or determine the result some other way.
In the first situation, players often shoot themselves in the foot by trying to use specific skills in situations in which they are clueless. How does a player know if the Order of the Star is a matter of divine lore, arcane lore, local knowledge, or history if he doesn’t recognize the name. And yet, players often respond with “can I roll a Hisory check” based on the fact that it is their highest skill and they want to roll that one.
In the second one, players treat the game world like a point-and-click adventure game. Like there’s a button labeled Climb, one labeled Diplomacy, and one labeled Religious Knowledge. Again, this causes them to sometimes choose the wrong skill. But it also causes them to focus on pushing buttons instead of thinking about the living, breathing world. In the long run, this can prevent them from coming up with complex plans that combine several actions. Or considering any action that doesn’t easily or obviously fit into a single skill.
This rule needs to be enforced and reinforced constantly. I like to use shame and sarcasm:
DM: “… and the guard refuses you entry to the Citadel.”
Player: “Can I roll a Diplomacy check?”
DM: “Sure, knock yourself out.”
DM: “Wow, that’s a really good roll. Anyway, that was fun, but what do you want to do about the guard?”
Player: “I meant I wanted to roll that check at the guard.”
DM: “Well, he’s impressed by your roll too, but he didn’t bring is twenty-sided die. Besides, he’s on duty and can’t play dice games with you right now.”
Rule #2: Only Roll When There is Chance of Success, A Chance of Failure, and A Risk or Cost of Failure
DMs make their players roll too many f$&%ing dice. It’s fun to roll dice, sure. But only when it’s dramatically appropriate. Otherwise, it’s a waste of time and makes die rolling seem trivial, robbing the game of dramatic tension and frustrating the players. Every time a player describes an action, the DM has to decide whether a die roll is called for. And he should do so by asking these questions:
Can the action actually succeed? If the action is impossible, either because its just f$&%ing impossible or because the difficulty is so ridiculously high the player can’t succeed, don’t roll. Either tell the PC that it is impossible or narrate the failure. Done.
Can the action really, truly fail? This is actually trickier to figure out because a lot of actions seem like they can fail, but they really can’t. For example, if the PCs are ransacking a dungeon room, barring anything magically hidden or designed never to be found, they will turn up everything eventually. If they are researching information in the library, they’ll eventually turn it up.The trick is decide whether the PCs are constrained. Assuming a lock is within the PC’s skill level, they will eventually pick the lock and get it open. But if the room is filling with water or monsters are beating the snot out of the PCs, the question is not whether they succeed, but whether they succeed in five rounds. That is something they can fail at.
That is why it is not enough that success and failure are possible. We also have to ask whether failure carries a cost or penalty. In the case of the lock being picked while the room floods, the penalty for failure is death. The party is risking something if the action doesn’t succeed. Searching for a trap and trying to disarm the trap both have a risk: you might blunder into the trap and set it off before you find it or disarm it.
The assumption is that, lacking any constraints, the party will keep trying something over and over until they succeed. The DM should take that into account. When the party attempts an action, assume they mean to keep trying until it succeeds. If the party could freely do so, then it is not worth rolling. They succeed. And beware not to impose constraints that don’t really exist. “Because it will take an hour” is not a constraint. “Because it will take an hour and the place will explode in two hours” is a constraint.
It is also important to note that “missing out on something” is not the same as a risk or cost of failure. If the party is trying to pick a lock on a door that leads to massive treasure, there is nothing that keeps them from trying until they succeed. There is nothing that establishes a failure point. A risk or cost of failure is something that requires the party to decide whether it is worth continuing to try (time is running out to escape from the bomb) or else establishes a point of final failure (the bomb went off, you died). If there is nothing in the scene that would (a) cause the party to stop trying to succeed or (b) keep them from being able to try again and again, just give them the success and call it a day. The roll is a waste of time.
A Digression: Dump The Penalty for Failure Until You Fail Forever
Some games include the “penalty for failure until you fail forever” rule. This is most often associated with lock picking for some bizarre reason, but it crops up in other places too. Basically, it amounts to this: every time you fail at something, you suffer a penalty to retry it. After a certain number of failures, you have to conclude that the thing is beyond you and you can never succeed.
