I can’t remember the last adventure I ran that actually moved in a linear fashion. Often, I find that if I am running a pre-gen, or have built an adventure with a definite plan of execution, it ends up one of two ways. The players swing the story around like a rat flail, mangling the world until the story fits their play style (while I weep in despair over my beautiful creation), or I reign them in and keep them straight on the planned course only to find them not enjoying the game as much.

The Cover of The Book, The Hobbit
“Then the prophecies of the old songs have turned out to be true, after a fashion!” said Bilbo.

“Of course!” said Gandalf. “And why should not they prove true? Surely you don’t disbelieve the prophecies, because you had a hand in bringing them about yourself? You don’t really suppose, do you, that all your adventures and escapes were managed by mere luck, just for your sole benefit? You are a very fine person, Mr. Baggins, and I am very fond of you; but you are only quite a little fellow in a wide world after all!”
– from The Hobbit, by J.R.R. Tolkien

Needless to say I gave up trying to run linear games a loooong loooong time ago. Mostly out of self-preservation. I adopted a new philosophy towards adventure design: Island Design. Island design style isn’t new. It doesn’t involve sitting in a Hawaiian shirt with a margarita and throwing seashell darts at a corkboard with notes (Well sometimes it does). What island design style is, is a way to rethink what the important parts of the adventure are and how the players get to them.

Island Design Style

  • Take all the things that you feel are important to the story. These should be things such as the Big Bad Evil Guy/Gal, Important NPCs, Important Story Arcs, Important Story Elements, Items, Rewards, Character Goals, etc.
  • Keep them all in your head, but arranged as little islands floating in the water without any direct connections to each other.
  • Arrange the islands in a loose order via proximity and where they should come in the story. Islands that the PCs should get to first, keep up front. End game islands should float around in the back.
  • As the PCs progress through the story let them find their own paths between the islands or move islands into and out of their path by modifying elements of the game being played.

Linear plots are fine, but flexible ones can be more realistic and more fun to play. Island Design keeps your plots loose and malleable, making them easy to adjust.

One plot element may be the hidden fortress where the BBEG resides. The PCs might discover the location in some way that you hadn’t thought of. They might bribe an NPC who knows the location, or sneak in with the laundry delivery. This might get them there earlier than you had intended, but that doesn’t mean you have to bring them back on course. If you don’t want them to fight the BBEG yet, then move his island farther away and make up a new island or bring something more appropriate in. Maybe a minion NPC is there instead of the BBEG. Maybe they find the fortress unoccupied but without the treasures it would normally have because the troops were away mounting a war someplace else. These elements can be turned into their own islands and moved back in at a better time.

If a story goes in the wrong direction, it doesn’t mean everything about it needs to go. The elements of it can be re-skinned for later use. If the PCs get the artifact that was needed to awaken the dead god, maybe the BBEG doesn’t spend time trying to get it back but researching a new spell, or killing 100 people from a certain area in revenge. The giant cave beast the PCs would have encountered as they tried to infiltrate the BBEG’s lair could now be sent out with a squad of thugs and trainers to help assault the king’s city.


  • When the players find an interesting solution and you decide to go with it, they feel like they’ve just “beaten” the game. Their character’s actions had an impact and their ideas came into play. The world gets fleshed out a bit because they feel they circumvented an obstacle instead of following a predefined course.
  • If something goes awry, or is unplanned for, it is easier to change the plot around to accommodate.
  • The Game Master can leave some of the plot in the hands of the players and let them do things that are more interesting to them.
  • Even though the plot progression hasn’t been laid out beforehand, it feels more linear to the players since they never, or rarely, had to backtrack and try something again.
  • In the end the Game Master has more control over the general plot progression since he or she can change things on the fly.


  • Players who like order and following the Game Master’s lead might find it hard to adjust.
  • Depending on play style, Game Master’s can feel like there is little structure to the story since it is being built as it is played.
  • Playing in a more rigid system can have some issues when it comes time to adjust statted properties.
  • More time might be taken to research and prepare new things at the gaming table.
  • Putting things under definite categories becomes harder to do. A document or file for NPCs will be shredded once you start moving them around in different places in the story.

