When I was younger and thus poorer, I liked to make gifts for people rather than purchase them. One year I came home from college for a weekend that coincided with my cousin’s first or second birthday party, and I decided to make him a small puzzle with a picture of a duck on it. I handed it to his parents, and during dinner they helped him put it together. I didn’t pay terribly close attention while he did, but whatever conversation I was having at the time was interrupted when I heard his jubilant cry upon completing the puzzle, staring at it for the first time.
You see, up until that moment he hadn’t recognized that there was a cohesive picture among the pieces, just that each piece had some colors and shapes. He didn’t know why certain pieces fit together, only that they did. It wasn’t until he saw the whole picture that he’d realized he’d been assembling a duck, and it was the biggest surprise in the world to see something recognizable appear out of nothingness.
I know a very smart GM with very poor taste in sports teams who likes to point out that nothing in a game is real until it’s explicitly presented to the players. The party can assemble any number of clues toward some in-game mystery: the brutal way the mayor was murdered, the large pawprints outside his windows, the full moon that night, the furry creature spotted fleeing the scene. But until they see it with their own eyes, they can’t be certain that they’re hunting a werewolf. It could just be a transformed druid, or some otherworldly monster, or just a very hairy person with terrible footwear.
The trope for this is Schrödinger’s Gun, and it’s probably the third most powerful weapon in a DM’s arsenal (after a BFG and an actual, at-the-table gun). The whole idea is that the GM is free, and in some systems encouraged or required, to change the details of the plot if it makes the game better for it.
In the example above, say the players determine that the murderer is a werewolf and load themselves up with anti-lycanthrope weapons and spells for a hunt. This is fine, since you figured it was a werewolf all along. But for some reason or another, you’re now thinking that maybe it’s not such a good idea. Maybe you came up with a better story between sessions, or maybe the players had a more awesome idea that you want to convert to on the fly, or maybe you just see how easy the fight will be when the party is overly prepared and you feel like being a jerk. You are, at any time, allowed to change what actually happened because, until the players see it, what happened could be almost anything.
Generally, I think this trope is understood in terms of story progression. Like the above, it refers to something that causes a DM to change their mind about some aspect of the plot that the players haven’t seen yet. But generally, that’s not how I end up using it (perhaps because my plots are always great to start). Instead I tend to use it for balance purposes.
I believe that the ideal fight is difficult but winnable. About a million years ago, I said this:
I like high difficulty. I tend to find that the combats that the players still talk about are the ones where they felt closest to death…Hard-won combats are more memorable and more enjoyable long-term than easy ones, even if at the time it feels terrible.
A fight that’s too difficult is frustrating for obvious reasons, but a fight that’s too easy is equally boring. Generally a good fight should consume some resources (hit points, spells, daily powers) but not tap them out.
So when a fight is starting to lean one way or another, that’s when I break out Schrödinger’s Gun. If a fight is too easy, an extra monster or two may appear halfway through, or a bloodied monster may gain a power that doesn’t actually appear in its stat block. If a fight is too hard, I can do the opposite: the planned minions don’t appear, or the monster loses its bloodied buffs or opts not to use a power it has available. Usually when I invoke this trope, it’s to give a creature a spontaneous “vulnerable X all” to speed the fight up some so that we can end the session on time or finish with a particularly boring encounter. The players aren’t aware of the exact plans you have or the numbers that go into a monster, so you have the option to change them on the fly.
4th Edition is a shade better at this than 3rd for a few reasons. One is the minion mechanic, where new enemies can join a fight that don’t completely rewrite its difficulty. Another is the higher power level, where adding a new trait (like vulnerable or resist 5 all) won’t completely kill the pace of the fight or invalidate certain characters. The “bloodied” status is a great time to make changes to monsters, one I’ve stolen for use in other systems. I think the most helpful change is the ease of reskinning, where a monster with the same stats can be any number of things by just changing its description and the names of its powers. Based on what your players are doing, you can change a goblin hexer to a kobold sorcerer, an orcish shaman, a human cultist, or even a plague elemental on a moment’s notice. The mechanics of the fight are the same, but the feel of it has changed dramatically based on what your campaign needs.
There are systems, like FATE, that are actually designed with this in mind. By spending fate points (or just making sense), a player can cause things to appear that were always there. For example, a player declares “I know about monster X, specifically that it nests in sewers”, and a GM responds “Sure, why not.” At first blush D&D doesn’t seem like a system built to handle such changes by players, but most DMs I know do it at some level. Maybe it makes sense that there’s a bookcase in a room even if I didn’t describe it, so it exists now that a player’s asked.
Beyond that, D&D explicitly mentions changing things on the fly as part of its core ruleset. It’s just that it only mentions it for dice rolls, and instead of “making things up as you go along”, the books call it “fudging die rolls.” Which is the same thing, but you get to feel like you’ve subverted luck as you do it. The 3rd Edition Dungeon Master’s Guide 2 (a book even less read than the 4th Edition version) goes as far as this:
…you should make your world serve the game, not the other way around. No part of your world is set in stone until it becomes part of the game. You might have an emotional connection to some elements of your material, but your players don’t, because they haven’t encountered those elements yet…Look at your background information as a work in progress, subject to instant revision if the moment demands a change that would result in greater entertainment…Even details that do become part of the game can be fudged on occasion. You shouldn’t change details the players vividly remember, because that punctures their belief in your imaginary world. Minor background details, on the other hand, should never be allowed to get in the way of an entertaining choice.
The point is that everything that hasn’t been shown to the players is as good as every conceivable outcome. To the players they’re obviously fighting the campaign villain, but you as a DM can spontaneously decide that this is just a clone, or a loyal follower, or a mind-controlled pawn. You don’t even have to decide this by the time the fight starts, and occasionally not even right after it ends. Until it’s spelled out, all the players have is shapes and colors, and you’re allowed to change the final picture any way you can.
Note: This isn’t even the only way players are like one-year-olds. Remind me to complain sometime about NPCs and object permanence.
Word.
Most important thing in GMing is don’t get too attached to the world you’ve built in your head. I avoid games that require me to do it at all, if possible 🙂
… though I’m starting to learn that aggressive scene framing is a really close second.
Oh, and an interesting corollary of Schrödinger’s Gun is this: Even if you describe your world fully, the images the players form in their own minds aren’t going to exactly match what’s in yours. So even after something’s been brought into the fiction it needs to stay just a bit malleable until a consensus solidifies.