Back in part 1, I said that as a rule of thumb, you can assume you’ll have three to five scenes between opening and finale. Each of these scenes should last around half an hour each. That’s actually quite a lot of play time. Depending on your game system, that’s one or two fights (or a lot more, in some lighter systems); it’s two or three conversations with an NPC; it’s exploring a few rooms in a dungeon; it’s a lot of planning and the quick execution of a plan.
I said three to five scenes – in practise, this is a minimum number. Some games have much shorter ‘scenes’ – a dungeon crawl needs a lot more than three to five encounters, so double the number of scenes for that. You can also have optional encounters, alternate scenes (if the PCs accept the bargain, go to scene 4; if they refuse, run scene 5), especially long scenes and so forth. Three to five, though, is enough for most games.
Each scene should be relatively self-contained. Try to format your scenario so the GM can easily see where each scene starts and finishes.
Scenes fall into several loose categories.
Combat Encounters: The PCs fight bad guys
Roleplaying Encounters: The PCs talk to someone
Challenge Encounters: The PCs have to deal with a problem
Transitions: The PCs go from A to B in a notable fashion.
Locations: The PCs explore somewhere
Timed Event: Something happens at a particular time, whether the PCs are there or not
Triggered Event: Something happens in response to something the PCs do
Scripted Scene: NPCs do a lot of stuff, the PCs respond
Obviously, these types of scenes aren’t exclusive – you can have the PCs go to a location, meet someone there, roleplay a bit, and then beat up some foes all in the same scene. I’ve divided them this way because these are the ways that they start. A location scene happens when the PCs go there, a combat scene happens when they meet the bad guys and so forth. Most scenarios contain multiple types of scenes, by the way. Take a cop game where the PCs are investigating a murder. Most scenes are triggered by the actions of the characters – they go to the murder scene, they talk to the witness and so forth – but a car chase is a challenge scene, and the bit where the corrupt Lieutenant calls the PCs off the case is triggered when they get too close to the truth.
A wise man once said to me, ‘you can do whatever you want when you’re GMing a game, as long as you let the players kill something once a session.’ The great thing about combat is that it’s hard to get wrong – a scenario that’s just five fights in a row can actually be enjoyable, as long as all the PCs have something to do in a fight and you’ve got a good set of combat rules.
Con scenarios pose an interesting dilemma when it comes to combat – if someone’s dedicated three hours of their time and a few quid to your game, is it ok to kill their character? In practise, the answer is ‘not until scene 5.’ You should challenge the PCs in a fight, but the characters should not be in actual danger of dying until at least two-thirds of the game session has elapsed. (Exceptions can of course be made for really, really annoying or stupid players.)
Keep the combat rules simple; keep the combat interesting. You can’t assume a high degree of rules knowledge on the part of the players. Some people’s eyes glaze over when you say ‘attack of opportunity’. Combats should never rely on the players knowing the rules, but should assume a modicum of common sense. A dramatic place to have a fight – a rope bridge over a chasm, a belltower, a market full of scared peasants and livestock, a corporate office full of cubicles and meeting rooms – will serve you better than rules cleverness.
Make both the player characters and their enemies interesting in a fight. Battles which devolve into dicerolling aren’t fun for everyone. Give characters gimmicks, even if these gimmicks aren’t necessarily the best tactical option in a fight. Errol Flynn swinging from the chandelier is more fun to play than Meatshield the Warrior, even if Meatshield is a tactically better choice.
Make sure everyone has at least some way of participating in combat. This doesn’t mean that every character needs to be a killing machine, but the kindly old professor should at least be able to whack people with his umbrella and contribute something. Alternatively, you can have combat situations where only one or two PCs can participate, but ensure there’s some other equally vital role for the other characters to perform at the same time. ‘Mecha Pilot Rodin and Mecha Pilot Jimbo go and hold off the bad guys, while we get the McGuffin from the alien spaceship’ works. ‘Mecha Pilot Rodin and Mecha Pilot Jimbo fight while we sit here for ten minutes and watch’ can also work. ‘Mecha Pilot Rodin and Mecha Pilot Jimbo play for an hour while we sit here bored’ must be avoided.
