Despite being obviously chaotic, there is also structure within this fractal composition. Based on an image by geralt on Pixabay.

Last week, in Into Each Plot, A Little Chaos Must Fall, I discussed the inevitability of chaos in adventures, the cumulative effect of three compounding sources of randomness: Player Decisions (unscripted player-plot interactions), Die Rolls (unscripted character-plot interactions), and GM Thought Bubbles (unscripted GM-plot interactions). In essence, if the last quarter or so of what happens when you actually play an adventure with your group is predictable, you’re doing something wrong!

Along the way, I discussed a couple of important concepts. The classification of adventure elements into eight orderly layers – three from campaign planning, one representing the accumulated experience of the players of events from the campaign-planning layers in previous adventures, two from adventure planning, and two that break down all of the above into individual scenes.

This is on my mind a lot at the moment, because the current adventure in my Zenith-3 campaign is about to come to an end. Act I connected this adventure with previous ones, dealt with a lot of consequences from those past adventures, and with developments in the personal lives of the PCs. It also introduced a number of new developments and problems, some arising out of those consequences. Act II was about the PCs beginning to wrestle these new problems to the ground and reacting to the events of Act I, and deliberately set the stage for Act III, which was the story of the “Pagani Perps”, and how they connected to an ongoing plotline, the “Figure In Black”, which is the focus of Act IV, which is what has to be completed for play this Saturday.

I also wrote about the adventure’s content and how I constructed it, with a multitude of excerpts from the adventure itself, in In The Beginning… Not! – drafting plots from the middle, which is more concerned with the writing process than the playing process, and which employed a three-act structure to describe it.

And that, right there, is illustrative of the chaos effect on plotlines: what was planned as a three-act adventure became a four-act adventure, one game session per act! Certainly, any adventure design beyond any broad outlines for Act IV that I had undertaken at the start of the adventure would have been completely out of date and would need to be 90% rewritten from scratch before it could be played.

But the subject of today’s article are the three sources of randomness and how I handle them, in campaign design, adventure creation, and at the game table. This can be as much as half the job of actually GMing, of telling – no, discovering – an interesting and engaging collaborative story with your players. So it’s a subject of no small importance.

Come with me, as I show you how to add a little order to the Chaos – just enough to put it to work on behalf of the campaign.

Implementing Player Decisions

Player Decisions, which almost always seem like a good idea at the time, are simultaneously some of the greatest sources of predictable chaos and the second-greatest single source of spontaneous anarchy that can befall any plotline.

The ‘predictable chaos’ part of that equation can be managed by anticipating, even deliberately incorporating, decisions by one or more players, and including parts of the adventure that can only be accessed in that particular form and at that particular time as a consequence of one or more player decisions – adventure branch points.

I discussed adventure branch points in September last year (in a somewhat different context) as part of How Many Molehills Make A Mountain?.

In the article linked to in the caption, I made passing mention of “unplanned plot outcomes”. But these are inevitably going to occur from time to time; whenever a decision is required, no matter how many of the possible choices you think you have anticipated, there’s a chance of an unexpected response. The anticipation is an attempt to manage the risk of this occurring.

But players can interrupt at any moment with a decision, whether you’ve anticipated one or not. That’s the ‘spontaneous anarchy’ part of the story.

How you react to player decisions is the heart and soul of your GMing style. As such, it can be a very slippery subject to discuss, full of exceptions and general statements; many GMs don’t even think about it consciously, let alone analyze the question.

Obviously, if you’ve anticipated a player decision and planned how you will respond to the most likely choices, which is what I try to do in my adventure design, the problem is pre-solved. Players rarely see the angst and difficulty involved, the bullets you’ve sweated in wresting a coherent and satisfying plot out of some of these turns of events, simply because they took the most obvious path; but it’s never an entirely wasted exercise to anticipate and prep for some of these less-likely choices, because it’s still good practice at devising plot threads and twists. For more on this aspect of adventure design, and this subject in general, see Ask The GMs: Giving Players The Power To Choose Their Own Adventures.

So, the focus of attention has to be on the unexpected decision, or the unexpected choice at a decision point. Johnn was the first to address this situation here at Campaign Mastery in Say Yes, but Get There Quick. My response tends to be rooted in a quick analysis of the decision that classifies the decision into one of six categories: Good choices, missed opportunities, misinterpretations, failures of memory, bad choices, and out with the pixies.

1. Good Choices

Good choices are choices that make sense in light of what the character knows. There are two sub-types: those that would still make sense if the character knew everything, and those that would not.

Good choice despite what the player doesn’t know

In the case of the first sub-type, my response is to let the character set about implementing their decision, essentially improvising the rest of the adventure based on what I had planned. The choice made might involve greater difficulties than the path I had identified, or it might avoid some of the complications that I had planned to use as setbacks along the way; so be it. Of greater concern to me at that point in time is how this impacts on the even distribution of spotlight time – some setbacks exist for no other reason than to give a player some spotlight time in overcoming the problem, so short-cutting the problem-solving process in this way creates an obvious imbalance. The other major concern is what the NPCs (especially the villains) know at each point and how their decisions (pre-baked into the adventure) would change as a result. I will frequently try to use this question to generate a solution to the spotlight time problem, if I can think of a way to do so.

The third consideration is “how much of what they don’t know do I need the players to discover en route to implementing this solution, and how do I get that information into their hands?” This is usually looking beyond the scope of this one adventure to the bigger campaign-level picture. On more than one occasion, I’ve put an adventure in place to achieve some campaign-level objective only to have the players solve the immediate problem without ever going near the campaign-level objective; when that happens, I have two choices, and usually have to make a snap judgment between them: either to insert a new adventure into the campaign plan to do what this adventure was supposed to do, or to revise the campaign plan to take into account the failure to meet that objective. Unless it’s something that is absolutely crucial to the campaign, I’ll pick the second choice.

As a general rule of thumb, success should follow a clever choice. It won’t happen very often, so let the player(s) have their moment of glory with good grace.