Some Things To Keep In Mind

  • Pick out the game elements that are malleable and can be incorporated in different places. Make these your islands.
  • Making connections between islands becomes important. If you bring in a new island that the story so far hasn’t had any reason to lead to, you’ll have to throw some things in to lead the PCs there. You can also use unused elements that are already in the story so far, as long as you remember they won’t be available until later.
  • Remember, the players don’t see the planning and changes you make to the islands. To them it looks like one big semi-coherent story. Especially at the end. They will, however, realize that they had more to do with actually making it happen.
  • Keeping notes for the islands in a way that they can be shuffled can be a great benefit. Writing things down on index cards or in definitely separated sections of a document can be very helpful.

So, what do you think of my Island Design theory? Any thoughts, critiques, suggestions or questions? Did you notice I used the term Big Bad Evil Guy a lot? Ever use a similar way to lay out an adventure?

30 replies
  1. Simulant
    Simulant says:

    This is pretty much what I try to do with my group. I organize the story in scenes that I set up for them. Since I don’t desing my adventures in one piece, I can make up the appropriate scenes from session to session and recycle the ones i didn’t use the next week.
    The Islands approach takes my concept to a level that allows me to design the whole adventure/ campaign beforehand, which leads to a much better designed story since it facilitates the application of an overall background.
    I should defintively give it a try, because right now I’m finding it hard to keep track of my own stories.

  2. bif
    bif says:

    Pinnacle Entertainment Group (guys who made Deadlands) have been publishing settings for a few years with this concept firmly in mind. They call them Plot Point settings, where the major islands, i.e., plot points, are fleshed out in the books, but there is plenty of flexibility in how and when these set-piece adventure hooks are incorporated into the story. My favorite of these, 50 Fathoms, is a sea-faring, swashbuckling, drowned fantasy world in which the players are often literally sailing between islands. I suppose it lends itself well to the concept. Some of the best games I ever ran came from plot point settings.

  3. wampuscat43
    wampuscat43 says:

    The problem I’ve run into occurs when I deftly handle a derailment, only to find, six weeks later, that I’ve forgotten the exact details (“You’re getting HOW MUCH per week out of that black market you took over???”). It is fun to maneuver around and tweak your plot as you go, but you need good bookkeeping skills as well. It’s rare to find that perfect left-right brain combination.

    Great post. Thanks.

  4. John Arcadian
    John Arcadian says:

    @SIMULANT: I was pretty sure this kind of idea was something a lot of people employed, but as I was sitting down to think about how I actually arranged my adventure outlines this is what kind of stuck in my mind. I’ve noticed a definite shifting from more linear styles of stories to more improvisational GMing. Not as the years progress in the hobby, but as an individual group of players get older. (P.S. Nice red dwarf reference with your name.)

    @BIF: I’ve had minimal dealings with the Deadlands setting (mostly hearing about it), but always found the ideas cool. I prefer a setting design style like that, as opposed to say the way DND or Whitewolf does it. I’ve always felt hedged in when I can’t make something unique to my players. Conversely, I’ve also found that my players love to read something in a book and then merge themselves into that. There is almost a feeling of overcoming the setting if they’ve done something awesome (like take over a famous thieves guild, or defeat an enemy that is pinnacle to the world, etc.) It is one thing to take over any thieve’s guild, another to take over THE thieve’s guild.

    @WAMPUSCAT43: That is definitely an issue that comes up. The more you improv, the harder it is to rely on notes written before the improv. Usually I’ll take notes on my laptop as to what happens. Usually they are just bullet points of significant things that I can refer to later. Of course, the other beauty (and hazard) with being loose with plot arrangement is that even if something isn’t sticking to canon at a later point, the players are used to it.

  5. BryanB
    BryanB says:

    I do something similar. I refer to it as matrix design rather than linear design. I build the vehicle but the PC’s have the keys to the ignition. That vehicle is going to go wherever they want it to go.

    If the PCs do nothing, then I will have the setting continue on without them, until something catches their interest, Having proactive players is certainly beneficial to this type of game approach. Having proactive players with a solid PC background (even a short bio) is pure RPG gold.

  6. deadlytoque
    deadlytoque says:

    This is more-or-less what I do with my games.. however, when I saw your title, I almost had a heart-attack.

