Time loops are a relatively established genre of fiction, and while Groundhog Day wasn’t the first, I think it lays down a lot of the foundation for the genre and demonstrates an essentially perfect form of the basic structure. Someone realizes that they wake up in the same time and place no matter what, and this goes on essentially forever, allowing accumulation of knowledge and skills, while also leading to reflection, growth, and change.
Here are the rules and (sometimes) hidden assumptions of the original Groundhog Day time loop:
- The loop starts at the same time every day.
- The loop ends at the same time every day.
- The loop lasts one day.
- The loop ends upon death.
- Phil Conners is the only one “in the loop”.
- The whole world apparently goes through this loop.
- Phil Conners is stuck in the same geographic location of Punxsutawney, Pennsylvania.
That last one is the only one that is a little suspect, because it doesn’t seem to come naturally from the premise, or as a rule that you’d have to have in order to settle some ambiguity. Phil Conners is stuck in Punxsutawney, and by magical means, not allowed to leave, thwarted if he attempts it. We’re only shown a few of his attempts in the movie, but it’s part of the conceit that he’s there for good, so something needs to prevent him from just getting in a car and starting loops with a four hour drive to New York City or an hour and a half drive to Pittsburgh (and no, I don’t think that a big storm is an explanation for why this doesn’t work, though I do love the theming). The movie rightfully glosses over this, and it’s a conceit that most time loop stories don’t include, or include by somehow restricting the protagonist’s movement.
Groundhog Day also lays out some of the basic narrative elements of time loops:
- There’s an initial run through that sets characters and events without any foreknowledge.
- There’s a loop or two where the rules of the loops are established and the protagonist comes to grips with things.
- The protagonist indulges himself in a world without consequences.
- If there’s something the protagonist needs to wants, several loops are spent attempting different strategies and collecting information until he has eventual success.
- The protagonist demonstrates mastery of new skills, knowledge of people, and otherwise has become an imperfect local god.
- The protagonist suffers from ennui.
- The protagonist finds inner calm and peace, or does the “perfect” loop, which brings things to an end.
Not all of those are strictly necessary, but I think in the original movie, they go hand in hand as answers to some of the pressing narrative problems, namely, the lack of time pressure. A lack of time pressure saps conflicts of their tension and urgency, so the lack of time pressure becomes a major source of internal conflict for the protagonist. The time loop strips out purpose and meaning, so in some sense it becomes the antagonist, or at least the catalyst for internal change.
With all that said, what are the interesting ways we can change the parameters? We’ll take the boring ones quickly, and as fast as we can. Many of these have been done before, but I won’t be noting specific examples.
- Longer or Shorter Loops: Groundhog Day lasts a single day, which works well enough, and I think is probably the standard for a reason. A single day has a rhythm to it, waking up, having meals, going to sleep, in a way that two or three days do not. Making the loop shorter can make for a more focused (and likely shorter) story, while making it longer does the opposite.
- Unending Loops: The loops don’t actually end, except on death. Either the story involves a lot of death, or you play it for horror at least once when someone’s long, long life comes to an end and they get sent right back to the start.
- Permanent Death: It’s relatively hard to establish this, but not impossible, and can help serve to raise the stakes of conflicts within the time loop, at the cost of disallowing some of the fun things that time loop stories are known for.
- Conditional Permanent Death: A subset of the above, where you can die in the loop in certain ways and be just fine, but if you die in other specific ways, you’re dead for good. This is a pretty common compromise solution that in theory gets you the best of both worlds.
- Permanent Injury: It’s not your mind that goes back to the start of the loop, but your entire body. If you get injured, that injury will last for as many loops as it takes for it to heal. This is obviously much more restrictive than permanent death, but also easier to communicate to the reader.
- Limited Loops: There aren’t an infinite number of loops, meaning that there are only a set number of resets available to the protagonist. This obviously alleviates some of the problems with tension, but can be extremely artificial.
- Deteriorating Loops: The world is breaking down with each subsequent loop, getting more erratic or with parts of it missing. This is a subset of limited loops, and gives up one of the major features that a time loop provides, but comes with a sense of urgency and dread that a strict numerical limit on the loops does not.
Now for the more complicated things.
Time loop stories have a tendency to stall out once past the initial setup->exploration->exploitation, because there’s little pressing need to do anything (and often, what to do is unclear). The basic narrative structure of time loops, which most of them repeat a few times, is to have try/fail/try/succeed cycles in order to overcome various obstacles. Sometimes this applies to character interactions, like gaining information from someone or figuring out what makes them tick, and sometimes it applies to more straightforward combat, escape, or infiltration scenarios. These work for a little bit, but quickly grow stale, and then it’s time to move on to either another cycle (though there are limits on how many of these try/fail/try/succeed cycles you can do) or something else.