It should be noted that this is a very different reaction to the one I had under these circumstances when I was just starting out as a GM. Back then, as soon as a player understood the plotline well enough to short-cut the adventure, it would immediately be replaced with something that only happened to look like what had already taken place to this point in play. This had the advantage of making sure there was always plenty of play for the group, but the problem of discouraging clever choices resulting in dispirited players and half-baked plodding from start of play to finish. Why should players make an effort when they will succeed eventually anyway, no matter how faulty their planning?

Although the change in philosophy had occurred some time previously, the first time my players became aware of it was in Magneto’s Maze (described in My Biggest Mistakes: Magneto’s Maze – My B.A. Felton Moment) when, after a clever move by one of my players (and after a minute of jaw-dropped shock), I regathered my thoughts (pretending to still be shocked to buy thinking time) and let the player succeed even though it short-cutted the entire adventure. I noticed something very important at the time – that even though it (obviously) short-changed the other players at the table of their spotlight time, they derived almost as much satisfaction from vicarious participation in the clever players’ action. These days, I’m quite comfortable in simply telling the player “brilliant move, and not one that I expected – give me a couple of minutes to adjust”). It doesn’t happen often – maybe once in every ten or fifteen years – but I actually consider it a compliment to my ability to craft a coherent game world and rational narrative within that world – and to my players’ belief in my ability to improv, no matter what they throw at me.

Good choice unless you know everything

Things get more “interesting” when the player makes a choice that makes sense at the time, but is a mistake in light of the bigger picture because of things the character is not aware of. Obviously, the result is a plan that is doomed to eventual failure, but not for any reason that the player can currently anticipate. The player will do “A” and expect result “B” – and instead, get result “Q”. This constitutes a first hint that something else is going on that they have not taken into account.

My first mental question when this occurs is “How far can I let this progress before it becomes too late to solve the original problem?”. The second question is, “What happens to the campaign if the original problem isn’t solved?”. If the answer to this question is, “it’s catastrophic”, then it becomes clear that I can’t let the player’s choice continue beyond the point at which it becomes too late to solve the original problem; but, 95% of the time, it won’t be anywhere near that disastrous to the campaign, which gives me permission to let the players fail, if things get that far.

There is, of course, absolutely no guarantee that things will get that far. Every action that the players undertake, every result “Q” that results instead of a “B”, is a hint to them that they have made a mistake and need to rethink what they are doing. In effect, every action in pursuit of their erroneous choice creates a branch point in the resulting adventure that will bring the players back to the original adventure – possibly with more time pressure.

This is still clever play on the part of the player – the choice made was a correct one on the basis of what their character knows – and that needs to be acknowledged by the GM and rewarded. Hints as to the bigger picture that the GM didn’t intend to reveal until later in the campaign are that reward, and they can have long-term consequences for the campaign plan – and that’s something that I have to take into account, moving forward. That happened in the later stages of the previous Zenith-3 campaign, when the entropy drain into Karma’s space-time was discovered (briefly described under “Idea #18 – It’s Electrifying: Portals Are A Planar Battery” in Feel The Burn: Portals to Celestial Morphology Pt 4 of 4. This discovery took place almost a real-time year before I was ready with the plotline of “Dekhay and Ruin” – and turned the investigation of the phenomenon into an ongoing subplot before I was ready for it. The simplest solution was for the entirely-reasonable investigation to yield no positive results for ‘long enough’ (about 6 months, real-time), then to dribble out hints and clues as to what was going on over a further period of time (another 6 months, real-time). In game time, it was a three- or four- week time-span from discovery to resolution, as I remember it, but the pacing was dictated by the speed of plot. It just so happened that it took about a year for other plotlines that could not be delayed to be resolved, clearing the decks for me to have enough time to write the adventure in question.

In effect, then, the plan of action on the GM’s part when the player makes a choice that is only reasonable if you are ignorant of what the character doesn’t know, is to let them proceed while dropping hints through unexpected outcomes that there’s more to the story than the PC has taken into account, and to continue until the critical point (if there is one) is reached, or the players take the hint.

There’s more information on this subject to be found in Domino Theory: The Perils and Practicalities.

Ultimately, I view this eventuality as a failure on my part as GM – I’ve built something up to a greater immediacy of imminent threat than I should have, haven’t communicated something effectively enough, or have failed to anticipate this possibility in my planning.

2. Missed Opportunities

The proposed choice might be reasonable but miss an opportunity that I expected the players to take advantage of, often because the decision is being taken prematurely, or because the players think that the opportunity entails too much risk. Once again, the default choice is to let the players proceed, but I will be on the lookout for an opportunity to raise the possible missed opportunity with the characters through an NPC or a die roll if and when it becomes reasonable for them to recognize it. Depending on when that opportunity arises, it may or may not be too late to change course, or the player may respond that they “have thought of this option, but…” – in which case, that’s their choice and I’ll deal with any campaign repercussions after the day’s play.

3. Misinterpretations

More serious mistakes are made when the players misinterpret the situation and hence make a poor decision on the basis of this misinterpretation. The theoretically-best solution, and the one that I employed in my gaming for a very long time, is to let the players proceed with their choice and let them discover the mistake the hard way. The problem is that this opens the door to Confirmation Bias – where the players fail to recognize that they have made a mistake until the situation threatens to derail the entire campaign. I discussed this situation extensively in “I know what’s happening!” – Confirmation Bias and RPGs, in which I identified 9 possible responses to the problem and analyzed them, including the five “do nothing about it” options. These days, as a result of that analysis, my first thought is to analyze the proposal in terms of accepting or rejecting option #6, “The Players Are Right” – usually rejecting it, but not always – and my second choice is usually option #7, “Correct the error Immediately,” perhaps in combination with #8, “Die Roll Saviors”. The reason is that if you don’t take corrective action immediately, players tend to lock their misinterpretation in stone, where it will cause future decisions to be made incorrectly as well. I’ve even been accused of ‘reinventing history’ by correctly reciting campaign history without the misinterpretation!

