This image combines “Magic Eight Ball” by MZMcBride courtesy of Wikipedia Commons, with Crystal-Ball-Photography by Alexandra ??A life without animals is not worth living?? from Pixabay

I’ve written a lot of articles about luck and a lot of articles about plot, but very few about how the two intersect. Time to change that.

A linear plot, like that depicted in Figure (1) Below, is very boring. Nothing the players say or do – and, more importantly, nothing the players have their characters say or do – changes the outcome in the slightest.

While there may be some initial interest because of a novel premise or an interesting story being offered, interest, and participation rates, decline rapidly once it becomes clear that there is no player agency within the plot.

(Figure 1)

Adding Complications and Challenges

Things improve quite a lot when you add complications and challenges. These are hurdles that have to be overcome by the PCs, one way or another, in order to reach the conclusion of the narrative.

(Figure 2)

But as soon as you do so, luck enters the picture. The PCs can say something brilliant, or make a spectacular die roll; or they can say something stupid, or roll spectacularly badly. In the simplest, most general constructions, this has no direct impact on the plot, but can have a vast difference on the context of the outcome – the indirect impacts. For example, taking the wrong path might turn someone who might have been an ally into someone who is a hireling who expects to be rewarded handsomely for every time he pulls the PC’s feet out of the fire. Or it might be as simple as the PCs accumulating damage and wear-and-tear, and chewing up consumables – in the old days, this was an expected phase of the game, and many reams of advice were written on how to ensure this consumption of resources without going too far and without killing the PCs in a nothing encounter.

Still more advice was written on how to prevent PCs gaining excessively from such encounters, and how to ensure that they didn’t derail the plot, and so on, and still more advice was written to players on how to recognize and take advantage of the opportunities that inevitably arose.

But players like having agency, they like to be able to see that they are having an impact on the world around them. They like to be the architects of change in their environment, in other words.

For a while, you can fake that by building it into the stories that your games are telling, but that starts to ring hollow. Giving the players real choice scares a lot of GMs; we tend to be control freaks. But sooner or later, we have no choice.

Branching Linear Plots

If we’re lucky, we stumble across a half-way house like the one depicted below in abstract form. Figure 3 depicts a plotline with three outcomes and two major tracks. There are challenges to be overcome along both tracks – possibly even the same challenges, but because players only get to see the track that they are on, they will never know.

(Figure 3)

The difference that this makes is simple but profound. A lot of the time it won’t be noticeable, but every now and then the consequences of a past action will be seen to have had a measurable impact on the situation the characters find themselves in, or on the options available to the PCs for overcoming those challenges.

The way to construct such a plot is to write a linear plotline, complete with challenges, but building in branch points where the players can decide to do something different from a limited palette of broad options. Once you have the main plot outlined in this way, you can insert the divergences as variations on the existing material. Rather than redundant passages, it’s generally better to simply describe the differences to each encounter/challenge and each resolution, preserving the central plotline while presenting various chapters with optional or alternative content.

The downside is that this adds a new vector of player agency – now, what they think, and what they think that their characters think,.also matters. And that makes for ever-greater uncertainty of encounter/challenge outcomes. Now, they are choosing the terms on which they face each challenge; sometimes, they will talk, sometimes they will bargain, sometimes they will act, and sometimes they will avoid action.

The effect on player investment and interest can be dramatic, as Figure 3 also shows – while there are still troughs and valleys, overall the interest levels tend to be sustained throughout.

From the GM’s perspective, sometimes they will take one of the one or two obvious pathways, sometimes they will make a choice that the GM was not expecting, forcing an improvised response. The GM is not totally in the dark when it comes to such responses – he has the more obvious choices as foundations, and the personalities of the NPCs, and – hopefully – their goals and ambitions, to serve as guides.

The problem is that the resolution of an encounter, challenge, or situation is usually linked to the rest of the adventure like one in a chain of dominoes. Back before PC choices mattered, the dominoes were fixed in position and could only fall in the right direction; now, the chain reactions are wild and unpredictable. Certainty gets replaced with Uncertainty.

The Rise Of Fuzziness

I’ve already mentioned how much GMs hate uncertainty – at least a touch of the control freak resides in all of us!

GMs learn to manage this unpredictability by micromanaging their adventures. When you first start the adventure, it seems overwhelming, and at best you can barely predict how the adventure will end.