Don’t make the bad guys pushovers. If you’re going to have a fight, then it may as well last a few rounds. This is especially prevalent in gritty or high-tech settings, where a hit often means a kill. Don’t have the bad guy and his goons show up in a situation where one grenade takes them all out. Give the foes enough hit points/cover/other protection to survive two or three attacks from even the most dangerous player character attack. If you’re going to break out the combat system, make it count.
If your scenario is going to feature a lot of combat, then it’s a good idea to have a simple early fight against weak foes so the players can learn how the combat rules work, how their characters work and how the group fights together.
Be aware of the effects of damage in your system. Yes, a D&D cleric can patch up a mortally wounded warrior with a few spells, but he may need to rest afterwards. Will it break your scenario if the group pauses for twelve hours? In Traveller, one good hit can put a PC in hospital for weeks – is there a convenient high-tech medical facility with fast-healing drugs available to the PCs?
For most roleplaying encounters, it is enough to give the GM a brief description of the NPC’s goals and personality quirks, and maybe a few sample lines of dialogue or responses to likely questions from the PCs. Something like:
“Assistant Director Halsey is the PCs’ direct superior in the FBI. He’s a harried bureaucrat, suffering from stress ulcers and premature baldness. He just wants the case closed, and isn’t particularly interested in tales of UFOs and paranormal phenomenon. Just get him an answer that he can write up in a report and he’s happy. He’s here to give the player characters their assignment, and to caution them against blaming everything on aliens.
Halsey tells the characters that four children have vanished from Williams High School in the last week, and that local police are baffled…
Q: Where were the kids last seen?
A: At the school. Two vanished between classes; another walked into the shower in the locker room and didn’t come back. The last one went to the guidance counselor, and vanished while the counselor’s back was turned…’
There’s enough there for the GM to play the NPC. Optionally, you could mention an actor as a visual reference (‘he’s like Taub from House‘), or give a few mannerisms or catchphrases (‘he winces visibly when the PCs mention the supernatural’), but don’t go overboard. Three or four lines of description is probably overkill. As long as there’s a quick sketch of the character, and notes on what goals he has and what information he can impart to the PCs, that’s enough.
One common mistake is to include too many NPCs in a roleplaying encounter. Assistant Director Halsey is fine on his own, but don’t put him in a conference with the PCs and three other NPCs. It’s hard to play more than one NPC at once.
You can have multiple NPCs present, as long as the players can talk to them in sequence, not at the same time. So, if Duke Fotherington and Mayor West are both at the ball, then the PCs can meet the mayor first, then the duke, or split up so you run the conversation with one NPC first, then the other – anything, so long as the GM is only playing one character at a time.
If you absolutely have to have multiple NPCs in a scene, then:
Make sure there’s a strong contrast in speech patterns and demeanour
Have one NPC do most of the talking
Involve the player characters as much as possible. Have the NPCs ask questions of them, demand their participation, appeal to them for support – anything to avoid the absolute death that is two NPCs having a conversation or an argument in front of six bored players.
Any roleplaying scene has to have sufficient content to last several minutes. Give the player characters something to react to or argue over. Yes, the players may come up with their own cool things to roleplay about, but you can’t rely on the players’ own creativity to carry the whole scene.
For example, I recall one Vampire con scenario which started off something like this:
Scene 1: The PCs are invited to a party
Fair enough. It’ll do as an intro.
Scene 2: At the party, the assembled vampires ask the PCs what they think of the prince and the party.
And here the game runs aground. Only one or two of the PCs had any listed opinions on the prince, there was no real description of the party, and the ‘assembled vampires’ were a faceless crowd without any personalities or even individual names. The result was thirty minutes of painfully stilted play.
GM: So, kindred, what do you think of the party?
PC: Er…it’s great.
GM: And… the prince.
PC: I dunno. Also great. It is so generous of him to invite us to this elegant soiree.
PC: You said that very crypticly.
GM: Did I?
PC: I talk to someone else.
GM: He also asks what you think of the party.
PC: I leave.
GM: You can’t.
At the same time, you don’t need to spoonfeed everything to the players. Just give the GM some ideas, and ensure that the players have enough information on their situation to make stuff up. Let’s tweak the Vampire party to make it a little easier to run.
Scene 2: The PCs arrive at the party. It’s being held in a private ballroom in the biggest hotel in the city, and feels like a Mafia wedding. Lots of men in suits, women in expensive dresses, red wine and concealed weapons. The servants are all mortals or ghouls; there are around fifty vampires here. Some of the vampires have brought mortals with them, so you’re supposed to keep to the Masquerade here.