    The WORST games I have ever participated on were all set on ships of various types, where the GM (always the same guy) had decided that the best way to make us stick to his plot and not try and deviate from his VERY STRICT linear progression was to drop us on a boat (or airship, or space cruiser) and make us follow the orders of some NPC — more a GMPC, really (who would inevitably act as the ONLY conduit between us and the other NPCs, AND a convenient deus ex machina).

    Once a story was over, we’d get back onto the ship, and we would be dragged to the next story location, where would we jump through the hoops he’d laid out for us, and go home. The best game session we ever had was when we basically mutinied, but when he started to get pouty over the potential death of his GMPC, we backed down, and I quit the game.

  7. John Arcadian
    John Arcadian says:

    @BRYANB:: Why Matrix design? If it has anything to do with Blue and Red pills . . . I’ll take the red pill.

    @DEADLYTOQUE:: Wow. That is kind of funny. Sorry to elicit a near heart attack! I’ve actually seen that tactic used in a few games myself, but with people under the lead of an army, or commanded by a king and always having to report back in at court, etc.

    Now that you mention that I’m tempted to change the title of the article to just Island Design Theory.

  8. BryanB
    BryanB says:

    @John Arcadian
    How to explain Matrix design… hmm

    Let us compare it to linear play.

    Linear goes like this:

    Scene A, B, C, D, E, F, G. The players might skip a scene but the events will tend to follow A (the carrot scene), B thru F, and then end with the BBEG in scene G. They will happen in that order with hardly any variation.

    Matrix can go like this:

    Scene A (the carrot) goes as planned. Scenes B thru G are thought out as what might happen, but they aren’t neccessarily tied down to the linear pattern. You might get scene E or D coming before B, depending on the character’s decisions.

    Matrix is more like Sandbox style, on which we had a Gnomestew topic a little while ago. There are NPCs with motivations of their own. There are things that are planned to take place. The PCs come in and change things, alter plans, and create conflict or cooperation with the NPCs. Sometimes what happens is entirely based on NPC reaction to what the PCs did. Sometimes the PC’s go after an NPC based on what they think is the best way to stop his goals. You might even end up at scene G in the end, or NOT.

    If you think of what I term as scenes like what you term as islands, then we are on a similar train of thought. 😀

  9. zencorners
    zencorners says:

    One of the greatest methods for facilitating this sort of planning are post-its. I’ve been using this method of late to plan my 4e game. I generally put one concise red herring, plot point or Big Bad Evil Guy/Gal (BBEG) on a post-it and sort similar themes in the “pre-planning”. Each post-it is usually fleshed out prior to the game into a paragraph and maybe 2 to 4 encounters.

    This “method” has worked wonders for me in my latest 4e game. It’s great for letting the players decide on the course of their adventures, as well as allowing me the freedom to incorporate elements that I like into the game.

    Getting the players back onto the plot, in essence takes giving them “difficult” decisions. Recently my players were faced with splitting up the party. On top of this they were forced to defend a group of slaves, and then decide whether or not to ambush the slavers while only being at half strength. This tough decision was made by them to see out one plot point versus another. But it was their decision. This is one of the highlights of my games, where as the GM I see the players deciding on the course of the adventure willingly, not being “led by the nose”, so to speak.

  10. Scott Martin
    Scott Martin says:

    I like this presentation. It often feels like we are describing close points on a continuum– matrix, sandbox, island, flowchart, scene based, and “don’t over prep” all lead similar places, but they’re not quite identical. I wonder if there’s a good way to isolate the elements that lead from one state to another… and if that would help people transition in a direction they like, or gain the most advantages while avoiding the biggest disadvantages. Of course, that’s leading into deep theory land…

    I like the presentation of Island Prep above, particularly the pro/con breakdown. Could you explain drawback 4 a little more: More time might be taken to research and prepare new things at the gaming table.? I think I understand this drawback, but not why it ties to the Island style.

  11. Martin Ralya
    Martin Ralya says:

    Great article, John — this concept really resonates with me. The term “island” makes perfect sense, and I knew exactly where you were headed after reading only the title. I could benefit a lot as a GM by running my own games this way.

    My group just finished up a two-year D&D 3.x campaign (in which I was a player), and looking back on it I believe our GM took a very similar approach, especially once we passed 10th level and had a ton of power at our disposal. It worked very well, and it was perfectly suited to the kind of game he was running.