(Figure 4)

As each branch point is achieved, more of the adventure becomes certain (because it has already been played), and there is that much less scope for deviation and unpredictability, so the rest of the adventure also becomes more certain, as Figure 4 shows.

Once you accept the premise that the end may be uncertain but will become less so as you go along, you start learning how to manage the situation.

More importantly, while you can’t steer the plot without violating player agency, you learn how to use stimuli and adventure content to nudge players in the right direction when the adventure threatens to become becalmed or to devolve into chaos.

Players respond to the increase in agency with greater interest and enthusiasm, provided that they appreciate the significance of their choices. The more trivial their choices seem, the more like a faux-agency the scope you are providing them seems (even though you may have other ideas about how significant their choices will be in the end).

What you end up with is illustrated in Figure 5, below.

(Figure 5)

The moment that something takes place that the GM could not possibly have predicted in advance (even if he did), and the adventure proves robust enough to continue and be affected by the choice at the same time, the players will become aware that they have been granted access to a whole new level of Player Agency. Their PCs lives become truly theirs to command, within reasonable limits, and they can make choices confident that you will not only make the lives of their PCs interesting (and fun to play), but keep the campaign moving forward.

This seems very desirable, to me.

Three approaches, plus one

There are three well-known approaches to this situation.

1. You can attempt to outline every possible outcome, at least in general terms.
2. You can have a vague idea about the overall plot, a clearer idea about the immediate situation, and simply improvise around player choices and rolls as they happen.
3. You can have a more solidly structured overall plot, with clearly defined branches, and simply improvise around unexpected variations as they occur.

Method one is the reason why computer RPGs sometimes seem like a choose-your-own-adventure book, and why embedding more flexibility in such media is so difficult – it’s a VAST amount of work compared to the basic utility of a linear plot.

(Figure 6)

Figure 6, above, gives you a reasonable impression of just why this is so. Even with some options conflated, and keeping the branches simple binary choices, five branch points yields 20 outcomes in the illustration – and it could easily have been worse. Thinking about the structure of an adventure in this way is what leads GMs to shy away from offering this level of choice to players.

It takes a lot of confidence to progress to option two, and the adventures are rarely as satisfying as the sweeping epics that become possible with greater structure and advance planning. Nevertheless, a lot of experienced GMs would hold this up as the pinnacle of the art – Johnn was always an advocate for this approach, for example. Even more would argue that it gives necessary experience and expertise, and that’s an even harder position to argue with.

That said, it’s always been my contention that option #3 is the best way to go. It’s the one that I most frequently use in my campaigns, and is one of the reasons why they last for decades. It’s the one that I’m currently using for the Zenith-3 superhero campaign, and the one – in more episodic form – that my co-GM and I use for the Adventurer’s Club pulp campaign. The Zener Gate (time agents) campaign, now approaching it’s final adventures, was always envisaged as an Option #2 campaign, explicitly and directly translating character experience and expertise into greater Player Agency. The Doctor Who campaign is a sort of half-way house between options #2 and #3 – the plots are robust and carefully planned, but much of the narrative and some of the adventure specifics are improvised. Anything can happen, so long as I get to the next plot point intact.

But there is a fourth option, one that I’ve been planning to write about for a long time. The initial drafts were an epic entitled “The Trouble With Disaster” that became hopelessly bogged down back in 2014 and was set aside for later redevelopment. I came back to it in 2017 and figured out how to fix the problems that had derailed it the first time around, only to strike new problems.

Okay, I know I’ll get asked about it sooner or later.

I originally structured “The Trouble With Disaster” as follows:

1
   A
   B
   C
   D
   E
   F
   G
   H
2
   A
   B
   C
   D
   E
   F
   G
   H
3
   A
   B
   C
   D
   E
   F
   G
   H
… and so on.

I got about 80% of the way through – 8500 words written – when I realized that I was having to copy-and-paste the same explanations into every A section, and another set into every B section, and so on, and that if I restructured the article as follows:

A
   1
   2
   3
   4
   5
   6
   7
   8
B
   1
   2
   3
   4
   5
   6
   7
   8
C
   1
   2
   3
   4
   5
   6
   7
   8
….and so on,

then I could eliminate more than 1/3 of the content as redundant, making the article a lot less tedious to write – and to read. But I kept getting sidetracked into the subject of today’s blog post, which was too big a topic to ignore – so, four years ago, it was again put on hold.

