Set Pieces

Set piece scenes are often the most memorable scenes in any session. When planning an adventure, they deserve special attention.

Plan to Improvise
You don't want to have everything planned out in detail, but you do want to have enough planned so you are ready for most likely actions the Heroes will take, and have enough interesting twists to make the encounter fun. You want to plan out the goals and tactics of the NPCs and most importantly, provide lots of stunt-able features in the environment.

Here is a planning template I use:

Connection to PCs
Of course, the scene must be connected to the Heroes somehow. Normally this is fairly obvious, but sometimes, you might want to make it clear. Ask yourself, "What is the purpose of this scene in the adventure?" or "Why are the Heroes going to participate in this scene?" Ideally, the scene should be about something more than just killing the opposition and taking their stuff.

Once I understand the reason for the scene, I try to pick an exotic or iconic location. The setting of Heirs to the Lost World Is rich with great locations for set-pieces. Just on the swashbuckler side of things gives you:
Ship boarding action
Classic tavern brawl
Court room
Dungeon or Inquisition torture chamber
Smuggler's bayou
Governor's mansion
Auto-de-fe ceremony
Horse and/or carriage chase
Completely of partially wrecked ship
Of course, for extra coolness, mix in a little Voodoo, Aztec, Maya, or American Indian elements.

Major NPCs
Next, plan out your major NPCs in the scene. For each, be sure to write down a motivation and a short, easily distinguishable hook to aid your narration. In this step, do not put the NPC's goals in the scene, but rather their overarching motivations. This will help you decide their actions when things don't work out as planned. I normally start with the following categories for my major NPCs:
other (perhaps a repeat of one of the above NPC types or something different altogether)
Of course, do not feel confined by these categories or your set pieces will start to look too similar.

Before starting the set piece, make sure the goals of everyone are clear. This includes the PCs and the NPCs. If unsure, consult the character's motivations when developing the set piece.

NPC Actions and Tactics
Using the NPCs motivations as a guide, think about their goals in the scene. How will they attempt to accomplish those goals? You may want to think about a couple "if, then" statements to react to likely actions from the PCs. Also, be sure to think about when the NPCs will surrender.


Stunt-able Features: "Action Scenery"
If you want the players to use the environment in clever ways, you must pre-plan potential stunt-worthy items and call them out when describing the scene. If using a battlemap, draw these items on the map or use tokens to represent them. Classic items include, ropes, chandeliers, wagons, awnings, etc. but anything will do. Try to have potential ways to jump, climb, swing, slide, throw, drop, etc. For example, a for a scene taking place in an Inquisition torture chamber, be sure to describe all the torture implements and mark them on the map. You know they will be used in some great stunts. Another technique you can use is to place items with unspecified contents, allowing the players to spend Destiny points to decide. For example, you could draw some crates and barrels on the map while hinting or reminding how the Destiny point rules work.

When thinking of the terrain for the set piece, variety is the key. Try to avoid the "10 foot room". Multiple levels are always great (balconies, stairs, etc). Try to have terrain that allows for the use of non-combat skills like acrobatics, athletics, and swim. Also, add something to change things up by including an environment complication. This could be a change in the weather or some other natural change that occurs in the middle of the scene. Alternately it could be a hazard somewhere nearby or other feature that provides tactical options. This could be a choke point, something that slows or speeds up movement, a potential danger like a fire or pit, etc.

Minor NPCs
Most set pieces need additional NPCs. A good rule of thumb is to have at least one intriguing henchman and a couple types of mooks. To help drive the action, it is best if at least one NPC has a ranged attack. As before, do not feel limite by these categories.
Ranged Mook


Nothing builds tension like a burning fuse. Not all set pieces need them, but a time-sensitive complication can be great. It does not need to be a scene-ender (like a fuse leading to a bomb that kills everyone). That would not be much fun. Rather, have a "clicking time-bomb" that will release a complication on the scene if the Heroes do not deal with it. For example, they see a rope that is starting to fray and will break soon, dropping _____, or they notice a small fire starting to spread. Alternately, you can just have a planned twist mid-scene. This could be an environmental change mentioned above, or something completely different. For example, maybe the ship they are on starts to sink or maybe the militia arrive.

Moral Complications
While perhaps not appropriate in every set piece, moral complications can make the scene more than a simple battle. The two classic examples are the innocent bystander and the possessed friend acting as enemy. To think up moral complications, try to tie them to the PCs' personal motivations or complications. Sometimes, they may be played more for the humor value (ex: an alcoholic PC witnessing a rum barrel slowly leak) or can be much more serious (ex: chained up slaves trapped on a sinking ship). While it can be difficult to plan, it is also fun to come up with conflicting moral complications between individual PCs or for an individual PC, forcing difficult choices.