This is a stupid, arbitrary rule. Why does failing to pick a lock make it harder to pick? Trust me, it doesn’t. And if the difficulty of the check is low enough for the player to be able to roll it, it is not beyond the PC’s abilities. The rule only exists to prevent players from rolling over and over until they succeed. Well, so does my rule. And my rule does it much, much better.
So, dump the “penalty for failure until you fail forever” rule if your system has it. Just ignore it. And while it you’re at it, just ignore any of those codified rules about when you can and can’t retry an action in favor of a common sense approach.
Rule #3: One Roll is Usually Enough (Unless Something Changes)
Once you’ve decided that a die roll is actually called for because there is a chance to succeed, a chance to fail, and a cost or risk associated with failure, the next decision is whether to break the action down into one die roll or several. The answer is almost always that one roll is enough.
Rolling the same check over and over is boring. And, truthfully, the idea of “attempts” is silly. Its easy to say each blow against the door in an attempt to break it is an attempt, but how do you know when an attempt at lock picking ends and another begins?
Instead of focusing on individual attempts, focus on the situation. Specifically, when does the situation change. If the PCs are trying to pick the lock on the tower door to rescue the beautiful monster from the evil princess before she sacrifices it to her dark god, the PC is going to keep trying to work that lock until something changes. Like they hear the monster scream and gurgle and die. Why break it down into multiple rolls? What sense does that make?
Corollary to Rule #3: Reevaluate the Action for Rule #2 Before Every Roll
Unless it is dramatically appropriate (see Rule #3a below), one roll is enough unless something changes, right? Well, it is important to keep Rule #2 in mind before every roll. Before every attempt. That is, after a single attempt (whatever that means) fails, ask yourself whether the next attempt actually needs a die roll or not. Usually, it won’t.
Imagine the PCs are trying to break down a door. On the other side of the door is an ogre enjoying his Ogre Treats Cereal. If the PCs smash open the door on the first try, they will be surprised to see the ogre and the ogre will be surprised to see them. Neither side will be able to ambush the other. Initiative will be rolled as normal. However, if they fail to smash open the door, the ogre will realize someone is trying to get in and he’ll prepare an ambush.
So, something changes. Therefore, it is appropriate to call the first roll a single attempt that can succeed or fail by itself. And then the party can try again after that failure.
Imagine the PCs do fail. The ogre is alerted and gets ready to ambush the PCs. The PCs don’t know it. They’ll find out when they finally get through the door and walk into an ambush. Meanwhile, they decide to try the door again. Do you make them roll for the second attempt?
The answer is NO! Why? Because even though they might succeed or might fail, there is no longer any risk. They can just keep battering at the door until it breaks open. The ogre has already prepared his ambush. He’ll wait for the PCs. The second attempt is an eventually success: “you give the door a few more solid kicks. Eventually, it flies open with a heavy wham! And a javelin flies out of the door into your chest.” Done and done. All it took was one roll.
Now, let’s go back to that “Princess Sacrificing the Monster Scenario” for a moment. Because, I can already hear readers screaming at me that I’ve just made the entire outcome of the adventure hinge on a single lucky or unlucky die roll. Yes. Yes I did. The monster lives or dies based on how quickly the PCs can pick the lock. And while it is perfectly valid to boil it down to a single die roll, this is also a case where you could drag it out by using multiple die rolls. HA! You didn’t think I was going to say that, did you?
Rule #3a: Rule #3 Doesn’t Count if The PCs Can See The Ticking Clock
When there is a source of rising tension that the DM can easily communicate to the players so that the players are aware of the tension and can use it as a cue to change their minds, it is okay to break a complex action down into multiple “attempts” and require multiple die rolls. But each attempt needs to represent something. Each attempt might represent a minute of time passing. The party may or may not know when the ritual will be over (maybe knowledge of the ritual will help them figure it out), but the DM can describe what they hear through the door. The rising crescendo of the princess’ voice as she incants, the roar of the soul-vortex as it expands, and so on. The DM needs to be able to ratchet up the tension with every die roll and remind people things are getting worse to keep the PCs sweating bullets and maybe give them a chance to decide to change their approach (“Get out of my way, I’m busting this door down!”). In short, there needs to be a ticking clock and the party needs to be able to see it.