Okay, so where was I? Oh yes – the fourth option.

The fourth option is to understand luck a lot more solidly than most GMs do, and to use that understanding to prune that multitude of possibilities down into a core or Anchor plotline and a couple of critical branches, plus some structures that give the impression of full player agency while collapsing the possible choices down to a manageable number. In other words, to scale the degree of randomness to the criticality of the event to which the domino in question connects, while tossing in extra opportunities for uncertainty to manifest when it doesn’t matter to the bigger picture.

And another word for uncertainty, at least in this context, is randomness – or Luck, if you will.

No, I think the best approach is to describe the technique for creating an adventure with this structure, one that controls the impact of Luck. And THAT’S what this post is all about.

1. Overall Summary

I always start by generating an overall summary of the linear plot. This becomes my overall guideline; either an encounter fits into this guideline or it’s extraneous – though sometimes the whole reason for a particular adventure is one or more of those extraneous encounters, serving as a delivery vehicle for an NPC to be significant in the bigger picture later.

I compress and compact this as much as possible, putting any necessary explanation in footnotes rather than cluttering up the outline of the plot.

Example: PCs enter dungeon. Goblins (adventurers welcome, toll, escorts). Trap (warning signal). Kobolds torturing bugbear (Drow ‘advisor’). PCs learn of the Sapphire Star (1) while Drow learn of the PCs (2). Bugbear Ghost. Kobold War Patrol, bugbear counterattack, crossfire cliffhanger.

From this, it can be assumed that the PCs reached the Dungeon last game session, and (from the encounters specified) that they are somewhere around 5th level (party of 4). It succinctly frames the adventure action, gives reminders of the relevant backstory and the context of the encounters, offers at least one plot twist right from the outset, (Goblin Welcome) and hints at a physical stratification of multiple underdark societies – Goblins, Kobolds (enhanced), Bugbears (some enhanced, some not), and Drow – though there may be more intervening levels to this cosmopolitan underdark yet to be revealed.

2. Bullet-point breakdown

This starts out being a simple rearrangement of the overall summary, but each point is expanded until the structure is something close to complete.

Example:

become

As you can see, this breakdown is far more complete and specific. It doesn’t include final narrative or dialogue, but does give key narrative signposts and dialogue cues. Most importantly, it follows a logical sequence – PCs reach a location, get a general impression, get more specific information, interact.

It should also be observed that in addition to the welcoming tone of the Goblins (the already established plot twist), this outline adds a second (which justifies the first), the worship of the Goddess Of Peace (who seems remarkably martial in many ways, a blend that should appeal to any number of PCs). It also offers hints that the opposition in this dungeon might be a lot tougher than might be expected (using the Goblins as an exemplar) which is an initial glimpse of the bigger picture, and incorporates a couple of occasions when rolls will impart additional information.

There are still a few finishing touches and additional details required, for example the price list, descriptions of some of the NPCs. A list of mini-encounters in Thatch should also get made, some with basic triggers – if a PC interrupts a Goblin, he will get told “How Rude!” and feel a warning tingle as though a lightning strike were imminent, for example.

The name of the village is very important – “Thatch” gives the right tonal signals, enhancing the description of the settlement and adding to the general impression. Finally, the pollution at the ceiling is very important as a signal of plausibility, a notation that actions will have consequences.

3. The Anchor Plotline

Once those additional details are dropped in where appropriate, the Anchor Plotline is complete. This is very much the format that I generally use to run the Dr Who campaign.

I think an example of that would be fairly redundant, since I’ve already described the differences between that plot format and the example above, so let’s move on.

4. Minor Branches

A minor branch is one that doesn’t alter the overall plotline, as spelled out. Insignificant branches may be bypassed.

For example, the PCs may decide to head for the largest building that they can see from the cavern entrance, and stick to that despite the instructions on the sign. That bypasses the construction site and it’s additional information about the construction of the huts and replaces it with a new plot sequence at ‘the biggest building they could see’.

First up, I have to decide what this is – the communal bathing facilities, an industrial operation, or the temple of Lithis. Or, better yet, I could have the PCs make a spot check and describe that, and the central tower, in general terms, and let the PCs decide where they want to go. Any alternative will end with an NPC pointing and saying “The Governor’s Tower is that way. They look after visiting adventurers like yourselves. Be at peace!” (I like that last snippet of phrase – I might even make “Be at peace!” the general way of saying goodbye.

