The Alexandrian has been writing (a lot) about railroading over the last weeks and I read an older article by the Angry DM from a couple of years back a few days ago, in which he touches on Schrödinger’s Gun (among several other things). Which made me think some more about both subjects and how they are related, and resulted in some of my own thoughts I want to share. The reason why you should read this article and not just the two I just linked, is that this one almost certainly will be a lot, lot shorter. ^^
So the basic question that has been pondered a lot among gamemasters for the last years, and almost certainly to some extend for well over three decades, is to what extend GMs could or should change their prepared and randomly generated material to compensate for shortcuts the players unexpectedly discovered. There seems to be a certain segment of paleo-academia disciples who advocate that GMs should be completely impartial and the game essentially randomly generated by dice, which will be used exactly the way they fall. What the players will do is entirely left up to them, with the GM not taking any action to promote or inhibit any decision the players might make. To keep it short and civil, I personally don’t see any appeal in that kind of game.
I am much more interested in games that have some kind of story, but one in which the final outcome is very much determined by the course of action the players decide on, with a slight random factor introduced by the dice. But I think that to have an interesting outcome and an interesting journey to get there, the adventure needs to begin with an interesting setup. Which in practice means that I begin an adventure by establishing that the village is getting attacked by a monster at night. The monster has a reason and the monster has a plan, and if the players don’t do anything to prevent it, the monster will continue with its plan. But what the players decide to do about it is entirely left to them. Theoretically they could decide they don’t care and leave the village to its fate, but in practicae all players understand that preparing an adventure takes work and time, and if they like their GM, they always go and investigate and won’t decide that todays game only last 5 minutes and go home. Which is why I always play with people I already know or who are good friends of my good friends. Among friends, people are normally happy to run with whatever makes the whole group happy, even if it’s not 100% exactly what their personal preference would be.
As a GM, I don’t have any predetermined outcome planned. I assume that the players will be able to solve the mystery and put an end to the monster attacks, saving most of the villagers. But if for some reason the players make descision or false conclusions that result in their failure, then that is what is going to happen. If the players have a good idea to become friends with the monster and help it in destroying the village, I am not going to stop them. If they screw up so badly that they can only watch the village go down in flame, then that’s how it will end. I think I am fairly laid back and don’t make it very hard for the players to accomplish the outcome I assume to be the most likely they want to get. Which in this scenario would be finding the monster and stopping it before it causes too much more damage to the village. But I think it’s important that other outcomes are possible and that the players know that other outcomes are possible. There needs to be a chance that the players will be unsuccessful and there needs to be an option for them to decide on a different goal than the one I assumed they would want to pursue. Because I believe that this is what ultimately gives meaning to the adventure and to all of the players actions. If events A, B, C, and D will happen no matter what the players do, then there will be no feeling of accomplishment at the end, and no feeling of urgency or suspense during the adventure. So if the players come up with a plan that goes straight from B to D while completely avoiding C, then I think a GM should still go with that. (Which is why I am no fan of the adventures published for D&D 3rd edition and Pathfinder, as those usually can not progress unless the players are going through each of those events in the predetermined order.)
But sometimes C is really, really cool…
So perhaps even though the players came up with a plan to go straight from B to D, that makes taking a stop at C unnecessary and a bit illogical, the GM could still secretly have them go though C anyway? A tempting idea. But this is the path that leads down to the dreaded railroad! Most usually people are using the term railroading for any kind of action by the GM that makes sure that events happen in the way the GM planned for them to happen. Which is almost universally, and rightfully, regarded as bad, bad, bad! Becauser players will notice. Perhaps not at first, but eventually they will notice that none of their clever ideas to get an advantage that surprises the GM ever work out. But those moments when the players manage to be clever and outsmart the villains are often the best thing about roleplaying games. Those are the huge difference between roleplaying game and videogames, and the one area in which videogames just can’t compete in even the slightest way.
But this event C is still really, really cool! And if the players don’t encounter it, nothing that follows will make much sense and the rest of the adventure be just boring! They might think they finished their lame job of killing the goblin kind and return home, and never learn that the goblins are really just one group in a huge army lead by a vampire lord. It would be like ending a James Bond movie after the opening scene, or the Hobbits turning around and go back home after delivering the ring to Rivendell.