Of course, time is not the only possible resource that each attempt can waste (remember, if you are rolling, there needs to be a cost or consequence). If the party is gathering rumors in town, money can be the ticking clock. Crossing a desert? It’s hit points or fatigue. Building a thing? Materials can be broken, used up, or wasted. But remember, the party needs to see the resource dwindling or the consequences thereof.
Just remember that rolling the same die roll multiple times is boring by itself. And, eventually, no matter how much tension you inject into the scene, the players will eventually recognize they are just doing the same repetitive task over and over. So use this technique sparingly, keep it brief, and use it for the really big stuff. Otherwise, one roll will do it.
A Digression: Auto-Success and Metagaming
If you have been paying attention, you will notice that my rules are handing the PCs a lot of automatic successes. And if this is a problem for you, well, you should stop doing it. After all, it’s your fault. Getting the most out of your skill system means cutting out a lot of the crap. You are wasting everyone’s time requiring skill check for useless actions that have no downside. You’re just rolling dice for the sake of rolling dice. I totally understand that some doors in a dungeon will be locked or stuck. That makes things fun and flavorful. But it doesn’t make it worth wasting time on. If the party has a trained member who is skilled enough to possibly pick the lock, that is good enough. You don’t want them rolling forever on it while everyone at the table gets bogged down by a string of bad luck. And you don’t arbitrary limits like: ‘you only get three tries and then the lock apparently self-destructs.’
Put another way, when you designed the obstacle and didn’t make it possible to succeed, possible to fail, and give it a price or consequence, YOU decided it wasn’t worth wasting time on. You didn’t add in any risk. Any drama. You did that. Not me with my rules. You. You. You. My rules are just here so we can ignore your pathetic attempt at meaningless challenges.
You may also be sitting there worrying about the evil players who will catch on to my dastardly rules and realize that, whenever they have to make a die roll, it is something important and possibly dangerous. Of course, if they have special abilities and resources to spend to improve their checks, they will only ever use them on important things. They will always try to aid each other when its possible because its important to succeed. They will try their hardest at every challenge, those metagaming bastards.
Yes. They probably will. They will realize that die rolls only happen when its important and risky. And they won’t accidentally waste their best abilities on meaningless garbage. And that is what you are complaining about: you can’t trick them into wasting resources on stupid, usless crap and they won’t have those resources when something really big and important comes along. I don’t know about you, but I’d call that a pretty s$&%&# thing for a DM to want to do.
Besides, when you get down it, you are basically complaining that the heroes are adjusting their efforts based on the level of risk and the importance of task. Its like, when things are most important, they actually try harder. What crazy behavior! Yes, SOMETIMES, the PCs won’t have any in-world way of knowing this door is more important to break through than that door. Fair enough.
But… if you only waste die rolls (and therefore table time) on stuff that is important, you will waste less time on unimportant crap that can be handled with one quick remark, and fill your game with more meaningful challenges for the PCs to spend their resources on. They will still have to manage their resources. Its just, they will be managing them between important things.
Rule #4: Don’t Make the PCs Ask Questions
Remember when I said that PCs can only ask questions or declare actions. Ideally, they should never have to ask a question. A question is a speed bump to role-playing. And it means nothing is happening in the movie of your game. Let me explain.
The idea of role-playing is that the player is presented with a situation. The player projects himself into the mind of the PC and decides what the PC does in that situation. That’s role-playing. But when the player has to double check whether or not something is in the PC’s head to make the decision, they have an extra step between situation and decision. A speed bump.
Meanwhile, thinking about things and figuring things out are not actions. When someone is thinking about something, outwardly, it is kind of hard to tell what’s going on. They could be thinking, they could be daydreaming, they could be pooping. But the one thing they aren’t doing is moving around doing something interesting to watch.
There are a lot of DMs out there who will describe a circle of strange runes on the floor but wait for one of the players to do something that indicates they would like to make a knowledge check (like ‘examining the runes’). This is f$&%ing insane. When you see a sign written on the wall across the room in a language you understand, you don’t have to wander over and examine it (unless the text is too small). You just read it. Pretty much instantly. And if its a language you recognize but don’t comprehend, you recognize the language but you don’t know what it says. If you see a thing and you know what the thing is, the information pops into your head unbidden. That’s how brains work. And eyes. If they didn’t work that way, we’d spend all our time examining things and pondering things and squinting at signs from three feet away.
