I try as hard as possible to create the scene without keeping the characters in mind. I try to get inside the mind of the villain, and set up the location using his preferences:

“Okay, yep, I put a security guy with an MP5 submachine gun here. The guard changes every 4 hours. A new guy just got hired for the 3 pm to 7 pm shift. He doesn’t take his job very seriously. Here is a motion detector, but no guards. Security cameras are placed here, here, and here. The man in charge of the building looks like this, has lived in town for so many years, is married to such and such, he has a young daughter, and lives in an expansive house on the outskirts of town… He has had no security cameras placed in his private office, as he finds the the thought of people being able to watch him working distracting.”

Then I introduce the characters to the scene. They may do anything they please without threatening the flow of the adventure, because I DO NOT KNOW HOW IT IS GOING TO END. I have not planned the adventure to go in this way or that, but to just happen. The NPC’s should be reacting to the players, not the other way around.

My players love it.

This is a great cinematic device, and it went really well in our group’s campaign. It lets the characters be prepared, and keeps the players surprised at the same time. You can read the recap of our session by our GM in the comments on Law’s LJ page under the subject “Trying this out”.

For example, the mastermind would say, “Our first step is to distract the guards so that Mr. Hacker can gain access to the wireless router rack unit and quickly disable the bank’s already lackluster security protocols.” I’d say, “The bank lobby looks like this, this, and this.”

They’d put together a quick plan (I’ll trip on the floor mat and knock into you, and we’ll feign an argument; if the guard doesn’t take the bait then you ask for a manager and we’ll etc., etc., etc.), then the scene was played out, with some pre-arranged encounter triggers, but also a fair amount of improv based around their actions, rolls and ideas.

I think the best parts of a good heist are the unsuspected things that pop up, both complications the characters didn’t foresee, and also the characters’ “aces in the hole” that aren’t revealed until they actually happen–having to set up those kind of switchbacks sort of ruins the delight when they’re pulled off.

The GM’s job in the the Evolution method, and perhaps I should have include a few lines on this, is that you need to create encounters to sweat the player’s plan. No plan survives encounter with the enemy. So the GM’s job is to make things complicated.

For instance in the Italian Job, the plan for stealing the safe was about as perfect as it got. But when they captured the truck and opened it up, the wrong type of safe was inside. That is a complication, not a failure.

That said, the GM should pick the where the complications occur based on how exciting the complication will make the heist. The idea is to shake the players up and see if they can keep their cool, and think on their feet.

When I create opposition for my players in the Evolution method, I do not always put failure of the heist as a the negative outcome of the scene. The idea is to complicate, not crush.

So using a getaway as an example, the obvious failure condition to evading the bad guys is a combat encounter. An alternative is that players still evade the bad guys, but their safe house, where they were heading is now compromised, and now they have to find another place to lay low.

In my experience, when you let players design their own challenges, they will make things way more difficult, and way more fun than you could just on your own.

How do you get your players to feel good about letting you in on their plan if they know you are about to design something tailored to stopping that plan? Which is exactly what Peter is objecting to.

If they are planning on sneaking in, they are going to want to know what defenses they past, just so they can feel successful. So, when you do your session two planning, do you take that into account?

I think I’ve got the old school rivalry thing going in my group. I’d be shocked if they didn’t create a plan in session one, then rethink it 180 degrees over private (me excluded) emails up until session two.

an example: we were hired in shadowrun to get someone out of the city. we immediatly stepped in a normal car and drove off, but still the chinese mobsters had a plan and knew exactly where we were and what we were doing.

No, really.

Firstly, read Chalk Outlines Waiting to Happen. Grok how that game makes the fun, movie-style elements of the heist front and centre, and puts the boring stuff in the background. At all times, play proceeds on the assumption of overall success, but with drama and failure along the way nonetheless.

Secondly, watch Dirty Dungeons and get some tips on how to reward players for good ideas. If you happen to have access to Wilderness of Mirrors you’ll note that WoM is basically a game built around the Dirty Dungeon concept with a spy paintjob.

Thirdly, remember that the best parts of heist movies are the bits where something totally impossible and unexpected happens, and then you flashback to reveal it. That “Event -> Flashback” structure is important. If we knew exactly how the event would unfold beforehand, it would lose its entertainment value. So, if you’re not comfortable with allowing full player control, consider allowing players to surprise you, so long as they can justify it with a crazy flashback. “No, wait, it wasn’t the SWAT team that left the vault… it was us!”

Fourthly, heist stories are usually rife with what we gamers call PvP action. Intra-party conflict is always a riot in these situations, because you never know when it might explode and bring down the whole plot. At the same time, intra-party conflict is often just a ruse for the benefit of the mark or another party member. Leave the room and let your players dream up their own in-character conflicts — and whether they are real or fronted — without your knowledge or interference. Reward players at the table for any conflict in which you don’t need to do anything: a scene where players are arguing amongst themselves is a good time for a GM to prep a sinister surprise for later.