4. Failures Of Memory

Which brings me to variant four – a decision made in a particular way because there’s something the player has clearly forgotten to take into account. When this happens, I immediately give the player a roll to “remember” the missing information – with big bonuses toward success if it’s something the character is unlikely to forget – and if they succeed, I remind the player of what they have forgotten (but that the character has not) before accepting the decision as final. It’s also necessary to give the player room to re-orient their thinking – how much depends on the player; sometimes, that’s best achieved by taking a short break from play, sometimes you can bring forward a scene involving some other player’s character, sometimes you can simply proceed with play. Learning how mentally-agile a player is (and it does vary over time and with experience) and hence how much mental space they need to make such re-orientation, is one of the most critical aspects of getting to know a player.

Special note should also be made of any differences between how flexible the player thinks they are, vs how flexible they really are. You need to accommodate the latter while appearing to accommodate the former, which can sometimes be difficult – the trick is to fill the gap with something non-distracting. Obviously meaningless in-character side-chatter between another PC and an NPC is often the best choice, or having another PC focus on some element of the environment (flavor narrative that tells the other players more about the environment) while the decision-maker is trying to grasp the revised problem and formulate an alternative course of action.

Unfortunately, with some players, by the time they have finished reorienting their understanding of the situation, mild Confirmation Bias has set in, and the only solution they can think of is the one that has just been ruled out – “I don’t know what to do” is the result.

This is a more delicate situation than it may initially appear. If you simply lead the player through the logic that leads to a correct solution, the player will – at worst – think that you are running a plot-train, and – at best – feel that you have taken control of their character out of their hands, causing frustration and dissatisfaction on the part of the player. Neither is all that desirable.

I have to admit that I’m still looking for the best solution to this problem. At the moment, I’m leaning toward a mix of die rolls and involving other PCs or NPCs – in particular someone who can say, “explain the whole problem to me again, from the top,” and who can then put their metaphoric finger on the mental block that’s impacting the player in the spotlight. Solve that, or correct the flawed assumption, and the player in question is usually capable of evolving the correct solution on their own.

It’s unfortunate but inevitable that sometimes, this takes what should have been a snap decision and draws it out to five, ten, fifteen, or twenty minutes of game time. And, meanwhile, the other players are twiddling their thumbs or getting into side conversations (or worse yet, thinking up mischief for their characters to engage in). In truth, players are usually aware that the game has ground to a halt while they try and figure a way out of their problem, and this is likely to add to their frustration.

Again, different players will react to such lulls in different ways. Learning those reactions, identifying which ones are harmful to the game, and finding ways of preventing those reactions before they occur, is another essential part of the GM learning his players’ styles and personalities. This might make the choice of which PC to invoke as a sounding board obvious to the GM – he then just has to find a way to invoke that choice at the gaming table. This is an aspect of the GMing art that I freely confess I’m still learning, so take my advice with a grain of salt.

5. Bad Choices

Some choices are just bad, or silly. For example, there was a time when Stephen Tunnicliff was feeling particularly put-upon. One of his more outrageous schemes (to monopolize the Brazilian coffee crop) had collapsed, flaws in some of his patented technology had emerged unexpectedly, another PC had just dumped on him for re-exposing the game world to a threat that the PCs had only barely survived the first time around through a series of selfish and foolish decisions (a combination of manipulation by the threat in question and an entirely reasonable desire on the PC’s part), and a third PC had just contrived to steal control of his PCs company from him. The response was Stephen at his worst – in the heat of anger he petulantly decided to accept an invitation to appear on the Johnny Carson show and reveal the other PCs secret identities.

When he calmed down, and I explained an aspect of the game rules that he had overlooked (his character had paid character points for the company – if it got taken away from him, either he got the points back or I as GM would arrange for him to either get his company or an equivalent replacement back), he realized the enormity of what he had done. Fortunately, from the moment the PC had made the first foolish decision (the one that led to re-engaging the old enemy, over a year earlier in real-time), I had been planning a solution for when it all inevitably hit the fan, and when he announced his Bad Choice, I had time to conceive of a solution. First, UNTIL had perceived that the character might do something foolish when the stock market manipulations were discovered, and had intervened to cause “technical problems” during the key moments of the broadcast, protecting the other PCs; second, I implemented the plotline that I had been preparing for this (general) eventuality: after the broadcast, thinking that he’d at least gotten some payback, the character had gotten himself drunk (not an easy thing to do with his Constitution, he needed Asgardian Mead to make it happen). In Act I of the consequent adventure, the drunken PC was attacked and killed. In Act II, the character awoke, thinking it was a year earlier, and discovered himself to be in a stasis chamber that he had designed (in game session 7, several years earlier) as a last-ditch life-preserver. He discovered that the power supply had been interrupted as collateral damage from the destruction of his factory about 2 weeks earlier, but that the team had dealt with the threat responsible, and had announced that the character had seemingly been killed. Eventually, the backup energy storage had become exhausted, and the wake-up process had been automatically implemented. He also found that he had vague memories of being struck on the back of the head with enough force to render him unconscious and no memory of being placed in the stasis chamber. Teleporting to the team headquarters, he greatly astonished them by still being alive, and was equally-greatly astonished to discover all the things that “he” had done in the last year. It was subsequently discovered that his research in developing cloned replacements for any team member killed in action – at the heart of a plotline a year earlier – had gone badly awry when his personality-duplication process created an Evil Clone…

That’s how to handle an outright bad choice – immediately begin preparing contingency plans for when the train-wreck can no longer be denied, then give the player in question enough rope. The assumption is that when the player sees how disastrously wrong his choices went, he will avoid making the same mistakes a second time.

And, to be fair to him and his memory, that’s exactly what Stephen did. In fact, he ended up having a lot of fun resolving the various situations his “clone” had gotten him into, some of which turned into pivotal building blocks later in the campaign.

Let me be clear about it – it takes a lot of confidence in the players on the GMs part, and a lot of trust in the GM on the players’ parts, to implement this technique. Early in a campaign with new players, I would be far more likely to treat this as a “Misinterpretation” situation and – if necessary – be far more heavy-handed in putting a stop to things before they went too far. The events described coincided with the campaigns’ tenth anniversary – part of the reason why I had brought so many things to a head at once for the character – so I had the experience in dealing with the players that was needed, and the result was a memorable adventure for all concerned.