When this happens, there is one tool that every GM should at least now about, even if you never get to actually use it: Schrödinger’s Gun.
I think almost everyone has heard of Schrödinger’s Cat, which is a hypothetical cat put into a sealed box with a machine that can detect radioactive decay in a small sample of a radioactive element, and if it does so it releases a deadly poison that kills the cat. And I think almost no one who has heard of actually understands it. Erwin Schrödingers point was not that the cat is both alive and dead at the same time until someone opens the box and looks inside. His point was that this idea is ridiculous and the cat is either dead or alive, regardless of whether someone looks inside or not. The experiment would be the same way if you just check the machine if it has detected radioactive decay or not. The cat and the poison are only there to emphasise how stupid the idea is that you have to check it before it becomes reality. (Because neither a cat, nor a geiger counter are subatomic particles, and quantum mechanics only apply to subatomic particles, not to things made of atoms.*) However, you will only know whether the cat is dead or alive when you look. And that’s the part people are usually refering to when calling something Schrödinger’s Thingy.
The other thing I am talking about here is Chekhov’s Gun, a principle about writing stage plays by the Russian playwright Anton Chekhov, which says that “If you have a gun on the stage in the first act, it better gets fired in the third!” Which is really great advice. You don’t have to literally fire the gun, but when you introduce an element to a story that to the audience looks like it will become important, then it should actually become important. Say you introduce a treasure map and the hero goes on an adventure to find that treasure, but in the end he ends up in a completely different place, doing something much more important than searching for that treasure, and in fact he never actually finds it. That’s still okay; the treasure map did fulfil its purpose. It did not lead the hero to the treasure, but it did lead him to somewhere he wouldn’t normally have gone to.
I don’t know who had the idea, but someone figured out that you can combine those two concepts into one. Schrödinger’s Gun! It’s a gun that is introduced and that will be fired, but you don’t know yet when and by whom. Or in other words, nothing in the fictional world is fixed until the characters (and audience) have interacted with it or gained any knowledge about it. A cheap example would be that the players have three doors to pick from and only one will have the treasure they look for, while the other two are traps. In theory, the GM could decide that the treasure will be in the last one the players open, after they already opened two other ones and got hit by the trap. After all, the player would never know if they just had back luck or if the outcome was determined by the GM before they did anything. If you wanted to, you could be using Schrödinger’s gun all the time, without the players ever being the wiser. It’s not like you invalidated any idea or action the players have made. But when it becomes a pattern, the players will recognize it. Even if they don’t do it consciously, they will assume that all adventures are always so easy that they can never fail, or if the GM is kind of an ass, so hard that they will never win despite their best efforts. If the players guess lucky, allow them to feel lucky. If they come up with a plan to avoid a difficult fight, let them avoid that fight. If they are only lucky or clever when it suits the GM, they will notice. If you want to make a long face, because they got around a hard fight you planned for by outsmarting you or just making a ridiculously lucky gamble, let them see your long face. Because they will feel very happy and rewarding when they see their plan worked or their gamble worked off. So I say, use Schrödinger’s gun very sparingly. Let the players be lucky and clever when they are, and also let them be unlucky and stupid when they are. If someone tries a really stupid jump that would make the character splash upon the rocks, don’t handwave it and tell him he made the jump anyway. Let him fall to the bottom of the cliff and crash onto the rocks. But if you’re a mercyful GM, you might still let him have -5 hit points and force the other players to start a half hour rescue opperation during which the villain they were chasing gets away. You can lessen the blow, but it should still hurt.
Okay, but now to my actual point I want to make! (Finally!) I think sometimes it is okay to use the device of Schrödinger’s Gun without it being railroading. As I see it, it’s not automatically railroading when the GM makes something happen in the way he prefers it to happen. Assuming you don’t make it a habit, making little changes behind the scenes only becomes railroading when it invalidates the actions and plans of the players. Let’s go back to the three chests with the single treasure and two traps. If the players just open chests at random, there is no planning or descision on the players part that you would invalidate by moving the treasure around. However, if the players decide they want to look for clues, and perhaps examine the chests to see if there are flakes of rust on the ground below the hinges of one of them (indicating it gets opened and closed more often than the others), then this should work out for them. You might not have thought of it that there might be a way to tell which chests get used the most while the others are never touched, but now one of the players outsmarted you. You could tell him that no, there doesn’t seem to be any difference to any of the three chests and they still have to pick one at random. But I think you actually should tell the player that indeed, he finds the clue he was looking for, and you know add that bit of rust to the game world, even though it had not been part of your preparation. And why should you? Because it teaches the players that they should come up with creative ideas. It tells them that by being clever, they can outsmart the GM. It makes it clear that their thinking and descision does matter. And as I said before, I think this is the most important and best thing about roleplaying games. Planning and making descisions matter!
