TombThe streets here are a concrete labyrinth. I try to go one block east, towards the ocean, and find myself crossing another bridge over the grey waters of the Miskatonic, and I’m back on the north side of the city, climbing up towards the civic monstrosity that squats atop Sentinel Hill. Transport Police, their faces hidden by gas masks – to protect against “typhoid”, according to the peeling posters in the subway – watch me as I march past. I don’t dare ask them for directions, and I can’t go back underground. I have to stay on the streets, even if I get lost again. Maybe if I find higher ground, a vantage point… a doorman ushers me in, making a familiar sign with his left hand as he does so, but too late I realise that the building I’ve entered is one of the cryptic and terrible windowless skyscrapers that loom over the city, their tops lost in the oppressive, low-hanging clouds. I cannot go back – I have to climb, struggling up flights of stairs that are clearly not made for any human frame…

Why, I am writing Cthulhu City, now that you mention it. Or rewriting, in parts, as the book has its own ideas about what it wants to be. A sandbox, maybe, where the Pillared City of Irem was lost long ago.

* * *

At Gen Con, I ran two prewritten scenarios: Kevin Kulp’s Valkyrie Gambit for Timewatch, and Ruth Tillman’s Midnight Sub Rosa, which can be found in Out of the Woods. In both games, I screwed up and misread key elements of the scenario (protip: running a game on the day after those Ennie Awards is never going to go smoothly). In both games, though, I was able to recover from my error and keep the game on track. Neither group noticed that anything was amiss.

Confusion & Conflation

In Midnight Sub Rosa, I conflated two locations. There’s one house where the main action of the adventure takes place, and there’s a guesthouse where most of the assembled non-player characters are staying. In my haste, I missed the guesthouse and assumed that everyone was staying in the same place. If I’d noticed my error in time, I’d have simply corrected the players, but a good fifteen minutes of play elapsed between me describing the building, and me realising there was supposed to be a whole separate guesthouse down the road from the country house, and rewinding play kills momentum in a convention game. I had to get ahead of the derailed train while it was moving.  (if you notice a mistake just as you make it, you can correct yourself – “oh, no, wait, they’re not staying here, there’s a guesthouse down the road” – but that’s a very narrow window. Once you’ve spent five minutes in-character complaining about the cramped rooms in the main house, that opportunity’s gone.)

Removing the guesthouse introduced two problems. First, it made it harder for the player characters to sneak around and investigate the various bedrooms. In a six-person con game, though, that problem solved itself: some player characters distracted the NPCs while the others committed a little breaking and entering. The second issue was a bigger one. Midway through the scenario as written, there’s supposed to be a ghoul attack on one of the NPCs as he walks down the isolated tree-shrouded laneway between the main house and the guesthouse. By moving his bedroom into the main house, I’d removed the opportunity for the ghouls to ambush him, and I couldn’t have the ghouls attack the main house midway through the scenario.

The ghoul attack scene is in the scenario to be a sudden visceral shock and to eliminate a particular NPC. It doesn’t need to happen on that laneway. So, I invented a reason for the NPC to leave the safety of the house. I described him as a smoker, and then later had one of the other characters complain about the smoke. Soon, a player character suggested that he and the NPC step outside for some fresh air where they could smoke in peace. They wandered into the gardens… and the ghouls were lurking in the trees nearby.

If the location of the ghoul attack scene was important, then I’d have had to come up with some other solution, but here all I needed to do was eviscerate one particular occult expert. Once I’d done that, and given the players a fright, the game was back on track despite my screw-up about the guest house. The key is to know the purpose of every scene, even if you have to change the setting or content.

The Case of the Missing Villain

In Valkyrie Gambit, I forgot to introduce the villain of the whole adventure. The villain’s supposed to show up in the opening scene, setting up a dramatic reveal at the end. (“It was you all along! Shock! Horror!”), but the players and I were having such fun brawling with mutant cockroaches that I ended the scene without bringing the villain onstage. I could have added another scene where the villain pops in, but it would have stuck out like a strange growth on the scenario’s spine. The shape of the story in a roleplaying game isn’t discernible when you’re in the middle of play; it’s only seen in retrospect, when the players look back and see the sequence of events from beginning to end. In a convention game, where you’ve got limited time and only a handful of scenes, I couldn’t get away with adding a new scene to add a new NPC – it would make the game feel unsatisfying at the end, even if the players didn’t notice in the heat of play, because it would have robbed that opening scene of its purpose. Pointless scenes are always rotten, even if they’re fun in the moment. (There’s a tension between the game that the players are experiencing right now, and the story that they’ll remember and tell afterwards. You can have a really fun, action-packed game, and then discover when you look back on it that nothing actually happened, that it was just running around and rolling dice without any consequence. You can have a perfectly structured compelling story that’s boring and frustrating to actually play through. For a good convention session, both the game and the story need to sing.)

It’s always better to call back and reuse material in a convention game. If the players introduce a concept in scene 1, then try to bring that into a later scene, even if you have to force things a little. In 13th Age games, for example, I’ll happily twist myself into knots trying to work in all the players’ One Unique Things, because it’s more fun for them to have contributed something that actually plays a part in how the story plays out. In Valkyrie Gambit, one of the players decided to play with the Timewatch rules by having his future self show up to help out in that initial fight. That gave me a justification for my replacement villain – it was a time-shifted duplicate of one of the mutant cockroaches, breaking the laws of time by skipping out in the middle of that first fight.

Using the time-shifted cockroach as the villain was the most parsimonious solution – it incorporated two existing elements (cockroaches, and the fact that time travellers can duplicate themselves), so it gave a sense of unity to the whole game when the player characters met the cockroach again in the final scene. It tied everything together. Look for ways to link back to earlier events and ideas, or to echo them.

Distraction With Shiny Clues

Another common landmine – which I gracefully leapt over this year, unlike the steps at the back of the Embassy Suites – is the logical contradiction, where you accidentally say something that breaks the logic of the mystery. You describe, say, an NPC closely examining a weird statue, even though it’s supposed to be locked away in a glass case. In that situation, look for a way to correct the mistake that involves the player characters finding out more information through active use of their Investigative Abilities. You could, for instance, describe the museum porter come back in with the glass case, complaining about how he has to clean it every few weeks because a strange black mold keeps growing on the inside, giving the player character with Biology a chance to whip out her microscope, look at some mold samples and discover that they’re very similar to a toxic mold found in certain Egyptian pyramids or somesuch (the clue doesn’t have to be relevant; it’s there purely to give the players a little reward so they don’t notice the plot bandage you just slapped on.)

Convention games are a particularly manic high-wire act for the GM when they go awry – as everything has to fit into one three or four-hour slot, you’ve got to find a solution to problems in time for that big finale. Always keep your nerve – if you screw up, keep going instead of backtracking. Prewritten scenarios are just suggested routes, they’re maps of what might happen, not strict scripts that you’ve got to follow. If you go off course, keep going and look for another turning to get back on track. Do it right, and the players will never suspect a thing.