6. Out With The Pixies

Some decisions go beyond bad, to the epicly stupid. An example that came up on Quora recently demonstrates what I’m talking about: the poster wanted advice on how to handle a Paladin demanding a DEX check to pluck the eye from a sleeping PC and replace it with the Eye Of Vecna without the prospective victim noticing.

I don’t know about you, but my first reaction on being confronted with such monumental idiocy is to gape and wonder “where do I start?”

To qualify for this status, the decision in question has to be catastrophic for the campaign. So I would start by explaining that this is too much for a simple DEX check – removing the eye would require an attack, and that succeed or fail, it was practically certain to wake the other PC, negating any hope of the character not noticing – in other words, taking the starting point of the proposed plan of action and demonstrating the first major flaw in the proposal as a hint to drop the idea. I would also ask him to think about how he would feel if someone else were to do this to his PC.

If the player persists, deciding to drug his victim for example, I ask where he’s getting the drug in question and pointing out that the PC being targeted has a change at spotting the unwanted seasoning. But I also note that he is the type who won’t take a hint. That’s when I start piling consequences onto the character – even planning such an action is an alignment breach, and would attract attention and penalties from greater powers, who will take an increasingly hard line in telling him “don’t do it”. And if the player, stubbornly, persists, all sorts of creatures who want the magic item in question would start turning up to try and claim it, until one succeeds. If that then means that the Paladin in question acquired a personal quest to recapture the Eye, that’s fine – the campaign is back on track, with a new ongoing plot thread for the GM to play with.

And if the Paladin managed to hold onto the eye despite all this, and persisted in going through with his plan, I would let him attack the PC, let the other PCs wake up in time to prevent the removal of the victim’s eye, and let justice take its course – then thank the Paladin’s player for participating, see you later. Because keeping him in the campaign would do more damage to the campaign than letting him go. I would, however, be quite willing and even eager to talk about what happened and why with the player.

Most players won’t let things get that far. They will take a hint, and discover why it’s a monumentally stupid thing to do, and learn from it.

It’s worth making another point: I will generally not kill a PC unless they are doing something monumentally stupid. “The bathwater’s too cold.” “I throw the toaster into the water to heat it up”. Alarm bells ring…

Random Sources Of Chaos

The second source of randomness is the outcome of die rolls. These always entail the risk of an unlikely result, even if that risk is an unlikely eventuality. If there is no risk of failure, the GM should think about why he’s requiring a die roll at all instead of simply telling the player, “you have enough skill in [whatever] that you succeed automatically”. So die rolls should always be meaningful. See The Nimble Mind: Making Skills Matter in RPGs for more on that subject.

In fact, I always bear in mind the reasons for requiring a die roll when one comes up. These might be “the risk of failure” or “the potential for an improbable success” or simply injecting chance into the adventure so that the players feel the risk of failure.

How you interpret a die roll is always a function of the motivation for requesting a die roll, shaping the course of events thereafter. But that’s only the beginning of the story. Die rolls can be classified into four situations, depending on the outcome relative to expectations.

Expected Successes

When you expect a character to succeed, success simple permits the adventure to continue along it’s expected course. Of course, there’s no need to make it sound as easy as it was; there can be a wide gulf between an objective evaluation of the chances of success and the actual chance put in place by the GMs.

The Adventurer’s Club and Zenith-3 campaigns have very different philosophies on this point, deriving from system mechanics in the ultimate analysis – which is interesting since they both derive from a common system, albeit from different generations of that system.

In the Adventurer’s Club, my co-GM and I are always careful to assess any task in terms of how likely a character of the skill level of the PC should be in attempting the task in question, then applying modifiers to the die roll accordingly. A character of skill 14/- (i.e. 14 or less on 3d6) should be able to attempt tasks that a less-skilled character wouldn’t even think of, with an expectation of success. Even if we don’t make this clear to the players (and we usually don’t), the accumulated psychological impact of successes and failures influences the thought patterns of the player when he’s “in character”, to the point that it becomes possible to reverse-engineer the modifiers and gain an insight into our thinking when the difficulty was assigned. Part of that thinking is whether or not success or failure will hinder the course of the adventure beyond any tolerable level – in other words, do we want the character to succeed for plot reasons?

In the Zenith-3 campaign, the skills are defined in terms of how difficult it is for a character of given standing to succeed, the standings being unskilled/amateur, hobbyist, employable, professional, expert, and genius. Where a character falls within this spectrum is defined by their skill level in the subject. A skill of -35% or less is unskilled/amateur, -34 to -15% is hobbyist, -14% to 15% is enough for the character to earn an average wage or better using the skill, 15% to 60% are various shades of professional, 61%-100% is an expert, and 101%-150% is a genius. Modifiers to the chance (not the die roll) range from -75% to +75% – but the typical modifier for an easy task is +25%, so the system is weighted toward positive modifiers. However, some players have had trouble wrapping their heads around skill values of less than zero being “adequate”, so a planned revision adds 25% to the scores and loses the positive bias in the modifiers. The difference is that individual skill levels don’t have to get taken into account during adventure creation; instead, I decide on the minimum level of expertise required and how difficult it would be for the base someone of that expertise.

Expected successes are easy to deal with. But you always need to be prepared for the unlikely failure.

Unexpected Failures

An unexpected failure should always have an effect on the plotline. That effect could be immediate or it could be delayed, but eventually, the failure will send the adventure down a different path. Planning for this usually amounts to two things – identifying the plot consequences of the failure and finding an alternative mechanism of achieving whatever a success would have done, i.e. recovering from the failure. The failure could be a momentary aberration or hesitation, or it could be a serious misstep with protracted effects. I always assess the consequences of a failure first in campaign terms, second in adventure terms, and finally, in immediate terms:

• • Campaign terms: will the failure lead to a catastrophic problem within the campaign? If so, at what point will it become too late to correct the problem? How can I plausibly let the PCs recover from the failure at or before this point? What is the optimum point, in campaign plausibility terms, for that recovery to take place, allowing for someone to perceive the failure, decide to take corrective action, plan that action, put that plan into motion, and have the PCs become aware of that intervention – all preferably while avoiding a dues-ex-machina?
  • Adventure terms: the list is almost identical to the campaign list, with one notable difference: if it won’t be toxic to the campaign, I’m perfectly prepared for the PCs to have a partial success or even a failure in any given adventure, if that’s warranted – and can be made sufficiently entertaining along the way. That said, such failures are often radioactive to the campaign, or unlikely to be enough fun – and that means that recovery has to take place within this adventure, or a sequel adventure at the latest. The other elements of recovery planning have already been assessed under the ‘campaign’ heading, it’s just a question of timing.
  • Immediate terms: this is hard to describe generally, because it’s so dependent on what it was that the PC was attempting. But failure generally comes in three flavors: qualified success (a GM’s cheat), failure with the potential for an (eventual) second bite at the cherry and some short-term discomfort or trauma, and abject failure with no hope for redemption without active measures.