Think of potential ways the scene could end and come up with a short plan in each case. I have found that most scenes only have a handful of potential outcomes. By planning for them all, I prevent myself from subconsciously railroading my players to a particular one. You don't need huge amounts of planning, just jot down a few ideas for each possibility. In addition, plan out any rewards the PCs will get at the end of the scene. This could be treasure, information, contacts, etc. As always, be sure the adventure can still proceed somehow regardless of the outcome.

Other resources:

http://www.obsidianserpent.com/downloads.html - Downloads page on ObsidianSerpent.com, check out this Set Piece template.
http://www.rpg.net/news+reviews/columns/action07jul03.html - The first of a series on action scenes by Dan Bayn
http://arsludi.lamemage.com/index.php/48/anatomy-of-an-action-scene/ - Anatomy of an Action Scene
https://sites.google.com/site/amagigames/action-scenery - Action Scenery


One of my favorite parts of Heirs to the Lost World are the stunt rules. Here is a very brief summary of those rules.

To attempt a stunt, the player spends an Effort die (a visual cue that he is attempting a stunt), describes his cool action, then makes a skill roll. The Game Director uses a "keep going, roll with it" hand gesture if more description is required. The Game Director also sets the difficulty of the roll, based more on the effect of the action rather than the "color" in the narration. If he triggers his Mojo in a successful roll, then he earns a Destiny point. If a player fails the stunt attempt, he can turn his failure into a fumble to earn a Destiny point. As in all fumbles, the player narrates the result himself. Stunts are the main way players earn Destiny points, a key resource in play.

To develop these stunt rules for Heirs, I tried out three awesome games to test their stunt mechanics:
Feng Shui (The difficulty of a stunt is only determined by its effect, not its "color" or "trappings".)
Exalted (PCs gain a bonus to their rolls based on narrating cool stunts.)
Wushu (PCs gain a bonus for every cool detail the player adds to the narration of an action.)

All three of these games were influences on Heirs, but after playing each, I knew I wanted something a little different, while keeping some of their great ideas. Here is what makes Heirs stunts different:

-Heirs to the Lost World stunts give immediate (during play) rewards to encourage wild action with fun player narration. Experience points as a reward to encourage certain behaviors is not fast enough. Imagine playing in a D&D game where you got a hit point and attack bonus every time you described your action in a cool way. This would immediately change behaviors in your players. So stunts in Heirs give you a chance to earn a Destiny point. In my games, I use replica pirate coins as Destiny points, making them even more fun.

-Stunts keep the players interacting between the rules and the game fiction (back and forth). The game fiction gives stuff (content, ideas) to use in stunts, and players are rewarded for attempting stunts thereby increasing the amount of descriptive game fiction. It creates a positive feedback loop in that doing cool stuff rewards you with tools that allow you to do more cool stuff. These include tools for the player (details in the fiction to work into descriptions and future stunts) and tools for the character (Destiny points). For example, in an early playtest, a player used a critical success in a stunt to narrate slicing a water-logged zombie's guts open to find a starfish. This unlikely starfish was used (by players and the Game Director) in several further stunts, fumbles, and criticals throughout the scene, keeping the players laughing.

-Stunts are invented by the player, not a predetermined list that players choose from. I did not want to try to make rules for every potential stunt because this would stifle creativity, could never be exhaustive, and would need too much consulting of the rules during play. Any time a player comes up with something crazy, the Game Director can just say, "Let's handle that with a stunt."

-Stunts do not involve the "subjective judgement" of the Game Director to decide if a stunt is worthy and deserves a certain bonus, because it just gives the player a chance to earn the bonus (a Destiny point). In running games like Exalted, I felt social pressure to give bonuses to players, even if I did not really love their stunt description. It somehow felt personal to give one player a +1 and another a +2. In Heirs, I can just give the "keep rolling" hand gesture to get more description and reward the player with the potential to gain a bonus. It feels "safer" and less of the GM-is-all-powerful / semi-adversarial relationship.