These branches all return to the main plotline. I’ll build in as many as I can think of, one at a time.

The general structure of all of them will be similar:

13 preceding plot point
14 plot point being bypassed
15 rejoin

becomes

13 preceding plot point

14a alternative plot point, which starts with (in big letters) IF THE PCs …. and the condition that diverges the path.

14b plot point being bypassed, now starts with IF NOT establishing that it only happens if the condition in 14a is not triggered.

15 rejoin

This is the most efficient structure. You can have any number of conditional plot points between 14a and the 15, so long as the last one is the one currently listed as 14b. Observe that this uses the numbering schema to reinforce the point – “there can be only one 14 event unless the PCs separate”. (If they do, this sort of structure makes dealing with it a breeze).

Sometimes a more complicated structure is needed because there are multiple plot points in a branch:

13 preceding plot point

14a alternative plot point, which starts with (in big letters) IF THE PCs …. and the condition that diverges the path. Ends with a GOTO 16.

14b second alternative, starts with IF THE PCs… as above.

14c plot point being bypassed, now starts with IF NOT and must always come last.

15 second plot point being bypassed.

16 rejoin

Nor do the instructions at the end of each branch have to point to the same rejoin point. Some options, for example, might not bypass 15, while one does:

13 preceding plot point

14a alternative plot point, which starts with (in big letters) IF THE PCs …. and the condition that diverges the path. Ends with a GOTO 16.

14b second alternative, starts with IF THE PCs… and ends with a GOTO 15.

14c plot point being bypassed, now starts with IF NOT and must always come last.

15 second plot point being bypassed by one PC choice, and the rejoin point of one branch.

16 second rejoin – the plot is now back together again.

This structure is lifted straight from BASIC computer programming, which is why some may find it familiar.

Sometimes, greater complexity still is needed – for example, plot point 15 might be slightly different as an outcome of the choice and the resulting plot events. This means that 14a will still bypass 14 and 15, perhaps replacing them with something else; 14b will now point to a variant 15, which I will usually label 15b so that I can trace the plot thread; and 14c will lead to the normal version of event 15:

13 preceding plot point

14a alternative plot point, which starts with (in big letters) IF THE PCs …. and the condition that diverges the path. Ends with a GOTO 16.

14b second alternative, starts with IF THE PCs… and ends with a GOTO 15b.

14c plot point being bypassed, now starts with IF NOT and must always come last.

15 second plot point being bypassed by one PC choice

15b variant version of plot point 15 and the rejoin point of one branch.

16 second rejoin – the plot is now back together again.

I’ve inserted a 15b (a ‘b’ because it’s coming from 14b), and changed 14b – that’s all that I have to do.

These are the equivalent of a Magician’s Force as described a few weeks ago. The GM doesn’t care which path the PCs take, because they all lead back to the main plotline eventually.

5. Skill/Ability Branches

The same thing can be and should be done with die rolls and any unusual abilities that aren’t ‘always on’, whenever they can be anticipated. This is also a good time to review the standard plotline to make sure that any abilities that ARE always on have been factored in; I like to throw in the occasional tidbit to signal the players that I have done so.

For example, there might be bat-like beings congregating at the top of the cavern, hidden by the foul air and darkness that a character with Infravision might be able to spot from the cavern entrance. Are these more servants of the Enemy Of Life preparing for an attack on the community? Or just beings who can breathe in the befouled air (unlike most PCs) and have entered into social symbiosis with the Goblins? Either way, they must have a route in and out that isn’t controlled by the Goblins – so this represents an entirely separate branch for the adventure. That’s beyond the scope of the techniques we’ve dealt with so far, which have been small and local.

6. Major Branches

The technique may be slightly different, but it is not that dissimilar in structure. The branches of events are simply longer, and change the circumstances of subsequent plot points.

There are two elements to such major plot branches: insertion of the plot branch, and flagging subsequent plot points that may play out differently as a result.

Here’s an insert for a major branch:

03
04 branch point – IF PC’s …. GOTO A05
05
06
…
11
12 GOTO 13
   A05
   A06
…
   A17
13 rejoin

Observe that there can be a different number of plot points in the branch, though – for reasons of timing – it’s generally better for them to be of similar length if you can manage it. The other point to make is that I’ve padded the low digits with a zero, anticipating plot points numbering less than 100. Based on the example given earlier, though, in which two became nineteen before minor branches were even selected, that might be an underestimate – but this is still usually a safe assumption.