The great thing about unexpected failures is that the other players are usually aware of the failure, and of the course events were expected to take; after a moment of enjoying the suffering of their fellow-player, they become active co-conspirators to any attempt by the GM to get things back on track. In fact, if the GM doesn’t act quickly enough to instigate a correction, they may take matters into their own hands. That’s all fine, too – the important thing is that the campaign or adventure recovers from the failure, not how it happens.

Expected Failures

Improbable chances of success are a part of life when undertaking challenging things. A lot of adventure planning lies in identifying places where the PCs are unlikely to succeed in direct action and creating a route around that expected failure. A GM has two responsibilities when it comes to expected failures: making sure that the character is not permanently impaired by the failure (or worse yet, killed) – after all, the GMs put him in this position – and making sure that the campaign and the adventure are not permanently impaired by the failure, as described above.

Beyond that, they should convey a sense of “Okay, you’ve tried the obvious and it hasn’t worked – what’s next?”

As with expected success, achieving the expected outcome isn’t a huge problem if you’ve planned properly.

Unexpected Successes

The unexpected success is whole different story. It can shortcut the entire adventure if you let it stand, or even radically reshape the campaign. If a PC in the adventurer’s club campaign decided to assassinate Hitler, and we were foolish enough as GMs to let them make the attempt on a single die roll, no matter how improbable a success might be, by definition there has to be a chance of success – and that must be planned for.

A more realistic example: so you’ve set things up so that a rare plant is needed to heal someone, and the PC with the medical skill wants to make a medicine roll to come up with an alternative treatment. You’ve decided to allow the roll because of the near-certainty of failure, which will then propel the PCs into the main adventure. You announce that the character will need to roll an 18 on 3d6, twice, in order to come up with such a treatment (even though that’s not in the rules – a clear signal that the GMs don’t want you to succeed!) Despite the odds, you make the first roll – a one-in-216 chance – and then gather the dice for the second attempt. Now the GM is nervous – the odds are vastly in his favor, but where lightning has struck once, it can strike again (in contradiction to the old adage). And indeed, once again, the player rolls an 18 – something that will happen only once in 46,656 attempts, according to the odds. The player rejoices, the other players raise eyebrows and exchange high-fives (or their equivalent) with the successful player, and all eyes turn to the GM….

There are a lot of ways out of this bind.

• • The most obvious is a success that the player chooses to ignore because of the in-game ramifications – an alternative treatment that involves the amputation of all four limbs and an 85% chance of the patient’s death, for example. After all, he also knows how improbable his success was. This is the conditional success.
  • An alternative is the flawed success – “You remember perfectly exactly what you read, word for word. However, as you recall the words, you find a serious error in the statements, one that destroys your trust in the information.”
  • A third technique is the costly success, in which the success sets the feel on a path to inevitable success but reveals that the price of following that path is too high.
  • A fourth technique is the illusory success, in which the character appears to succeed but in reality achieves only a partial success – like assassinating Hitler’s body double.

There are others, but they are all variations on the theme of undermining the success to the point of it actually being a failure.

At the same time, it has to be remembered that the character has pulled off a miraculous success, and needs to be rewarded for that in plot terms.

Let’s take that “more realistic example” cited earlier. A success despite the odds might result in the character being completely convinced that there is no other course other than retrieving the rare plant, but knowing a way of eliminating contaminants from the plant so that the side-effects that would normally result from the treatment will not do so in this case. What’s more, a paper describing this refinement to the process – after the fact – will earn the character a tidy sum and be a genuine advance in medical knowledge, bringing with it civic honors and recognition.

It’s entirely reasonable to anticipate a medicine check under the circumstances. For the sake of the adventure, it’s a roll that the GM has designed to fail – but there’s always an outside chance of success (or there’s no point in having the character roll), and that means that the GM has to plan accordingly. It doesn’t have to be extensive – the couple of lines given above are a reasonable example – but you need to be ready with an answer.

Unexpected Die Checks

In fact, the only time when a GM can justify being caught short by an unexpected outcome on a die check is when the GM wasn’t anticipating a die check at all.

This is frequently the result of a player announcing an unanticipated course of action (see earlier in this article) and the GM insisting on a die check to see whether or not the action is successful (regardless of whether or not it will have the desired effect). Sometimes it’s the result of the GM forgetting a character capability, and sometimes it’s a consequence of the player applying an ability in a way that the GM hasn’t thought of “I cast Blade Barrier down the purple worm’s throat – what happens?” (an actual example from the latter days of the Fumanor: The Last Deity II campaign).



It’s at times like this that I fall back on the hierarchy pyramid, last shown in The Language Of Magic: A Sense of Wonder for the Feb 2019 Blog Carnival – again, in an entirely different context – that was first proposed in Blat! Zot! Pow! The Rules Of Genre In RPGs (Jan 2011) and re-examined in The Blind Enforcer: The Reflex Application Of Rules (April 2014).

* What do the official rules say (if anything) about any similar situation?
* What do the house rules say (if anything) about any similar situation?
* What seems the most “realistic” given the base assumptions of the campaign world?
* What seems the most appropriate interpretation given the genre of the game?
* What is the interpretation that works best in terms of the adventure plot?
* What is the interpretation that works best in terms of the health of the campaign?
* Are there any practicality considerations that should be taken into account?
* What is the interpretation that will produce the maximum fun?