  12. John Arcadian
    John Arcadian says:

    @BRYANB: I get it. Yeah it does sound very similiar. The place where I see terming this as a Matrix that might work better for some games is a perceived sense of structure around the movements of the scenes. If the players catch on that many things change based on their choices, then they might feel like there isn’t any structure or forethought. Terming it as a matrix, it seems more like things are getting moved around instead of morphing themselves.

    @Scott Martin:
    Yeah. Exactly! I think this isn’t new ground by any means, just a different way of terming the “Improv Gming” movement that seems to be gaining ground. Isolating the elements. Wow, that does get more into deep theory territory. If you go with a matrix and scenes like BryanB does, then I guess the connecting elements would be the lines or the grid. If we use Island Theory as the terminology, then I guess we could say that they are the currents or the prevailing winds, but ones that can be ignored or utilized by the way the players turn their sails.

    “More time might be taken to research and prepare new things at the gaming table.” My thought behind this was that you might have to spend some more time looking things up while sitting at the table waiting to game. If you pull the BBEG’s island away from the island that the PCs are getting to and you want to fill it with something else, you might have to spend some time prepping that something else. What are its stats, items, powers, etc. The PCs would logically find some treasure here. Ok, time to roll on the treasure generation tables, etc.Depending on the system it might not take much, but it will take some time away from the game.

    @Martin Rayla: Thanks! There is something about the idea of islands surrounded by water that makes it an easy image, for me at least, to grasp. High levels make it hard to run very linear games. That or whitewolf. Once you have a thieves guild, corporation or other similarly huge entity at your disposal, it gets hard to fit the PCs into narrowly focused categories. High level powers also make it hard to do linear games. They always provide unique solutions that fall just outside of the ways we plan for them in linear games. I’ve got a F.A.P. post about high level powers that is half finished.

  13. dmmagic
    dmmagic says:

    As you said, I found myself adopting this method simply to survive. I remember one campaign where I had developed countries, intrigues, political structures, armies, commanders, etc… the PCs had found a magic sword that was so powerful that countries were fighting over it.

    As the story progressed, though, they decided… well, I don’t know what the thought process was, but the Chaotic Neutral PCs decided to trek to the other side of the continent, steal a ship, and become pirates. Story dead.

    I’ll be trying another more-linear campaign this summer, though, and hopefully the players will agree in advance (as they have in recent years) to at least try and not do completely random things.

  14. John Arcadian
    John Arcadian says:

    @dmmagic – Did you let them play on as pirates? I can see many of the same countries, intrigues, political structures, armies, commanders, etc. being used with some changes. Countries could still be fighting over the sword. Armies could be sent after the PCs or waiting at ports (every ship has to land eventually), etc. Mind you, they would have to telephone in their attacks on the PCs, but its do-able.

    One tactic that I’ve used when PCs abandon the story, but I still want to cling to it, is that the story can chase the PCs. Just because they stopped fighting the BBEG, it doesn’t mean that some innocuous piece of treasure they looted can’t become suddenly important. Now they need to be tracked down. Even if they get rid of it, it could become a revenge thing.

  15. DocRyder
    DocRyder says:

    I’ve seen something similar called “the Onion Diagram”.

    Like Shrek the Ogre, villains have layers, called flunkies, henchmen and minions. 🙂 This people are connected to one another, and eventually, you can find your way to the Big Bad(tm) by following the right connections.

    The way you track these connections is a chart with a small circle in the middle, labelled “Big Bad”. Then you have concentric circles on the chart, around the middle one, with each of these onion-like layers divided into (at least) two sections (but four works well) more than the closer one in. Put the flunkies, henchmen and minions together in such a way that such folks on the same ring are connected in logical fashion.

    Plot elements can be placed into the onion as well.

    TSR also had a chart that was an easy to build chase and plot flowchart in a DM’s supplement years ago that I’ve used off and on over the years.

    I saw this and thought it might be something I’d stumbled on this morning in regards to game mapping: Why do so many fantasy games use maps that look like bearskin rugs? 🙂 I think I have an answer, but that’s one for another article.

  16. ulmus
    ulmus says:

    I find myself more and more often using a special FX design to adventure. I design in two tracks in parallel. The politicas, intrigue and why and wherefor make out one track. The other track is the special FX crew. How do I get good scenes, how do I set the mood, get interesting fights or cool enemies. The two tracks are run intertwiningly and usually add to eachother. I realize this might be a bit incoherent, but I think I’ll do a post on Special FX adventure design on my blog.