So, as soon as a PC is exposed to a thing they might recognize or know something about, they should recognize it or know it. Or at least make the die roll. Recognizing a monster and its particular traits and weaknesses should be part of the flavor text. Immediate flavor text. It should not require a PC to wander over and examine it to load the part of their brain that has the information in it. It should not require a player to ask!
DM: “On the floor of the room is a strange circle. Anyone who is trained in Arcane Mystical Knowledge, please roll a check.”
DM: “Arathicus and Bob recognize the circle as a summoning circle. A demon summoned into such a circle is bound, unable to leave it or return to its home plane unless the wizard lets it out.”
Of course, researching things is a different matter altogether. But then, the PC is DOING SOMETHING, aren’t they?
A Digression: When to Roll Knowledge Checks
Since I’ve brought up the subject of skills that essentially just determine whether a PC knows a given fact, you might ask whether or not they are bound by the same rules about when to roll and how often to roll. The short answer is YES!
Knowledge skills are a huge pain in the ass. On the face of it, it seems like they should always be rolled. A PC might know something, they might not, and, depending on the difficulty of the information, they could therefore succeed or fail. And not having the information usually makes things harder or impossible for the PCs. So, you should always roll one roll for each useful bit of knowledge. Right?
But, knowledge skills aren’t really skills at all. While you can make an argument that the PC is rolling to “remember” something, they aren’t that cut and dry. They are more of a random roll to see if some particular piece of information is in the PC’s head. Otherwise, there is no reason for the PC to have different knowledge skills. The skill would be “remembering” and the chance would be the same for every thing the PC was ever taught.
I will discuss knowledge checks more later. Because they can really, really f$&% up a game for no good reason. But ultimately, when it comes to knowledge checks, I tend to use passive scores. That is, if the difficulty of the knowledge in a given system is equal to the average roll for the PC (e.g.: 10 on a d20 plus the skill modifier in d20 games), the PC knows the thing. otherwise, not. So, both the training and the score are still useful, but I don’t have to waste time with die rolls. To facilitate this, I keep a list of knowledge skills by PC.
I go back and forth on the knowledge skill problem, but, for a time my players were trained to know that I would tell them anything they knew or recognized about the world the moment it became relevant and if I didn’t tell them, it meant their PC didn’t know it. We spent a lot less time with the players “examining things” and asking questions about this term or that monster.
But, look, if you like rolling for knowledge, it isn’t going to break your game (except in some ways we will get to in a later article). It is justified. So feel free.
Rule #5: Differentiate Approaches, Because Success Needs Consequences.
Now we come down to the most difficult rule to explain and the most difficult rule to follow. Sadly, it is also the most important. And it could be an entire article by itself. In fact, it probably will be.
Suppose the party is running away from something and they come to a locked door. They could double back and try to find a different route through the dungeon. They could bash down the door. Or they could pick the lock. Which they actually do is a fairly simple decision. Which skill is their highest: Dungeon Navigation, Door Bashing, or Lock Picking. A simple mathematical puzzle.
Now, suppose the party is evading something and they come to a locked door. They have the same basic set of options. But…
The thing is behind them but they don’t know where or how far. If they double back, they might wander right into it. Especially if they fiddle with the door for a little while and then decided to double back. And they might not be able to find another route to the surface before they get eaten. And…
The lock is a fairly simple lock. Its not hard to pick (easy check), but lock picking is complex and time consuming. So it will take three successful checks to pick the lock. Each die roll (success or fail) will use up a minute of time during which the thing in the dungeon is getting closer. Of course, the door can be relocked behind the party, slowing the thing down. And…
The door is stout and reinforced. It will be hard to kick open or bash down (hard check), but once it’s broken, it’s broken. The party only needs one success. And each failure might injure the basher. Of course, if they do bash the door open, the party won’t be able to lock it behind them and slow the thing down. And…
The lock is delicate. Once the party attempts to smash the door open, the lock will probably (definitely) be damaged and be impossible to pick.
Assuming you know all of this, which solution is best? Not such an easy thing to say, is it.
Most game systems treat all skills equally. Whatever skill you roll, the basic mechanics are the same. Even the games with fancy narrative dice and degrees of success and lucky or unlucky breaks. The dice are focused on the outcome of the action. That’s what they are there for. There is this belief that non-binary rules and degrees of success somehow change that. But they don’t.