Remember that any subsequent answer overrules one that’s already in place – the house rules trump the official rules, “realism” trumps the house rules when they are inadequate, genre trumps “realism”, plot needs trump Genre, campaign needs trump the needs of any one plot, practicality of implementation trumps everything else, and fun trumps all.

Beyond that basic principle, the usual considerations apply.

Readers should also find Narratives Of Skill: How To ‘Improv’ Outcome Descriptions In Advance to be useful information in this context.

The Purest Anarchy

It has been said that no adventure survives contact with the players. It has also been said that the GM is just a special category of player.

Given that players are the source of two of the most omnipresent kinds of randomness, the truth of the first statement seems amply demonstrated. This section is proof that – at least in the context of the first statement – the second statement is also true.

GMs go off-script regularly. And repeatedly. Usually for what seems like good reasons at the time. Some of the time, the reasons are, in fact, valid! More often than not, though, they are either a very bad idea, or an idea that won’t be properly executed because the GM hasn’t prepped properly for it.

One of the most frequent causes of offense is the misplaced explanation. Something is said or done in the adventure that isn’t explained at the time, and a player interrupts the GM with a question. He frantically checks through the surrounding text, only to find no explanation. So he makes one up on the spot. A page or two – or perhaps, a paragraph or two – later, the explanation around which this part of the adventure has been built surfaces, and contradicts the explanation the GM has already given. And suddenly, he’s in trouble, and his adventure is in a state of anarchy – whether it’s obvious on the surface, or not.

The next most frequent source of GM-induced anarchy is the flash of inspiration. The GM is part way through the adventure when he suddenly has a blinding insight into how he can make everything more dramatic, more exciting, more interesting, more fun! And, without properly thinking through the ramifications, he changes course abruptly, throwing away whatever is contradicted by the rest of the prepared adventure if not the whole thing entirely. This, of course, is just asking for trouble.

Studies have shown that rushing increases the rate of human error eleven-fold. I’ll say that again: decisions made in haste are eleven times as likely to be flawed or otherwise poor decisions. Now, GMs are used to making quick decisions under time pressure – but so are the military commanders and pilots that were the subjects of those studies. Amongst those not so trained, the rate can be even higher – as much as twenty-five times as high according to some sources, due to the stress involved alone (Reliability and Maintainability and Risk, by Dr David J. Smith, 7th edition, Appendix 6). What’s more, after 1 minute in an ongoing emergency situation, the error rate rises to ninety times the base, showing the effect of panic. The first task in any such circumstance is to get control of the situation so that the error rate of subsequent decisions is reduced to the high-stress value or better, a significant improvement! But Douglas Adams said it best, in The Hitchhiker’s Guide To The Galaxy: Don’t Panic! (the link is to an omnibus edition including all five books in the series).

I tend to classify these spur-of-the-moment brain-bubbles on a scale with eight divisions, though it’s important to note that you can only correctly classify them after the fact unless someone points your idiocy out at the time – I’ll offer an example or two of that as we travel. The eight divisions are:

1. Second-Skin Decisions
2. Wallpaper Decisions
3. Enhancing Decisions
4. Neutral Decisions
5. Decisions Of Inconsistency
6. Obstructing Decisions
7. Derailing Decisions
8. Catastrophic Decisions

Second-Skin Decisions

Minimalist prep often means that you have to invent characterization details on the spur of the moment to give an NPC color. Those are second-skin decisions, and they tend to be relatively harmless – noteworthy only when they fail for some reason.

Cliches will leap to mind at such times, but must be spurned lest your characterization be cringe-worthy; but, at the same time, you can’t make the character too unusual or you’ll distract from the substance that they are present to provide within your adventure. But avoid either of those extremes and you’ll usually be fine with such decisions – and the worst that will usually happen is that the character will be a disposable throwaway, personality-wise.

But sometimes you’ll forget that the character has already appeared in the game – “Greetings, Mon Ami! And ‘ow can I assist you today?” “Wait a minute, didn’t you have a Scottish accent last time we met you?” Whoops. On rare occasions, you can (literally) make the error a throwaway piece of characterization: (Texan accent) “Why, Ah surely did, pardner! Jest gettin’ in some practice for this-here radio play ah’m doin’ next week.” (No accent) “So, what can I do for you?”

Suddenly, the nobody is a voice actor with a gift for accents. Characterization, done.

This trick won’t work all the time – if the error is of a more factual nature, you may be in trouble. But it’s one to keep in your back pocket for the times when it is useful.

When it won’t work, my number-one solution is to admit that the player raising the question has got me – a minor mistake that can be easily overlooked and isn’t too disruptive of play, but that humanizes me as a GM, making more serious errors more forgivable – if they don’t happen too often.

Wallpaper Decisions

Every campaign has a look-and-feel built into its subtext that the players absorb almost subconsciously. All it takes is the GM deliberately varying his vision of various staples and fixtures, especially of architecture and social behavior, in his own mind – that will then ‘wash into’ and color his descriptive language, and the behavior of the NPCs, and that will be absorbed almost subconsciously by the players.

“Like every village you’ve ever seen within this woe-begotten kingdom, the streets are crudely cobbled and dirty, the roofs are soot-stained and low, rising abruptly to a peak above the chimney, the walls are plastered in faded colors, and the locals who aren’t dirty are covered in grime and filth. You’re sure that they bathe religiously at least once a decade. A serving-woman with greasy hair beneath her off-white cotton cap made of scrap linen tosses kitchen garbage and a pot of a soup with something green floating in it into the street, where it lies in a steaming puddle before slinking away.”

Wallpaper decisions are spur-of-the-moment “enhancements” and “refinements” to this look-and-feel. Sometimes, that’s exactly what they are, making the game world richer for their incorporation; sometimes, they simply perpetuate the established flavor; and sometimes, they will conflict with it – and make no mistake, there’s a big difference between contrasting with the established flavor and conflicting with it.