(To be honest, most players would have assumed most of that description anyway – it’s Vampire, and it’s a formal party, so it’s going to be in a swanky hotel or private mansion. Adding the fact that mortals are present pushes the players to be a little more circumspect about what they say. It’s just a little detail, but it’s an evocative one.)
Among the guests are:
Lady Jasmine: A femme fatale who flirts with any male characters. She considers the party dull and wants to liven things up – do the PCs agree?
Roark: A young vampire. He’s got a mortal woman, Cecily, on his arm. She’s obviously bedazzled by the glamour of the party, and is oblivious to the fact that Roark’s going to kill her before the night is over. Roark will play on this, letting the player characters in on the joke of her impending demise.
Hamilton: A stuffy elder vampire and a close ally of the prince. He considers the PCs to be beneath him, and will sneer at them.
That’s enough to give the GM something to work with. None of the three NPCs has any purpose in the scenario other than to serve as a roleplaying foil for a few minutes, so you don’t need to provide any stats or other background. (well, you might for Roark, if any of the player characters decide to save clueless Cecily).
Note that you only need to go to this much trouble if you specifically plan for a lengthy roleplaying scene as part of your outline. For other, incidental NPCs, you don’t need to provide as much detail. For example, in a fantasy game, you might have a scene where the PCs arrive at the king’s castle and are challenged by the guards. If you want a roleplaying scene where the PCs have to argue their way past the guards, then you’ll need to give the guards some personality and specify why the PCs aren’t allowed in. If being barred from the king’s castle is just a passing aside, then you can just say something like ‘the guards will grumble, but will let the PCs in if they present a good argument for seeing the king.’
Roleplaying encounters are heavily reliant on the skill of the GM to run, but that doesn’t mean the writer can just say ‘the PCs roleplay for an hour with the NPCs‘ and leave it at that. You must provide just as much support for roleplaying encounters as any other scene.
Challenges are problems the players need to overcome through some means other than hitting and talking. The problem can be a clear obstacle (‘how do we get into the castle?’) or a question to be resolved (‘so, we’ve just discovered that Bob’s the traitor; let’s come up with a plan to capture him’).
The challenge should be big enough to take up a whole scene’s worth of play. ‘How do we get into the castle?’ is a nice big challenge – the players can plan how they’ll set a fire to distract the guards, then scale the walls and hide in the stables on the far side, or sneak in on market day disguised as monks or whatever. The planning and execution of such a scheme will take half an hour of play.
By contrast, ‘how do we get past this door?’ probably isn’t a valid Challenge for a whole scene. It’s a single Lockpicking roll.
An aside: some scenarios have scenes that ‘spotlight’ one of the player characters, by giving that player a chance to do something that only that character can do. If there’s a thief player character, then there’s a scene where the thief get a chance to show off his sneaking skills by stealing a vital plot item. If there’s an Egyptian history professor, then there’s a scene where the professor must translate the hieroglyphics to find the lost tomb. These scenes are a handy way to include a potentially marginalised character in the plot, but there are two caveats. Firstly, make sure these scenes don’t last more than a few minutes. It’s boring for the rest of the group to sit there for half an hour while the thief solos the dungeon, or the netrunner hacks the network or whatever. Secondly, make sure that your plot is robust enough to work if the player character is missing or otherwise unable to act.
A good challenge involves more than dice rolling; it should force the players to discuss options, either in or out of character, and make tough decisions.
Some scenarios describe the PCs travelling from one location to another. It’s fine to include transitions like this, especially if they’re being brought to an important or secret location, or travelling in an unusual manner. ‘The PCs travel by jeep to the little village of Hosterfield, and are escorted by armed guards disguised as farmhands into the barn. There, a lift brings them down to the hidden submarine base…’
Players don’t really do much except react to transition scenes, so they should be kept short. They’re good for setting the mood or as a cooldown after a big fight.
In a location scene, the players explore a particular area. The size of the location varies depending on the game. The classic location scene is the dungeon, of course, but you could have the characters exploring a whole planet in a sci-fi game, or a single room such as a murder scene in an investigative game. The common element is that the players have decisions about the route they take through the location, and that they get to stop and examine important elements of the location.