Adding another “looper” in one way or another is a way of injecting something fresh into the story and livening up what might threaten to become stale, and of the time loop stories I’ve read, this seems to be a very common one. There are a few different approaches, but if the narrative point is to kick things into gear when they’re threatening to stall out, the “other” looper can either be newly inducted into the loop, or possibly was there all along, hidden or unnoticed.
The second looper isn’t just interesting for a different perspective, they add something vital: permanence. With a second person in the loop, there can be developments that take place outside of the looper’s head and carry on once the loop is finished. Granted, these are largely matters of relationships, but it’s a breathe of fresh air, and often comes with new information about the nature of the loops, which can help kick the plot into gear.
From a narrative standpoint, there’s less and less utility with each extra person added to the loop, and I think you hit a point of diminishing returns fairly rapidly. A second (or more) looper turns the story into a story about those characters that is much more “normal” and takes place within the backdrop of looping time.
Rather than a single time loop, you can have a series of time loops which take place one after another. The closest analogy is videogames, where you get a checkpoint after defeating a hard boss.
There are two benefits to this. The first is permanence, which a time loop story always eventually craves. The second is opening up new and interesting stories, essentially booting the story structure back to the start and getting fresh setup->exploration->exploitation. I’m not sure how many times you could get away with that, but it seems like the sort of thing that would do well in a web serial but not really anywhere else.
There are a number of different ways that this can be implemented.
- Naïve Progression: Time loops work exactly like a videogame, and there are “predetermined” checkpoints, whether known to the protagonist or not, which, once crossed, make a new start (and end) to the loop. I can’t see this working outside of a serialized context, but within it, it allows things to stay fresh for longer.
- Guided Progression: The time loops are under at least partial control of the protagonist, allowing them to set a new checkpoint for themselves. I think this ends up very similar to naïve progression, but there are a few wrinkles that are different, especially the risk of setting a checkpoint minutes before disaster.
- Multi-track Progression: Rather than a single “save point”, the protagonist has multiple, either limited or unlimited. This means that rather than a strict time loop, you in some ways have several, and can return to “earlier” loops or go to “alternate timelines” or even weirder things. In some ways, I wouldn’t even call it a time loop, but it would share a lot of things in common, especially with regard to the setup->exploration->exploitation and try/fail/try/succeed parts.
- Reversed Progression: I don’t know that I’ve ever seen this, but if we’re getting into esoteric stuff, so be it. Rather than changing the loop start to be later, after some stuff happened, you could change the loop start to be earlier instead. What would this mean? Why would you choose to structure a story this way? What would be the criteria for earlier loops? Honestly, if I were doing it, it would be a “how did things get so screwed up?” type of thing, with more information revealed the earlier you got in the loops and a protagonist who had amnesia, but it’s a kind of out there idea.
- Protagonist Switch Progression: Let’s say you make it through your whole loop, having won the day. Congratulations! It’s someone else’s turn! That new person gets to go through the loops on their own, following all the usual tropes, and perhaps with them not knowing you’re a looper and you not knowing that they used to be a looper. This form of progression has some benefits, namely the change in protagonist to keep things extra fresh.
I’m sure there are a number of variations on this, but I think these are the main ones, and what’s most interesting about them.
There was a proposed variation on loops, whose origin I can’t recall, where you can conceive of each iteration of the timeline as being numbered, and people only having their past memories on specifically numbered loops. For example, Sally’s number is 2, and she only gets memories from loops whose number is divisible by 2. By contrast, Michael’s number is 3, and he only gets memories from loops whose number is divisible by 3, and Trevor’s number is 5, and he only gets memories from loops whose number is divisible by 5. This means that in addition to normal loop stuff, there are ‘coincident’ loops where Sally and Michael are loopers together, or and a few where the three of them can all meet. When not in a coincident loop, it’s as normal, and when Sally is a looper but Michael is not (e.g. loop 4) she interacts with the baseline Michael who has never looped before. Looping Michael does not get these memories.
There are a few interesting things that happen here. One is that looping Sally experiences time at a faster rate than looping Michael, and looping Michael experiences time at a faster rate than looping Trevor. When they all have their first three-way coincident loop, it’ll be Sally’s 15th, Michael’s 10th, and Trevor’s 6th. This puts at least some strain on Sally, but also gives her a significant advantage over the two.
Phil Conners is the only person in the time loop. If you add another person to the loop, then you have a multiperson time loop. But people aren’t the only thing that you can add to a loop, not by a long shot, they’re just the most internally consistent. Instead, you can add things or effects that cross the loop boundary.
The most common of these that I’ve seen are things that directly affect the looper or their mental state, things like mental attacks, soul attacks, or other stuff that aims to be in line with the principle that there is a lone looper. This works well enough, and can introduce some element of danger, but I’ve seen it more than once, and therefore find it boring. Similarly, being able to kill someone so they either die at the start of the loop or vanish from it is interesting, but has been done before.
