We’ve reached the end of the road: The dice have been rolled. The mechanics have determined success or failure.
Now the GM needs to describe that outcome. (They need to complete the fiction-mechanics cycle by bringing the result back into the fiction.)
If the result is a success, this usually means answering two key questions:
- How does the intention succeed?
- Are there any complications (i.e., unintended side effects)?
If the result is a failure, the questions are:
- How does the action fail?
- What are the consequences of failure?
The process we talked about in Fictional Cleromancy sort of naturally elides into this: As you’re thinking about graduated results, you’re thinking about what the potential outcomes of the action can be. You can usually just carry these thoughts forward through the mechanical resolution.
The actual narration of what’s happening in the game world is, of course, more art than science. But when it comes to describing outcome, there are a few general principles that you can keep in mind.
INTERNAL vs. EXTERNAL FACTORS
First, consider the question of why the outcome happened. What were the determining factors?
Bear in mind that both internal and external factors can influence the outcome of a skill check. (The distinction here is between failing to crack the safe because you’re simply not skilled enough and failing to crack the safe because your lockpick was defective and snapped off.) A lot of GMs default exclusively to the former (the character made a skill check; the check was a failure; therefore it was the character’s fault), but it’s arguably more effective to remember that the randomness of the dice models the entire situation, not just variance in the character’s ability: Sometimes you fail a Steath check because a guard comes around the corner at exactly the wrong time. You fail a Jump check because the ground is unexpectedly slippery. And so forth.
Another way of thinking about this is that, in any given skill check, there are myriad factors that determine its ultimate success and failure. Some of these factors – generally the ones we care about the most – are known. (For example, in D&D we’re always interested in whether a character’s armor will protect them from an attack, so their AC is always factored into the attack roll.) A lot of factors, however, aren’t important enough or consistent enough for us to want to specifically track them, so we use a random number generator to account for all the different factors that could impact the success or failure of any given action (and then trust to the GM to adjudicate the result accordingly).
For example, let’s say that the PC goes to a library and makes a Research test in order to find a particular piece of information. The test fails. The GM decides that it’s because the library doesn’t own a copy of the book that would contain the information.
Some people struggle with this because, if the book wasn’t present in the library, then the PC shouldn’t have had any chance at success on their Research test. This is a fundamental misunderstanding, however: Nobody at the table knows that the book isn’t there until the fictional cleromancy of the random number generator (combined with the GM’s ruling of what that outcome means) gives them that information. The library’s ownership of the necessary book is just one of a multitude of different external factors that could result in failure. (Other external factors might include whether the book has been checked out; if the book has been shelved incorrectly; has the book been damaged; does the book exist at all; and so forth.) The point is that we don’t care about any of these external factors enough to track it or model it mechanically, and so they all get abstractly bundled into the random number generator.
And, because all of these factors are bundled into the random number generator, it’s the GM’s responsibility to creatively unbundle them as they describe the outcomes of action resolution.
But what if we DO care about whether or not the specific book we want is available in this specific library? Well, in that case the GM would specifically determine that – through a listing of all the books in the library; or a list of all the places where that book exists; or maybe through a random percentile check – and then, like the armor bonus to AC, directly factor it into the success or failure of the Research test. (For example, if the GM knows that only one copy of the book survives anywhere in the world and they know that copy isn’t in this library, the Research test would automatically fail.) But when you make an external factor like this explicit, it’s no longer part of the abstract factors being modeled by the random die roll.
(It should be fairly obvious, of course, that no matter how many factors you make explicit there will always be factors you haven’t accounted for when you’re making a skill check. If there weren’t, in fact, you wouldn’t be making the skill check: You’d simply be defaulting to yes or saying no. Saying that the outcome of the action is random is inherently saying that there are factors that may or may not affect the outcome.)
FACTORS INFLUENCING OUTCOME
The ways in which characters can succeed or fail are as varied and limitless as the panoply of actions they can attempt in the first place. With that being said, there are some general principles you can keep in mind when describing outcomes.