    Go get a look at my blog by the way, polyhedral.wordpress.com, it’s all fresh and new!


  17. kenmarable
    kenmarable says:

    @BryanB – That explanation sounds a lot like the old Planescape module “Tales from the Infinite Staircase”. I really enjoyed that set up. It was 7 (I believe) adventures where the first one was fixed as the intro to the storyline, and the last was fixed as the resolution, but the 5 middle ones could be played in any order. They even listed what the main NPCs were doing through the whole time so that if you played, say adventure 4 near the end, NPC X will show up to do whatever, or if you play it earlier NPC Y will be there doing something else instead.

  18. Dunx
    Dunx says:

    This idea might have been around for a while, but it’s new to me. I suppose this is what I get for dropping out of gaming for a decade.

    I’ve been running a Call of Cthulhu campaign and order of events problems have been some of the larger issues I have been dealing with. This is going to come in very handy when the campaign restarts later in the year.

  19. Aaronichi
    Aaronichi says:

    Hey, all, first post here. I am a new GM, I just ended my second session of my campaign. I have a rather, shall we say “creative” table. They always try and think of the least obvious solutions to the most obvious puzzles. They also have this “This is where I’m being directed? Screw that! There are monsters there, lets go drink beer and get whores!” attitude. The first 2 sessions, I created a crafted, linear story just to get used to being a DM, and to introduce them to my world. In keeping with their MO, they furiously tried to break everything I set in front of them. Arg, it was driving me nuts!

    I was using OneNote to organize my plot points, but I found I was still running around looking for rules that I should have thought of, and just plain coming up with bad railroading techniques to keep the players in line. They were not happy campers.

    Just this week, I started planning my sessions using Masterplan, and it has the same Plot Point style of organization mentioned above. It has some if/than type connectors, but it’s mostly a set of individual encounters. It also tracks rules for different plot elements and all that good stuff. Mix this flow with the group that likes to drive the plot, a few random encounters and skill challenges, and I think I could make my table an interesting place to be! Mind you, I’m not endorsing, MP specificly by any means, it’s just the first one I’ve tried and it fits what I need. I do think these tools are great for this style of story telling though.


  20. Eleanor Hingley
    Eleanor Hingley says:

    I love this as an idea! I’ve been mainly running live-action theatre-style games for the last six years and now I’m getting back into running tabletop, I’m finding that the amount of pre-planning I tend to do is making my games railroady, but I also find the necessary high improvisation levels of live-game running very uncomfortable. This is a great medium between the two! It allows me to put the time in planning so I feel prepped, but gets me out of the mindset that things have to occur in a linear fashion and I have to plot every logical leap to take the players from a to b.

Trackbacks & Pingbacks

  1. […] avoid railroading when planning several sessions, you may want to consider Island Design Theory. This allows you to plan ahead (good for your own sanity), while still allowing the sessions to […]

  2. […] brought this concept up before in a previous article, Island Design Theory, and it applies here in a big way. If you arrange all your game elements as little islands that can […]

  3. […] mean I don’t realize the necessity of good adventure prep. To that end, I try to prepare generic enough elements that I can throw in at any time. I don’t always cover everything I should. I usually end up making up NPC stats on the fly and […]

  4. […] relationship map above, you could keep all of your campaign info at the ready and rearrange it ala Island Design Theory. You could store all of your reference images for use in the game. You could create a puzzle with a […]

  5. […] Many experienced authors eventually share their craft with others by writing a book or books on writing. (We’re not focusing on grammar here, but on character and plot and story arc.) Two books on writing I can recommend to you are by Orson Scott Card and by Stephen King; there are many, many others. The Gnome Stew blog (see sidebar) has a set of 30 excellent posts on gamemastering; several posts in the list and others on the blog apply techniques from authoring to gamemastering. […]

  6. […] Island Design Theory discusses a way to keep the plotline of your story fluid enough to adapt easily to what the players do, without losing the thread of your plans. […]

  7. […] is in Island Design Theory, to borrow the name from the good people at Gnome Stew. When I read the Gnome Stew article on Island Design Theory back in January, I thought to myself I didn’t know it had a name. Put briefly, Island Design […]

Comments are closed.