In a more fantasy work, or if you were doing scifi, you could have specific things (or classes of things) that don’t obey the time loop. Perhaps a certain material is resistant to looping, a large piece of stone or something, and if you move it into the lobby of a building, it will be there when the loop restarts, moved from its original position. There are loads of interesting things you can do with something like this if you’re a looper, because it means you can change the starting conditions of the loop. You could possibly even telefrag someone by putting a bit of this non-looping stone right where you know their head will be at the start of the loop.
Contrarily, you could have something that stays broken as the loops go on. I can see that working well in a more magical-realism kind of loop, something like windows or other glass things staying broken through the loops as part of a deteriorating loop, especially if a deteriorating loop is a mirror for the protagonist’s deteriorating mental state.
Beyond that, I think there are a bunch of ways that you can have something that doesn’t reduce to multiple people in a loop, whether that’s some form of rules-based deterioration or changes to start condition.
In Groundhog Day, the loop seemingly applies to the whole world. Whether you’re in Punxsutawney or Miami, you remember nothing of what happens until the final loop. But what if this weren’t true?
I sometimes think about first lines to novels, and one I’ve had written down for quite a while is, “New York City was stuck on repeat.” I liked this line, or some variation of it, a lot, perhaps because it echoes “Billy Pilgrim had become unstuck from time,” the opening line of Vonnegut’s Slaughterhouse-Five. The idea suggested by that line is that time is repeating within New York City, but nowhere else in the world. You could, obviously, substitute out New York City for whatever location you wanted, I don’t think it’s that important and mostly just a matter of scope.
So how would this work? Time behaves normally outside of the NYC, but within NYC, keeps looping every day. In some ways, you can see this as changing our looper from a single protagonist to the entire rest of the world, and I think that comparison is at least somewhat instructive in terms of what narrative patterns you’d expect to see. Depending on the size of the geographic area, the fact that there’s a loop might become immediately obvious, especially in an area with seasons where the area inside the loop has different weather.
The biggest problem to solve, when deciding on a ruleset, is the boundary between the place inside the loop and the place outside of it. Can people cross? Can weather? Can radio waves? I don’t think there are any obvious answers to this, not even from a “what makes a better story” perspective, but they definitely need to be given some careful thought, because it’s certain that someone would try to find out eventually. You’d also need answers to whether things can enter or leave the loop, and what that means. Perhaps the easiest rule would be that you can see but not touch, in both directions.
(This is another of those ideas that I’ve been noodling for a long time, trying to find the right narrative that fits within the very interesting constraints. There are bits and pieces of scenes in my mind, a rusted sign explaining things to people in the loop, years having passed since anyone went in, perhaps an apocalypse outside while the inner loop is left perfectly fine, the static area of looping time studied as a historical curiosity once no one speaks their language. I don’t know what all those mental images add up to though.)
Outside the Loop
Normally, the protagonist is the looper, or at least one of them. But what if that weren’t the case? This is a story idea that I’ve been trying to crack/write for quite a long time, and I don’t think I’ve ever quite gotten there. There are, I think, two different methods to go about this.
The first configuration is a story that takes place inside a single timeline, from the perspective of someone who is not the looper. They find out that they’re in a loop, realize that this is the looper’s Nth time through somehow (perhaps by observation, perhaps by being directly told, perhaps through something loop proof), and then have to work to stop a looper who has presumably collected lots of skills and information. Stopping the loop depends on the mechanisms that instantiated the loop, which possibly requires some investigation or mystery-solving. What I like most about the story idea is the firm thought in the head of the protagonist that he’s been here before, and failed before, and needs to do the things that he hasn’t thought of before. There are also all kinds of built up unknowns, and it’s the ultimate in media res. I think there’s a good chance that I’ll end up writing at least a short story with this premise at some point in the future.
The second configuration is a story that takes place over many timelines. In the same way that a looper generally gets to know over multiple interactions with them across timelines, we get to know our non-looping protagonist. I think this is a very interesting, but perhaps less compelling story option, and thinking about it, I might actually have seen one or two instances of it before, in a manner of speaking. In some ways, this is just alternate timelines, a la the Gwyneth Paltrow class, Sliding Doors. The aspect of iteration inherent to loops is probably the key thing that differentiates this, and the idea of a looper as the agent of differentiating timelines. I don’t think you’d necessarily need the looper, so long as something differentiated timelines from each other, but at this point in pop culture, loopers are a known quantity, and can serve as either an antagonist or guardian angel or something that can help structure a narrative.
Looking back at some of this stuff, it’s shocking to me to see how little people have been experimenting with the conventions laid out by Groundhog Day. Not so shocking knowing how creative endeavors usually go, and how resistant both publishers and Hollywood are to new, unproven ideas. And not that shocking knowing that you take a risk when messing with what works, or deliberately going against narrative conventions just because you’re bored of doing what’s been done. But I do think that it would be nice to see a wider range of loops, and I think in another ten or twenty years we’ll see time loop stories with a broader set of conventions and narrative lessons learned, which hopefully keep them from being stagnant. I’m hopeful for that, because they really are some of my favorite stories.