-It is okay if a player cannot think of something cool to do. I've played high narrative games like Wushu that bog down because each turn, players must think of cool things to do. The game might be running along great, but suddenly a player just runs out of ideas and the pace stalls. Other players often jump in to help, but regardless the game seems to hit a bump in the road. In addition, the social pressure for good narration is too much for some players. Heirs should run fast because players must plan their stunts when it is not their turn. Also, if they cannot think of something cool, they can just make a normal attack or other normal action (but then they are not rewarded). Players who just don't think quickly enough can still enjoy other player's cool descriptions without slowing the game down.

-Stunts make the game more like an action movie, i.e. more cinematic. Unlike other games that have a "cinematic" version (which usually just means players get more character generation points or more "hit points"), stunt rules make Heirs actual play more cinematic. Fights end up like wire-fu with jaguar knights, pirates, and capoeira masters. Action is fast, and it is easy to picture the action in your head because you are not bogged down with bookkeeping, consulting rules, or difficult math and instead players are contributing to the narration.

RPG Inspiration Cards are Available!!

With the release of Obsidian Serpent Games' RPG Inspiration Cards (available at theGameCrafter.com), I wanted to post a sample use of the cards. If you would like to share how you use the cards, send your ideas to me. I will collect them and share them all.

As a player in a new fantasy RPG campaign, I used the cards to help create make my character's background using the RPG Inspiration Cards and a modified 3x3. A 3x3 is an idea I found on RPG.net. You can read more about the basic idea here: http://wiki.rpg.net/index.php/Midnight_PRE-GAME_Character_Background.

I modified the idea slightly for use with the cards. I laid out nine random cards into a 3x3 grid. The first row represents friends and family, the second row represents contacts and acquaintances, and the third row represents enemies and rivals. In addition, the first column represents the past, the second column represents the present, and the third column represents the potential future.

 pastpresentpotential future
friend/familypast friend/familypresent friend/familypotential future friend/family
contactpast contactpresent contactpotential future contact
enemy/rivalpast enemy/rivalpresent enemy/rivalpotential future enemy/rival

I used elements of the random cards to inspire my ideas using this framework. Here is my example:

My GM told me that the new campaign would be a fairly typical fantasy game mainly set in a large city. The game would involve intrigue and politics with the PCs eventually (potentially) becoming influential players. The city is filled with corrupt officials, powerful guilds, and rich thieves (think Fafhrd and the Gray Mouser's Lankhmar). With that information, I dealt my cards and got:

past friend/family
present friend/family
potential future friend/family
past contact
present contact
potential future contact
past enemy/rival
present enemy/rival
potential future enemy/rival

With this, I started to put my character's background together. I am a young, low ranking member of the city's wizards guild. Here are the meanings I assigned to each position/card.

past friend/family: Lovers - Almost one year ago, my romantic partner of several years disappeared on a stormy night, leaving no clues.

present friend/family: Wine - My father, a former bare-knuckles brawler turned wine merchant drinks more than he sells, barely keeping his business afloat.

potential future friend/family: Midwife - An influential member of the midwife's guild was a friend of my deceased mother. She has been making overtures of alliance between out two guilds.

past contact: Riddle - All wizard guild members are assigned an adviser. My original adviser left the city a little over one year ago, claiming he was on an assignment for the guild. Since that time, mysteriously no one in the guild seems to have any memories of him except for me.

present contact: Owl - My current adviser is a wise, experienced wizard who seems to have little time for me, but is nonetheless friendly. If only I could prove myself to him.

potential future contact: Wave - The merchant guild is slowly taking over control of the city by wrenching power from the decadent royalty.

past enemy/rival: Council - The previous merchant guild's council member started their rise to power. He was very antagonistic towards the wizards guild, but he has since lost his council seat.

present enemy/rival: Temple - The temple to a particular god is a fierce opponent of the wizards guild, going so far as to claim magic is evil.

potential future enemy/rival: Trickster - A fairly high ranking member of the wizards guild has always seemed to hate you for some unknown reason.

Social Conflict in RPGs

Situation A:  The role of the rules vs. role-playing when a PC make a social action against an NPC

Case A1:  A PC wants to convince a guard to let the PC into the castle
Case A2:  The PC wants to seduce a woman in a bar
Case A3:  The PC wants to compel a witness to admit to lying with his testimony
Case A4:  The PC wants to determine if an NPC is lying
Case A5:  The PC wants the NPC to agree to going to work the next day naked and singing a song to his boss. 
Case A6:  Same as case #5, but wants the NPC to actually do it (not just agree to doing it)  (Note that this is similar to case A1, but with more extreme stakes that are delayed).