But let’s consider another scenario. The players have tracked the kidnapped princess to a castle. The GM know that the princess is in the room at the top of the highest tower, and that there is a black knight waiting in the main courtyard. He also expects that the players will go through the main gate, fight (and defeat) the black knight, and then get to the princess in the tower. Which is perfectly reasonable to assume. But the players think it’s obviously a trap (which it is) and instead of going through the main gate, they wait until nightfall and call their eagle friends to carry them to the highest tower and then work their way down the tower through the castle. If they do that, they would most likely skip the entire castle and never run into the black knight, making all the preparation for naught. A bad GM would simply say “as you wait for nightfall, a dragon comes flying to the castle and sits down on the highest tower to sleep”.
With Schrödingers gun, you can do better. If the players decided to enter through the top of the tower on a whim with no reason to suspect the princess will be there, the GM could simply move her to a different location in the dungeons below the castle. They will still have to go through the courtyard and face the black knight, as the GM had planned, and it doesn’t invalidate the players plan. Landing on the tower works, just as they intended. And you could even reward them by letting them reach the courtyard from the other side, seeing the black knight and his henchmen watching the main gate from their hiding points. But since the players did not take the gate but came in through the tower, they are now the ones surprising the knight and his men from behind, not the other way round. Everyone is happy.
But perhaps the player did sit down to consider carefully where the princess might be kept and after thinking it through really deeply for a while, they come to the conclusion that she won’t be in the dungeon, for whatever reason. The reason they came up with might still be completely wrong and it would work out perfectly for the black knight to keep her in the dungeon. But in truth, even though their reasoning was faulty, they did come to the correct conclusion. Or they might expect an ambush at the main gate and want to avoid it. You could still move the princess out of the tower and into the dungeon, but this would deprive the players from their reward they have worked for. And I think the rewarding feeling of having been clever or just gambled and beeing lucky is more important than going through all the setpiece encounters the GM had prepared. So there is no ambush at the gate. A small sacrifice for the elation the players will feel when their clever plan works out for them. However, if they went with the tower because they thought they would find the princess there or because they expected an ambush, it doesn’t mean they don’t have to encounter the black knight. If none of their planning involved avoiding the black knight, then he can still be a Schrödingers gun. Keep the princess in the tower room, but move the black knight from the courtyard to the door of the princess chamber. Bam! Still a big part of your planning salvaged.
The principle of Schrödinger’s gun says that nothing is fixed until the players interact with it. But I think that it is important that they don’t have to encounter it or get any actual information about it. Simply starting to speculate about it should be treated as interaction by the GM. Once the players start to plan their actions on assumptions about a still unkown object or character, you should make no changes to said object that would invalidate their planning. Even if it really was just a lucky guess. You can still do it. but my advice would be to only do it when it seems to be absolutely necessary and completely unavoidable. Lucky guesses and outsmarting the GM are a very important part of RPGs, if not even the most important. Don’t remove this part of the game to have the story play out the way you want it to play out.
I feel more lenient when it comes to the reverse: When players come up with a completely wrong, but smart sounding reason while something should be in a certain way, and it actually seems better than what you had planned, I think there is little harm in essentially giving the player a reward they have not earned. Why ever go with something lame if you still have the option to replace it with something much better? Not only will the players feel smart about having made the right conclusion, they will also think that you are smart for having come up with that idea. ;)
*Actually, there is some debate about whether a large object made of atoms like a cat could be in superposition for as long as it does not have any kind of interaction with the rest of the entire universe, but this would be the case only for an infinitely small amount of time. However, since this is not an article on quantum mechanics and superposition, this really isn’t important in any way for this topic.