SKILL: The most obvious of potential factors. Sometimes you have the best game of your life and sometimes you screw up and fall on your face. A lot of things can impact success or failure, but sometimes you succeed because you’re just that good (or fail because you’re just not good enough).
KNOWLEDGE: Is the character familiar with this particular model of safe? Do they recognize the patterns in a game of chess? Sometimes having just the right piece of information makes the difference between success and failure.
POWER: Sometimes people succeed because they just put more power into the attempt, or fail because they didn’t. A guard raises his sword to parry the barbarian’s blow, but her mighty thews sweep it aside and crush the guard’s skull.
FINESSE: And sometimes actions succeed because of the precision with which they are performed.
ENVIRONMENTAL FACTORS: Slippery floors. Frigid weather. Jammed locks. Floors that buckle under foot. Piles of fetid garbage that get in the way of your swinging sword.
TIME REQUIRED: How much time does it take to complete the action? High margins of success might indicate that the action took less time for the character to perform. A failure might result from something taking too long.
LUCK: Sometimes the biggest reason a character succeeds is because they’ve gotten lucky. The giant’s sword was going to take their head off, but it deflected off a falling piece of rubble. They were about to slide off a cliff to certain doom, but they grabbed a piece of scrub brush and miraculously its roots held.
THE TARGET: Whether the target is an object or a person actively opposing the character, they can obviously have an impact on the success or failure of an action. These are the locks that are devilishly difficult or the gullible guard who easily falls for your lie.
BYSTANDERS: In addition to the character directly targeted by an action, it’s possible for other characters to either interfere or assist in the attempt (whether wittingly or unwittingly).
TOOLS: You’re only as good as your tools. Lockpicks break, elven blades slide through seams in armor, inferior IC makes a system vulnerable, and luck charms crafted by your beloved can give the edge in a mystic duel.
These obviously don’t represent the totality of factors that can affect outcome, but hopefully they’ll provide a little inspiration.
(Way back in 1999 I wrote Dice of Destiny for Pyramid Magazine which mechanized this process by assigning qualities similar to these factors to individual dice in a dice pool system. If you find yourself struggling to diversify your outcome descriptions, you might want to check it out.)
MAKING FAILURE INTERESTING
Something else to remember is that the gatekeeper of mechanical resolution is that failure should be interesting, meaningful, or both. In other words, it should have consequences.
This can be one advantage of using external factors in explaining failure: If the character’s research at the library reveals that the book they need only exists in one place, for example, their next action will be to figure out how to get access to it.
What this means, in practice, is that failure generally should NOT cause a return to the status quo. This doesn’t necessarily mean failing forward, but it’s usually best if the outcome of an action – regardless of success or failure – should in some way change the situation. FATE refers to this as “blaming the circumstances”, and the advantage is that the new situation creates new options (which prevent the situation from stagnating or becoming a dead end).
(All of this also applies to success, but as I’ve mentioned previously this generally takes care of itself: Success implies that the character is one step closer to achieving their goals. A stated intention can almost always be summarized as “I want to change the current situation” and, therefore, the success of that intention automatically carries with it a change in the current situation.)
KEEP IT PITHY
Vivid descriptions are great, but try to get the ball back to your players ASAP.
A necessary corollary of making the outcome of an action interesting by giving it consequences is that you will have created a situation which (ideally) demands a fresh response from the PCs. Once you’ve established that new context, give the players the opportunity to make that response.
PLAYER DESCRIPTION
Instead of narrating the outcome themselves, a GM can instead prompt a player to provide the description. (Often this is the player attempting the action, although it can also be outsourced to other players at the table.)
For example, the GM might say, “You’re spotted as you try to sneak onto the mansion’s grounds. Who spots you?” Or, “You make a loud noise as you climb in through the roof. What’s the noise and how are you responsible?”
(Providing specific improvisation prompts like this – instead of simply asking a generic, “How do you fail?” – is generally more effective because it focuses the player’s response. You’re less likely to get a blank look if you ask a player to finish painting a picture instead of just handing them an empty canvass.)