In some RPGs, all of the above cases are handled with role-playing only.  The dice nor rules would ever be consulted.  The GM would just decide the effects based on the merits of the player's words, the situation, and the NPC's personality.  In case A4, the GM may or may not intentionally give some "tells" to hint to the players that he is lying.

My problem with this is that a player who has trouble thinking on his feet will never be able to convince someone of anything (and therefore could never successfully play that type of character).  For example, I could never be able to portray someone who could pick up a woman in a bar (case A2) because I don't have that skill as a player.  Also, the GM might not be a good judge of the qualities of a pick-up line.

My preferences is to use role-playing combined with rules.  Several RPGs use this method.  A player's role-playing serves as a modifier (normally set by the GM) to a skill roll of some sort.  Other modifiers are based on the situation and NPC personality.  For example, case A5 and A6 would have huge negative modifiers in most systems.

Case A4 can still be problematic.  If a PC fails his "detect lie" roll but the player still feels like the NPC is lying, should the player portray his character as if he believed the NPC even though the player does not?  Perhaps the "detect lie" skill only gives an additional way a player can detect a lie (if the player was unable).  This is connected to the next topic - social actions against a PC.

Situation B:  The role of the rules vs. role-playing when an NPC makes a social action against a PC

Case B1:  An NPC wants to convince a PC guard to let the NPC into the castle
Case B2:  An NPC wants to seduce a PC in a bar
Case B3:  An NPC wants to compel a PC to admit to lying with his testimony
Case B4:  An NPC wants to determine if a PC is lying
Case B5:  An NPC wants the PC to agree to going to work the next day naked and singing a song to his boss. 
Case B6:  Same as case #5, but wants the PC to actually do it (not just agree to doing it)  (Note that this is similar to case B1, but with more extreme stakes that are delayed).

In many games, the GM will argue for the NPC and the player will decide his own actions based on what he thinks his character would do.  Hopefully, he will portray his character's personality if the rules have a mechanism for this.  Many games have a reward system (such as experience points, Fate points, etc.) for when a player chooses to disadvantage themselves when in-line which they character's flaws.

One of the big questions I have is:
Should an NPC be able to make a roll to compel a PC to take a certain action?  If so, how do you enforce this and what are the consequences?

Many players have a problem with this because it limits their agency.  But how is this different that a character being knocked out in a physical conflict?  Of course, being unconscious limits a character's agency as well.

Let's take case B1 as an example:
A PC is a guard and an NPC with a high persuade skill wants to get in the castle.  Can the GM role-play the situation then make a roll for the NPC (something like a Persuade skill vs. the PC's willpower) and say, "He convinces you to let him in."?  Would it matter if the NPC used magic (either to boost his own persuasion skill or to reduce the PC's willpower)?  Why or why not?  A player may choose to portray a weak-willed character and may role-play it when he wants, but should the rules force this?  If so, should the rules use a carrot or a stick?

Perhaps case B4 is the most straight forward.  Most systems would allow the NPC to make a detect lie roll or would force the PC to make a bluff roll.

Fear and morale and mind control effects often all under Situation B as well.

It would be interesting to classify how specific games handle social conflict.

Games either:
1.  Have no social combat system
2.  Only have social combat for PCs affecting NPCs (sometimes this is explicit but often the autors never mention it one way or the other)
3.  Have social combat both directions (including NPCs affecting PCs)

Sometimes, it is not a clear distinction.  For example, many games only have Fear and Morale as social combat against PCs. For example, a character may suffer a Fear penalty if he fails a Willpower roll.

PC vs PC sometimes is a special case.


It seems like games handle social combat against PCs in three ways:

1. Ignore it.  Many games do not have any rules for social actions against PCs and only deal with it via role-playing.  A subset of these games also lack rules for PC social actions against NPCs.

2.  Carrot.  Some games reward players for allowing NPCs to affect them socially, especially when it causes problems for the PCs.  This is sometimes called "good role-playing" as the player accurately portrays his character's quirks and disadvantages. Example rewards include additional experience points, Fate points, Destiny points, etc.  The reward could be immediate (like Fate Points) or delayed (like experience points).

3.  Stick.  Some games punish players for not allowing NPCs to affect them socially.  Again, this could be an immediate cost (like costing Willpower [Exalted] or Fate points [FATE]), an immediate penalty (like CoC???), or a delayed penalty (like reduced experience points)

Some games also have a mixture of these.

Note that some games include social combat, yet still do not address Situation B.  For example, Burning Wheel's Duel of Wits.