If spatial relations are important, then provide a map.
If there are secrets or traps to be discovered, then highlight these in the text.
The chief danger in a location scene is making the location so large that it takes a long time for the players to explore. As soon as maps and rooms become important, the players become paranoid and cautious. Assume about ten minutes per room or important item in a room.
Timed events happen when a certain amount of time has elapsed. You can use in-game or out-of-game time for this. Both have their advantages.
Events based on in-game time feel less artificial. The enemy is going to attack at dawn, so the attack happens after the PCs all go to sleep for the night. The bomb’s set to go off an hour into the flight. The king comes home from hunting at twilight. The game feels like a living world, as events occur beyond the player characters.
The mistake I’ve seen made time and time again is failing to correctly estimate how long in-game events will take. The most common mistake is to have events separated by hours or days of in-game time. One scenario had the first timed event happening at noon on the first day, and the second event taking place on the evening of the second day. The problem there is that players are rarely willing to let time pass without doing something. There was nothing for them to discover in the scenario until the second event happened – the plot didn’t kick in until then. If the players had been willing to let go and skip on by 24 hours, then the game could have continued, but instead they kept running around this little rural village, interrogating everyone and chasing after every non-existent hint and rumour. By the time they were finally willing to allow time to move on, there wasn’t enough out-of-game time left to finish the whole plot.
A good GM can push the group on, but you should still avoid excessively long gaps. All the action in your scenario should fit into 24-36 hours of play in most cases. The longer the gaps between scenes, the greater the chance that the players will wander off or stall. If the ritual to summon Cthulhu happens at midnight, then the players should learn of this in the nick of time or just too late, not ten hours in advance (not unless you want the cultists to wander into a minefield).
Using out-of-game time is a trick that really works on in a con game. You’ve got to finish your game within a three-hour time slot, so you can use this to your advantage. Having an event that happens, say, two hours into the game can set up the finale. This works best in games where a race against time is part of the plot. The triggered event should be one that can happen at any time in-game.
Use triggered events to respond to the actions of the players. For example, when the characters kill the henchman of a bad guy, then the bad guy might respond by attacking some ally of the player characters. Like timed events, triggered events make the game world seem more real to the players, and makes your plot less static. They’re especially good in games that would otherwise seem very predictable.
Make sure you don’t make any unwarranted assumptions about the situation when the triggered event happens. For example, if there’s a triggered event in your game that states ‘when the PCs find the alien artefact, the FBI show up at the crash site and arrest everyone’, be careful of assuming that all the player characters are present when that happens. Some of them might be there; some of them might be off investigating something else, or trying to buy shotguns. Will your plot still work if the FBI don’t arrest all the characters? How hard will it be for the GM to adapt the scene to unexpected circumstances?
Finally, scripted scenes involve a lot of pre-scripted actions from the NPCs. The scene where the evil villian monologues to the captured player characters and launches missiles towards all the world capitals, the dragon smashing its way into the castle to devour the king, the cult ritual that summons the Great Old One – that’s all scripted stuff. If the most important thing in the scene is the action of an NPC, then it’s scripted.
Keep scripted scenes to a minimum. They may be necessary to move the plot along at times, but they’re still a crutch for a lazy writer. The whole point of a roleplaying game is to be interactive – scripted scenes force the players into a largely passive role.
Above all, don’t let the scene’s resolution hinge on the action of an NPC. The PCs must always be the ones to determine the ultimate outcome of a scene. This may be done in concert with an NPC – for example, in a court case, you could have an NPC judge listening to the testimony of the PCs, or have the PCs pass judgement on the pleas of an NPC, but you should never cut the PCs out of the action.
BUILDING ON YOUR OUTLINE
This taxonomy of scenes is a little artificial – I don’t think I’ve ever consciously sat down and gone ‘combat scene…combat scene…roleplaying…triggered…combat…’ – but by calling attention to the mechanics of each scene type and its length, the above list should help you when you’re turning your planned outline into a scenario. Break down each major event or encounter into one or more of the scene types, and see if you’ve got your three to five major scenes that make up the spine of your game. You can have smaller scenes and spotlight scenes and optional encounters and so forth too, but if you haven’t got three to five clearly defined major encounters, then there’s a problem with your outline.
Next week – the finale (of your scenario, not of this series…)