A contrast is self-justifying by the difference between specificity and generalization. Not all places, even in such a run-down setting, will be uniformly dank and dingy; some will be better, and some worse; it’s all relative to the baseline. A conflict is a contrast so striking that such a justification would ring hollow; to make it plausible in this context, it needs to be justified in some equally-exceptional way. When you do so, it transforms an Inconsistency in the game-world’s characterization into a deepening and enriching addendum.

The only way you can tell that this justification has been successful is when the players accept it and begin to construct a generalized statement about the exceptions that demonstrates that the exception has been integrated into their world-view.

For example, I might describe a typical inn one way (hessian lining totting away under the floorboards to capture spilled ale creating a perpetual odor of stale socks throughout the common areas) and one particular inn as being spotlessly clean, immaculately presented, with timbers fitted to near-perfect tolerances and rich tapestries upon the oaken walls. Especially if the exterior doesn’t look all that exceptional, this contrast is so strong that it simply doesn’t belong in this setting. So, it needs an extraordinary justification – at the time, if possible. “It’s run by an Elf, a refugee from the Silverwatch Reaches.” If the player’s response is, “Okay, from now on we’re only staying at Elvish Inns if we have any choice in the matter,” and in the next village they visit, they ask “Is there an Elvish Inn here?”, then they have bought the justification and integrated into their world-view of the campaign. If their response is “it makes a nice change, but are they charging arm-and-a-leg prices?” then they haven’t really bought into the justification, and the world-view has been muddied just a bit by the exception.

Such are the rewards – and potential penalties – of making wallpaper decisions on the run.

Enhancing Decisions

Reinventing campaign history, game-world relationships (especially political or economic ones), pseudo-scientific principles, campaign metaphysics, villain motivations, and the like on the fly are all “Enhancing Decisions” when everything works perfectly. And something far less attractive when it doesn’t.

We all have to do this from time to time. We’re confronted with a hitherto-undiscovered plot hole and have to wallpaper over the cracks right now before they can derail play. We create something spontaneously to fill a particular need or desire (“The Imperial Capital of Korea”) and then have to make sense of it.

When it works, it makes you look brilliant. When it works really well, maybe you even are brilliant. When it doesn’t, though, it can be catastrophic to verisimilitude for months or even years to come. The most dreaded phrase a player can ever tell a GM when he isn’t courting it is “I just don’t buy it”.

Dealing with such problems can be a complex problem, and is beyond the scope of this article. Instead, let me refer you to my (still unfinished) series, The Elephant In The Gray Room, which addresses such problems in escalating order of seriousness.

Neutral Decisions

These are theoretically possible – I think. I’ve never seen one. Decisions that aren’t designed to address a deficiency of character, plot, prep, system, or setting, or to improve one or more of these things? Talk about your black cats at midnight in coal cellars that aren’t actually there – good luck in finding one.

Actually, let me take all that back – there is one situation that I can think of that could qualify. It’s when you identify a plot hole before you get to that part of the adventure, and so drop in a replacement or supplementary explanation for whatever is going on that avoids the hole; your goal is not to ‘improve’ anything, it is to substitute something for an equivalent of equal value (by all standards that can be applied) that has some sort of logical or inherent problem.

Decisions Of Inconsistency

Once you pass below the “neutral decisions” limbo bar, you’re into problem territory. It doesn’t matter what your motive was in indulging the moment of madness, it has created more problems than it has solved, and it’s only a question of severity.

The least cataclysmic outcomes are decisions that produce an inconsistency in a character or a character’s behavior beyond that which can be explained through rational means. A character can’t be generous one week and greedy the next; a civil war can’t be won by one side one week and by the other the week after – unless a mutable history is part of your game setting, of course! – and so on.

At the very least, there needs to be an explanation, and that’s where you find the decisions that fall within this category (if the consequences are more severe, like unraveling a character’s motivation for a particular action, you may need to take stronger steps). Such explanations can be delivered more or less as an aside to a subsequent conversation being role-played. “Leonora’s been giving everyone a hard time this last week, her husband left and took the kids and it’s really soured her outlook on life. I hope she gets back to her old self again soon.” The sooner you insert such an aside, the better and the more plausible it is; if it’s part of the same adventure or even the same day’s play, that’s ideal. But, before you can drop-in such a narrative point, you need to identify the need for it – and that requires recognizing the inconsistency or having it pointed out to you.

And if it’s too late, the affected adventure has come to an end and a new one started, in which the inconsistent character has no role? Well, you can either alter the plot by inserting a subplot that has not purpose other than patching the inconsistency, or you can choose to wait until a logical opportunity presents itself within the campaign. Remember to take the logical passage of time into account if you choose the latter course – “Hi there, good to see you again. I’m sorry if I was a bit of a pill last time – a family crisis and I’m afraid I took it out on everyone. All good, now!”

That’s not to say that you can’t employ a more powerful solution from one of the categories below – just remember the information I gave on error rates earlier, and plan it carefully! Maybe Lenore was being mind-controlled….

Obstructing Decisions

An obstructing decision is one that makes sense in the context of the Adventure but imposes a road block on a subsequent part of the campaign, usually by preventing or inhibiting some consequence at the campaign scale. At the very least, this requires a subplot to alter the situation in-game such that the consequence is “back”.

That becomes a lot more problematic if the players correctly identified that consequence as an outcome of the situation if they didn’t intervene and deliberately chose actions to avoid the outcome that the GM desired but that they did not. In such cases, a single subplot probably won’t be enough, you’ll need to implement a more significant solution. You will need to construct a circumstance, or more probably a chain of circumstances, which convinces the players that the outcome you want is inevitable and that at best they have delayed that inevitability, and that’s beyond the scope of an obstructing decision.

Derailing Decisions

Slightly more serious are decisions that you realize after the fact will derail the adventure you’re currently playing. That’s for two reasons: first, the players will almost certainly react immediately to the flawed brain-bubble’s content, entrenching it within the campaign; and second, because you don’t have time for leisurely reflection and planning, you have to make a decision on what to do right now.

Of course, this isn’t the worst-case situation, even within this category; worse by far is when the players react to the moment of madness by calling you out. Why? Because (a) this engages your natural defenses; and (b) because you are more prone to either entrench the mistake within your own thinking because of (a), or to invent something else on the spot to resolve the problem. 90% of the time, either response will only make things worse. This is where those error rates really come into effect.