What makes the game interesting, what makes the game a role-playing game, is the decisions the players make. What actions they take. But to focus on that, you have to treat different actions differently. And to do that, you have to make sure that the PCs actions have consequences, regardless of success or failure.
When writing a complex skill-based encounter, the first thing you do is come up with the problem or obstacle. Then, you need to come up with the different approaches (the different methods of solving the problem) the PCs might attempt. After you list the approaches, you need to answer two questions for every approach: (1) Why would the party take this approach over the others? (2) Why wouldn’t the party take this approach over the others? If you can come up with more than one for each, good for you. If you can’t come up with one pro and one con, you need to rework your situation.
For example: the guard won’t let the party pass. What are the likely approaches and the pros and cons?
- Plead/Persuade. Pro: No real risk. The guard just says no. Con: Not likely to work.
- Bribe. Pro: Likely to succeed. Con: Expensive.
- Intimidate/Threaten. Pro: Doesn’t cost anything. Con: If it fails, guard may call other guards to fight.
- Bluff/Deceive. Pro: Doesn’t cost anything (again). Con: If the guard figures it out, he won’t listen to the PCs again. Might report PCs.
- Sneak In. Pro: No interaction with guard needed. Con: If the PCs get caught inside, they will be in big trouble.
Notice, I can’t come up with a good pro for Bluff/Deceive other than “doesn’t cost anything” which means it is no better or worse than Intimidate/Threaten. So, when I write my adventure, I might put in some ‘fraudulent papers’ that the party can stumble over that give the bearer permission to pass the guard if they can convince the guard they are real. That gives them a reason to choose that option because, with the fraudulent papers, it becomes ‘likely to succeed.’ Also, that actually becomes a pretty favorable approach, but it only works if the party finds the fraudulent papers, so that’s okay.
Corollary to Rule #5: Approaches Are Actions, Not Skills
Notice also, I haven’t broken down this list by skills because SKILLS ARE JUST TOOLS TO RESOLVE ACTIONS. I’ve come up with approaches. Sure, some of them might mirror some skills. But if the PCs try to sneak in by hiding in the back of a merchant’s cart instead of rolling a Sneaky Skill check, I can go with it.
As a general rule, I always prefer to have to figure out which skill or ability check to roll to suit a particular action. I want to be able to wiggle on skills and abilities. If my list of guard approaches looked like this:
- Diplomacy Skill
- Stealth Skill
- Intimidate Skill
- Bluff Skill
When the PCs describe their action as “rolling in on the back of the merchant’s cart,” I’d have to try and cram that into one of those four skills to figure out how it should work and what the consequences are. On the previous list, its pretty obvious where it falls. It must be “Sneaking In” because that is the one that avoids all interaction with the guard and leaves the PCs wandering around inside with no authorization and the potential for big trouble. That works even if I decide that “riding in the back of the cart” doesn’t require a skill roll. It just works. And I might decide that because now they have to stay hidden once they are inside, so they will have plenty of dramatic rolls to make in a few minutes.
As a rule, I tend to set skills, abilities, and difficulties at the table except for a few vague notes (like: difficult check or not likely to work to remember me to set the difficulty high). That list of approaches for the guard at the gate is actually pretty close to the sum total of the notes I would bring to the table for that encounter.
Players can only ask questions or declare actions, and they should never have to refer to skill names to do it. Only ask for a roll if the PC can succeed, can fail, and there is a risk or cost for failure. One roll is sufficient, unless that roll changes the situation. Of course, you can roll multiple times if there is a ticking clock the party can see, but don’t overdo that. And when designing complex encounters, focus on approaches and make sure each approach has at least one reason to prefer it and one reason to avoid it.
With a few corollaries and a lot of explanation and rambling, those are my five rules for getting the most out of your skill system. With those to serve as a general framework, I’m going to explore a variety of topics related to skill-based adventures and encounters, mysteries, investigations, and interaction scenes. Hopefully, folks will find my advice useful, either to write engaging mystery adventures and campaigns or to improve the flow in the non-combat portions of the game or just to start fights with me in the comments section because I appear to have an upper extremity lodged into a lower orifice. After all, I did tell you it was perfectly okay to hinge the climax of your adventure on one lock-picking check.