Using the technique in this form grants the player a limited degree of narrative control. As such, it tends to work great in storytelling games (where it becomes part of a wider tapestry of methods for sharing narrative control). When used in a roleplaying game, on the other hand, I’ve generally found it problematic: It doesn’t really give the players any narrative autonomy (since they can only take narrative control when the GM gives it to them), but periodically forces them into a potentially disruptive and undesired authorial stance.
(In other words, if you want players to have that kind of narrative control, you’re probably better off playing a game that’s designed to do that.)
But that’s not the be-all or end-all of the technique. Instead of having the player get into an authorial stance and describe how the external world affects their character’s intention, you can instead have the mechanical result serve as an improv seed that informs how they play out the scene.
This can be particularly useful for social scenes: Instead of playing out an entire seduction attempt and then rolling to see if it succeeds, for example, you can make a Seduction attempt and then roleplay the scene based on the mechanical result.
This, however, begins to transition us into a discussion of fortune positioning, which is what we’ll covering in the next installment of the Art of Rulings.
Your research example is interesting. Personally, I don’t have any problem in principle with letting the roll decide whether a library has the specific book they need. The GM can hardly be expected to catalog the entire contents of every library the PCs might encounter, and some people might object to them making an arbitrary ruling that the information *can’t* be found regardless of the PC’s skill level. However, it does create a couple of potential issues.
First, what if the player wants to try again? (After all, the PC wouldn’t know if the book they need isn’t there or if they just haven’t found it yet.) If a failed roll *always* means the book isn’t there, so there’s no point in rerolling, that’s fine (as long as the players know it). OTOH, if *sometimes* the failure is interpreted to mean that and other times it’s not, then we’re back in the realm of GM fiat.
Secondly, how do you explain the fact that the PC’s skill level in research affects whether or not the book they’re looking for actually exists? (If they have access to several *different* libraries or archives, you could say that a higher skill level helps in determining which one is most likely to have the information they want, but what if they’re searching in only a single location?)
I’m going to answer those points in reverse order.
Re: Skill test determining whether or not a particular book exists. Bear in mind that this is not a test to determine if a particular book exists or not. It’s a test to determine whether or not they get the information they want. The existence or non-existence of a particular book is just one of many potential factors that could explain why they didn’t get that information, but which you don’t care enough about to make explicit (and therefore roll into the random number generator).
(In some systems you might also use a Research test to determine whether or not you can locate a copy of a specific rare book. But such a search would probably not be limited to a single library and, “Does the book exist?” isn’t likely to be a factor you don’t care about.)
Re: Retrying the check. This largely spills over into a question of when you allow people to retry skill checks in general. There are usually three standards:
(1) You can retry as often as you like. (You may have failed to pick the lock in the first minute you were working on it, but you can spend another minute fighting with it. There may or may not be consequences for each failure beyond time spent.)
(2) You can’t retry the check until you become more skilled. (This specific roll determines whether or not you are capable of picking this lock. In some cases, other improvements to your skill check — like the use of better lockpicks — might also allow a retry.)
(3) You can retry the check, but only if you modify your approach in some way.
Whether or not you allow someone to retry searching a library depends on how you, as a GM, would categorize that check. In this specific case, we’re talking about a situation where the PC has made the check once and you’ve decided to give them the partial failure result of, “You manage to identify THIS SPECIFIC BOOK which would contain the answer, but it doesn’t look like this library owns a copy of it.”
(I classify this as a partial failure because it doesn’t give you the information you want, but it does give you a method by which it could be obtained. One possible response to this, of course, would be to look for THAT SPECIFIC BOOK somewhere else. But we’re ignoring that option for the moment and just considering the question of retrying the Research test in this library.)
(1) If you classify it as a skill that can be retried as often as you like, then: Yes. Go for it! (How do you explain a subsequent success if THAT SPECIFIC BOOK doesn’t exist in this library? Maybe there’s another source for the information. Maybe a copy of that book was misfiled and digging deeper they track it down.)
(2) If you can’t retry the test, then: No. The result of their test represents completely exhausting all the possible resources this particular library has. They’ll have to look elsewhere for the information.