A few examples:

Song of Ice and Fire
Call of Cthuhlu

What Gamers Want Out of Their RPG

I'm sure it has been around for a while, but I just came upon Levi Kornelsen (?) excellent website filled with gaming goodies at sites.google.com/site/amagigames (formally at www.amagi-games.org).  In particular, I really like his list of things gamers want out of their game found at http://sites.google.com/site/amagigames/the-what-i-like-glossary (url updated).  This seems so much more useful and friendly than the player types lists found in Robin's Laws of GMing and the DMG2 (and other lists).  I like how they are unburdened by preconceptions, and the list has a few interesting motivations that I had not thought of before.  Go check it out, then come back here.

I have a couple of ideas for additional entries.  Unfortunately, I do not have any ideas for cool names for the entries, but I will describe them as best as I can.

1.  The joy of feeling special and powerful.  It is the excitement of using all your kewl powerz that make you unique.  I don't think this is the same thing as Ludus nor Fiero, but similar.  I think those people who like to tell you stories about their character often have this type of motivation.  The story is often something like, "and then I used my Sneak Attack ability to kill the evil wizard that no one else in the party was able to effect at all!"  This might sound like I have a negative view of this motivation, but I do not.  I have some fun using kewl powerz myself.

2.  After a lot of self reflection, I think my own primary motivation and "What I Like" is the feeling of being smart.  I want to figure out the riddle, come up with a cool way to defeat the villain (using only a head of lettuce, a roll of duct tape, and a spoon), find the murderer, etc.  It is similar to Ludus, but not just in terms of the mechanics of the game.  I do want to use the mechanics well, but I also want to be smart about the other elements of play (the plot/situation my interaction with other characters, etc).


Conflict vs. Task Resolution

After reading about the differences between these two concepts a while back, I remember thinking, "Yeah, I get it." and going on. Later, as I read more posts and listened to podcasts about it, I stopped getting it. To me, the distinction between these terms fell apart upon close inspection. I think a main reason is that different people use the terms differently. Despite this, I felt it was a false dichotomy. Here is the evolution of my thinking:

Stage 1: Task resolution is about small things and conflict resolution is about a bigger thing. I then read several posts specifically saying the difference is not about the scale of the conflict, so I evolved to:

Stage 2: In Task resolution, a failure means X does not happen, but in conflict resolution, a failure means Y happens. After more thought, this seems like it had more to do with stakes setting than task vs. conflict resolution, so then I moved to:

Stage 3: The differences between the two is more of a dial, than a switch. To quote Vaxalon from story-games post from Sep 11th 2006

"Resolution can be nested.
Let's take a standard DnD combat. The heroes are bursting into the evil temple to rescue the princess.
In the middle of the fight, stop time right where the fighter is swinging his sword at a temple guard. Here's some of the layers that may be in play at that moment:
"Does my blow hit the guard?"
"Does it do enough damage to remove him from the fight?"
"Do the guards keep us from getting to the altar before the priest sacrifices her?"
"Do we save the princess?"
At what point does it become conflict resolution?
I would argue that it becomes conflict resolution when the answer is nontrivial. "

But this means that the definition is dependent on the individual's tastes about what makes something "nontrivial".

Stage 4:
I was still unsatisfied about the differences and I started to write up my thoughts when I came across Eero Tuovinen's thoughts from story-games, Feb 2010. This has everything I wanted to say, but could not put together. Thank you Eero, for clearing it up for me.

Rewarding Players

I am a big proponent of the idea that game tastes and preferred play styles vary greatly. With this in mind, groups should be intentional in picking a game that matches and supports the preferred style of the group. There is no game that works for everyone. Heirs to the Lost World was written with a certain play style in mind and will not work well for people who prefer different play styles. (side note: one of the purposes of this blog is to make sure people understand that style so they will know if Heirs is for them).

I like games and GMs that encourage and reward the preferred play style. This will reinforce the preferences of the group. Many old-school/traditional RPGs use Experience points to do this rewarding. I feel that the reward is too late. People respond more to an immediate reward, not one that comes hours later. Many games do this with things like Fate Points, Bennies, Hero Points, etc. James Bond 007 was my first exposure to this idea, and I fell in love with it. Since I want Heirs to be a wild, cinematic game that encourages stunts, it gives players a chance to immediately earn Destiny Points for attempting these stunts. The players can then use these points to attempt more stunts. Being awesome gives you the tools to be more awesome. A positive feedback loop of awesomeness.