Okay, so there’s a big difference between a real emergency and a plot-execution error in a roleplaying game. For most players and GMs, their emotional involvement will still be high enough that reactions will be similar, if not quite as intense. Pilots and racing-car drivers train in simulators to prepare themselves for emergency situations, checklists are used because snap judgments in a crisis are suspect, and more checklists are employed because we tend to skip over important things in routine situations, and so on. Fire departments and police train all the time to make better decisions in emergency situations, and the military invented all this training. The GM may be in a lower-stress environment, with less of significance riding on his decisions, but all his training has to be acquired on-the-job. So I think it’s justifiable to use the same 11-fold factor for all those situations.

That means that as soon as you become aware of the problem, you have to stop play and call a five- or ten-minute break. This takes the urgency and stress out of the situation and gives you some time to think. Go somewhere where you can’t hear the players, who will probably commence a side-discussion if your group is like mine, and think.

Priority #1 is to protect the campaign. Priority #2 is to salvage the adventure if you can. Note that protecting the campaign doesn’t necessarily mean slavishly following the master plan that existed before the mistake; by throwing part of that plan away, or substantially reinventing it, you may be able to mitigate the effects of the mistake. Protecting the campaign might mean embracing the mistake, making it canon, even if that means completely revising the rest of the adventure and a future part of the campaign.

The first thing you need to do is therefore to analyze the consequences of the problem that you’ve created. Only then can you begin to assess potential solutions. I’ve examined such solutions in detail in The Elephant In The Gray Room, Pt 3 of 5: Significant Repairs and, to a lesser extent, The Elephant In The Gray Room, Pt 4 of 5: Major Structural Repairs.

Catastrophic Decisions

Some spur-of-the-moment decisions are just catastrophic, or (hopefully only) potentially so. The more serious the problem, the more justified you are in employing radical solutions to fix the problem; that’s the whole premise behind the “Elephant In The Gray Room” series.

When you become aware that you have just inflicted a potentially mortal would upon your campaign, the first step is to get it on life support. That might mean shutting it down for a few weeks, or it might mean inserting a number of standalone adventures of no campaign significance. If you’ve got any ideas, this is a great time to drop in some adventures that create plot seeds for a sequel campaign!

The next step is to perform triage on the campaign. That can be as drastic as completely throwing away the master plan that you’ve been using and completely revising the remainder of it, or it can be less severe – if you are lucky.

I like to develop, and have on standby, “cataclysm plans” that are designed to completely reinvent a campaign, as an outcome preferable to destroying it completely. A friend and fellow GM with whom I used to correspond sporadically about 20 years ago encountered a problem of this magnitude, dropping an arcane device into his game world that would have corrected the one vulnerability of his major villain and leaving that villain unstoppable. When he realized his situation, he asked me for help. As I recall, we had about six weeks before he had to pull the pin on whatever solution we came up with.

I told him to go and invent a new campaign with the same PCs and a variant on the current villain, and to presume that half-way through it, the PCs had all been exposed to some sort of mind leech that wiped their memories. I would create and send him the outline of an adventure which would transform campaign A into campaign B. In Act 1 of the resulting adventure, I destroyed the game world completely; in Act 2, I justified the survival of the PCs; in Act 3, they discovered who had destroyed the universe, how, and why; in Act 4, they were confronted with hobsons’ choice: re-knit the tangled skeins of reality together in a slightly different and unpredictable manner, or unleash the universe-killer on other worlds. Predictably, they would choose the first option. In Act 5, they would implement their choice while dodging the opposition – an entity they couldn’t hope to defeat directly, but who they could “undo” indirectly. The whole thing was rooted in a couple of throwaway lines in the campaign background that he had provided for the first campaign, which everyone (including him) had assumed was a mixture of poetic license and references to one particular being, the God Of Evil. I simply decided that they were all wrong, and that they really referred to a long-forgotten threat that was about to stage a big comeback.

And so it was that the characters in question found themselves in a new game world, with some superficial resemblance to the old one (same maps) with no idea of who anyone was, or who their characters now were, or what they had done in the past (but perfect memories of a past world that no longer existed). Campaign B resumed as though they had been playing it all along, with the players scrambling to work out who was who and what was going on. And thought it was all part of the plan from day 1 because of the tight integration with the old campaign’s background! Only when the campaign wrapped up did the GM reveal the true story to them.

That’s the scale of surgery that might be needed to navigate your way through the problem you’ve just inflicted on yourself. Or you might be able to get away with inserting a new adventure, or a new layer of plot, to solve the problem.

Catastrophic problems have catastrophic consequences and sometimes require catastrophic solutions.

This article is all about how you cope with, and control, the causes of randomness within an adventure, and the inevitable chaos that results. By far the most dangerous – by virtue of being unplanned and unplannable – are “brilliant ideas” and “moments of inspiration” on the part of the GM, especially because these are likely to be relevant to the situation at hand and implemented without sufficient thought and care, on the spur of the moment.

These moments of madness can be positive and beneficial to the campaign and the adventure, and when that’s the case, all is well. But each carries the inevitable risk of the other side of the coin – and the coin itself is weighted by virtue of haste. You are eleven times more likely to get these decisions wrong than to get them right.

The obvious solution is to decide that you aren’t going to implement any of these brain waves. That solution not only contradicts human nature, it risks making the campaign a sterile place that can’t react to unexpected choices by the players, or worse, that denies the players agency within the campaign. So buckle up and prepare to ride the whirlwind; chaos in your RPG is going to be with you for a while.

I received a very kind offer of free sample merchandise at www.paperlesspost.com to try out their products and processes and review them for Campaign Mastery. I had hoped to find things that I could re-task – a virtual parchment to use for maps and scrolls, and so on. Unfortunately, their products seem to be too mainstream for that sort of application (bet that’s the first time they’ve had that complaint!), geared more for creating custom virtual invitations and the like, so I declined the offer. But I wanted to say a public thank-you in this footnote, and put the service on readers’ radars in case it’s of more value to you.