When it comes to consequences, it may not even be necessary to have a branch, but if it is, you treat it in a similar way – either as a minor or major branch, as you deem necessary. For clarity, event though it derives from branch A, any major branch would usually be labeled B.

That means that you are limited to 26 major branches, which is usually vastly more than enough.

It should also be noted that major branches can have minor branches of their own!

7. Nested Branches

I’ve already noted one possible major branch, but another source of them is when things go horribly wrong. I always recommend reviewing an adventure to look for potential places it can derail, and constructing at least a rough plan for getting things back more-or-less on track.

What if, for example, the PCs decide to kill the goblins and loot whatever they can find?

Logic: a couple of defenders will arrive from nearby homes; most will attempt to delay the PCs while one or two go for alarm horns, which they will blow three times (once could be an accident, twice maybe not, three times removes all doubt). A group of militia will arrive next, and attack in coordinated fashion. They will be followed by the Priest Of Lithis, who will begin raining Divine Retribution upon the party. They may or may not get through the village at all, but at least a couple of them will probably get the chance to withdraw (under steady bow-fire). Survivors who escape and return under a flag of truce would be able to barter for the return of their dead members’ bodies (less any equipment they were wearing or carrying); anything more than a temporary truce will require retribution of 100gp + 50gp per Goblin killed.

This outlines a completely separate branch of the adventure, one that either ends it completely or leads to a complete reset (with added complications that the GM will have to insert into the adventure, probably on the fly).

This major branch is completely contained within the “Lurkers In The Mist” Branch, which is called a nested branch.

Ordinarily, if the PCs said something insulting or ungracious during the Governor’s welcome, that could create another major branch, but under the circumstances here it could probably be handled as a minor branch – but the toll and other prices probably go up accordingly, and Big Burly Goblin probably adds “I’m sure I’ll be seeing you again,” to the PCs. There’s no need to write a lot of this stuff down; what you already have is sufficient guide to permit adaption as you go by most competent GMs.

What you can end up with is something that – graphed symbolically – looks something like this (and observe the major branches spawning new major branches of their own):

(Figure 7)

8. The Luck Factor

Almost everything I’ve written about so far has been in terms of players making choices and the GM making allowances for alternative choices.

Let me rephrase that: Almost everything I’ve written about so far has been in terms of a factor outside the GM’s control changing the flow of the adventure, and the GM making preparations for alternatives.

So far as the GM is concerned, there is no difference between a choice by the Player and a die roll by the player – there will be an expected outcome, and any number of less probable outcomes, the most significant of which should also be outlined. In general, for most die rolls, that means a set of four outcomes: Success, Failure, Major/Critical Success, and Major/Critical Failure. Sometimes, it may be desirable to insert a pair of intermediate values – a bare success, or close failure – but that’s rarely necessary except in diplomatic situations where nuances of subtlety might need to be explicitly addressed – making a verbal faux pas, but being able to salvage the situation.

Before you can really do this properly, you need to understand luck thoroughly, as it manifests in your games, and as it manifests in real life, and how the two relate.

Let’s use d10s to simulate a simple poker machine for a moment in order to illustrate this point. The goal is to get three numbers the same (the higher the better), or any three successive numbers, or one or two tens.

First reel: roll a 7. Is that enough to call someone lucky? Probably not.

Second reel: roll another 7. is that enough to call someone lucky? Not yet.

Third reel: There’s a 1-in-10 chance of a 7. There’s a 1-in-10 chance of a 10. Nothing else will pay off, and the 10 will only pay a pittance. The roll is a 3 – it was far more likely to be a nothing result like this. The character definitely isn’t lucky.

What if it had beaten the odds and come up a third 7? Is the player lucky yet? That’s a more difficult question. They obviously have achieved a 10-in-1000 success, or 1%, of getting a triple; it is not the best possible triple (that would be triple 10) but it’s in the top 1/2, which would only happen in half of that 1% of cases. The payout would be substantial. You would have to say this character was moderately lucky.

What if this was their first pull of the lever, and the only pull they could afford to play? Suddenly, their luck quotient goes sky-high, and never mind that there were a few better payouts available. Alternatively, if they have fed 200 coins into the slot without winning a big prize, the payout odds would almost certainly leave this as a net loss – so, suddenly the player isn’t lucky at all.