In movies, some of the best parts are when the hero fails miserably. Therefore in Heirs, you can even earn Destiny points by turning a failed stunt into a fumble. You even get to describe your own fumble result. These player-narrated fumbles often become the most talked about moments of play.

Destiny points are also a way to encourage good role-playing. Instead of punishing players who do not portray their characters well (by skimping on experience points, you also earn Destiny points when you role-play your complications or motivations in ways to get you into trouble.

My play group uses replica pirate coins as the Destiny points. It makes it easier on the character sheet and encourages players to use them more frequently because they are fun to toss around. As the Game Director, I make sure I give them out frequently so players do not try to horde them. Also, they are NOT used in character advancement.

Niche Protection

Back in my D&D 3.x days, one of the things I liked the most in the source books were the prestige classes. They seemed like such a cool way to connect the setting with the mechanics. They were rules rich with flavor. This seemed at odds with my dislike of class-based systems in general. I always felt confined by them and preferred point-buy character creation. Class-based systems do have an advantage with regard to niche protection. Characters have a shtick that makes them special. In addition, when a player chooses a particular class, this serves as a flag to the GM about the types of challenges he or she wants to face in play.

In Heirs to the Lost World, I wanted to use a point-buy character creation, but I also wanted to keep some of the benefits you get with niche protection. I did this with two mechanical ways and two "suggestions" in the book.

The first mechanical way Heirs provides for niche protection while still being a point-buy system is the use of Paths. Paths are sets of skills and abilities that characters can "buy", but they are optional. They are kind of like prestige classes if there were no regular classes. A player could easily portray a thief by buying the appropriate skills and Assets, but he could also take the Thief Path as an option. Within each path, players pick from a short list of special abilities. Paths also are a great way to tie in the setting. Players can take Paths such as the Eagle Knight, Jaguar Knight, Lightning Warrior, Blood Mage, Capoeria Master, Alchemist, etc. These Paths are unique to this setting.

"Discounts" on Assets
The other way Heirs mechanically supports niche protection is through "discounts" on certain Assets. Assets are special abilities a character can take. If you have a particular trait at a level of 4 or higher, certain Assets are one point cheaper. Therefore, the Traits that you select at 4 or higher determine which Assets are discounted. For example, while any character can take the Inspire Asset, it is cheaper if your Presence is a 4 or higher. In fact, all Assets dealing with social interactions are cheaper if you have a Presence of 4 or higher. The same applies to other traits such as Notice, Agility, etc.

"Something Special"
In creating characters, I give advice that players come up with "something special". Players should think beyond the basic archetype and think of something that makes their version of the archetype different. This addresses a problem with class-based systems where virtually every version of the same class looks the same. Many Assets can work as the "something extra" or players can just make up something new. For example, a player should not just be a Conquistador but rather could be a Conquistador who "went native" and married an Aztec blood mage and is now uncertain about his religious beliefs.

Adventure Creation
In the Game Directing chapter, I give advice on creating adventures based on character's complications and motivations. This will ensure that every PC has a reason to go on the adventure. Often this means customizing every adventure to connect each PC to it. The adventure planning sheet (in the previous post) can help with this planning.

These four methods help produce some of the benefits of niche protection within the point-buy character creation system of Heirs to the Lost World.

Niche Protection at saveVSdm.com - The first of apparently several posts on niche protection.

Adventure Planning Sheets

I used to struggle with planning adventures - planned way too much. This was bad because either the planning was wasted (because the players did something completely unexpected - as they always do), or I subconsciously railroaded them to fit my plan, which resulted in poor play. Now a days, I try to plan using a scene structure with a few set-pieces. That way, I only need a few "if they do this, I'll use this scene, otherwise, I'll use this scene" kind of things, with a little, "if they ever do this, use this scene". I think if I pre-plan the most likely choices of the players, I have a better adventure, even if they do something completely different.

One nice thing with this planning is I can make sure I have scenes in the adventure connected to each character's complications/motivations. To help out with the planning, I made a little Adventure Planning Sheet. Get it from the downloads section at: www.obsidianserpent.com/downloads.html.

This sheet can be used for brainstorming, or virtually the entire adventure plan. When I use this sheet, I still make up the stats for the NPCs, but I usually use my generic NPC cards with slight modifications. The big important set-pieces take a little more planning. I'm now working on a set-piece planning sheet. These big encounters seem to work better when I have already come up with interesting twists and stunts. The planning sheet will help make sure I have everything thought out.

Other Resources:
5 Room Dungeon By: JohnnFour.
The Adventure Funnel.
How I GM, Part 1: Plot-Writing.