(3) If they can retry the test only if they modify their approach, then: Maybe. But the player will need to explain how they’re using the library in a different way. (Maybe they try to collate all the references to THAT SPECIFIC BOOK and see if they can piece together an answer from secondary sources. Or maybe they try to use the library to identify a local expert instead of finding the information directly.)
Your response to my second question doesn’t really seem to address what I was getting at; I may not have explained myself clearly. I understand what you’re saying about the die roll representing all the external factors that might affect the result. I also understand that a failed roll doesn’t necessarily mean that the book they’re looking for isn’t there, but you did mention that as one possibility.
Suppose the PC fails a Research check and you, the GM, decide that it’s because there are no books in that library that contain the information they’re looking for. But if the PC had a higher Research skill level, then they might have been able to pass the test with the *same die roll*. Which makes it seem like the PC’s skill level itself actually affects whether or not the library contains the needed information. Most of the other factors you mentioned — slippery floors, inferior tools, time pressures, etc.. — can be overcome by a sufficiently skilled individual, but “the library doesn’t have the information you’re looking for” wouldn’t normally be affected by skill.
I realize that’s is a very meta way of looking at it, and if the PC’s skill had been high enough to succeed, then you would have interpreted the low die roll differently. And like I said, I don’t actually object to your approach. I’m only mentioning this because it’s an odd sort of side-effect.
Ah, I see what you’re saying. You can also see the same effect if you imagine the same character coming back to the library later with a higher skill bonus that would have succeeded on the initial check.
If this was a Retry Standard #2 situation, you’d have created a weird situation: Their skill has improved so they can retry the attempt, but if their retry succeeds you’re going to struggle to explain how since you’ve already established that the library simply doesn’t contain the information.
But what’s really happened is that, when you said the library doesn’t have the information, you’ve implemented this as a sub-category of Retry Standard #3: Your initial check has permanently established whether or not you can succeed on this check, and you’ll have to modify your approach if you want to try again.
Introducing the issue of a second (or third or fourth) character who wants to retry a failed skill test can complicate things, but it still largely boils down to the same three categories. (Where it becomes complicated is that your assessment of which Retry Standard applies to this particular skill test can be different for multiple characters retrying the same test than it is for the same character retrying the test.)
In practice, I’ve found that the question when to allow or disallow retries for a particular check is a messy and imprecise thing that often depends heavily on circumstances. (And systemic attempts to codify when retries are or are not allowed often create their own oddities of logic.) And a lot of it does depend on how the initial failure was described (and how that changed or established the given circumstances of the game world).
A few other examples:
– You fail to convince the Duke to give financial support to your latest venture. Should you be allowed to retry that check? Should your 6 companions be allowed to make their own checks? (Does your answer depend on whether or not your companions’ interactions are part of separate conversations happening at some later time? Does your answer to the former depend
– The rogue with a skill bonus of +15 fails to disable a clockwork mechanism. Should the fighter with a bonus of +0 be allowed to step up and make a check? (Does your answer become different if they try it in the opposite order?)
Further complicating this is the issue of group skill checks. Does it make more sense to model your interaction with the Duke as a single check made by the character with the highest score? The character who takes point on the conversation? One of those characters but with +2 bonuses from Aid Another (or your system’s equivalent) granted by everybody else participating? Or does everyone make a check and as long as one of them / a majority of them / some arbitrary number of them succeed the attempt succeeds?
Another angle to be considered on retries is retrying when circumstances change, either due to the player’s actions or due to external factors. For example, the player makes a research check to search the library for information, fails, and you decide that that means that the book (or books) that contain that specific information is not present in the library. Would you allow the player to come back and make the same check in a later scene, multiple in-game days later, because the specific books needed might have circulated into the library?
Or to take another of your examples, you fail to convince the Duke to give financial support to your venture, so you go to the caves, fight a few troglodytes, dig up a couple of ice sapphires, and then bring them back to show to the Duke, as a demonstration of the potential riches his investment could bring. Do you let the player retry the check? Let his original result stand but modify it with a bonus for this new evidence?