Where do you draw the line when it’s on shifting sand?

Some people find the almost-instinctive understanding of luck that a gambler develops to be as attractive as the thrill of success itself. These are the people who look at the payout combinations and a small sample of rolls and try to work out what they should realistically expect to get out of the exercise they continue to play, and all goes well. My uncle used to be like that – every week, he would take a fixed amount to the local bowling club and feed it into one or another of the poker machines. Overall, he broke better than even, according to his own accounts, and every now and then he won something substantial – but he never went over his self-imposed limit, and it was never about playing to win for him; “that,” he told me once, “is when gambling can become a problem.” On another occasion, he commented that most people forget the pulls that don’t pay out; they can’t pull that lever fast enough. He, on the other hand, paid attention to them, and remembered them, however vaguely; he could glance at the current total credit, subtract any winnings, and tell you almost instantly how many failures there had been. When he got to the end of his allotted and budgeted expenditure, he collected whatever winnings were showing on the machine and went home. I doubt that he’d had these thoughts when he first started; they were opinions that had formed and firmed over time.

If you still enjoy the luck aspect of these games, you can also play them at Casimba, which offers decent odds to all online games you choose there.

All this is directly relevant to RPGs. If there’s a branching path that only happens on a 10 on d10, you might or might not include it in an adventure. If it happens on 1-5 on d20, that’s a whole different story. If this in turn has a branch that only happens on a 1-5 on a second d20 roll, should you outline it in an adventure?

Some people will look at the overall combination – 25% of 25%, or 6.25%, and note that this is less than the chance of a 10 on d10, and therefore say no. Others would point out that what a d20 rolls on one occasion has no bearing on what it will roll on a subsequent occasion – so the correct assessment is that there is a 25% chance of needing that branch IF the first alternate branch comes into play, and that argues a ‘yes’.

Those with some understanding of probability from their school days will know that the second is the more correct interpretation, but I think that in practical terms, the 6.25% has to be acknowledged, too. So I might prepare an outline of something for the second 1-5 outcome, while not giving it as much development as the more probable branches. This, to my mind, respects both results.

But there’s a counter-argument: Most PCs are competent characters, which is to say they generally have positive modifiers to any roll that they can influence. A +1 on a d20 alters the odds by 5%, a +2 by 10%, a +3 by 15%, a +4 by 20%. If it’s an ability check, that 25% of 25% might actually be a 5% or 5% chance, which only happens 0.25% of the time. That’s only worth a line – at most – and if the consequences of a critical failure of this type were reasonably obvious from the context and surrounding circumstances, I would probably even forego that line. The only justification for it, in my mind, would be to put limits on how harshly the failure was treated, and outline a recovery path.

Complicating matters still further is a third factor: the scale of catastrophe, and the rewarding of luck. Players expect to be rewarded for a good roll, and for things to go badly wrong on a bad roll. “Harsh but Fair” used to be a description sought by every GM, and even now, it’s a maxim many live by. The GM has to deliver, or the players feel cheated somehow; and it’s then up to the GM to fold the stroke of brilliance or sudden ineptitude of the PC back into the plotline and keep it on track, in other words, to deal with the fallout.

Getting into the habit of contemplating, how briefly, the shape of the measures needed for that coping, at the time you anticipate or call for a die roll to be made, can be a GM’s life-saver.

The benefits can be psychological as well as overt and obvious – knowing that you have prepped for everything that you can reasonably think of enables you to relax and play other unanticipated situations on their merits, and in general function with greater confidence.

Have you ever watched a TV drama and thought “he (or she) looks totally confident” or “He (or she) doesn’t look very sure of him/her self”? Players can tell when you are and aren’t confident. You’re generally too busy being one or the other to notice the telltale signs that you’re giving off, but they are there.

If the impact of a bad roll, or a good roll, is of sufficient magnitude that it can demand emergency measures to rescue the adventure, you should at least put some preliminary thoughts into the adventure about how to cope with those turns of events, in exactly the same ways that have already been described. These might not be a very probable branch – but they need to be at least outlined roughly (to the Figure 1 standard, a paragraph), just in case.

Understanding luck is just as important as understanding storytelling when it comes to crafting a good adventure. You need to fill your writing with both. And character, and tone, and style – but those are actually secondary priorities. Driving your structure with luck and story are essential.