Funicular Simulator 2021: Post-mortem

Here’s a post-mortem for Funicular Simulator 2021. It’s a little long, but we hope it’s interesting.



What we did: Back in March, we decided to build a game together. We weren’t sure we’d enter IFComp, but we were absolutely sure the game would be no more than five minutes long. That way, we figured, it would only take a few days to make.

Did it work? It did not. Funicular Simulator takes half an hour to play, and we spent weeks of full-time work making it. There are a few reasons why the scope ballooned:

  • Themes and genre: We quickly realised the game would need to be more than five minutes long. We wanted players to make meaningful connections with NPCs, ascribe meaning to inscrutable things, and decide how they wanted to spend eternity.
  • Collaboration: In most of our brainstorming sessions, one of us would utter the cursèd phrase: “right, we definitely shouldn’t do this, but what if…” and then the other one would get excited about it.
  • Feelings: Quite early on, we realised we were creating something we really loved, and we wanted to make it as polished as possible. This is about the point we decided to commit to IFComp.

Fortunately, we started making Funicular Simulator 2021 months ahead of the IFcomp deadline (perhaps subconsciously self-aware enough to know that we’d go massively over the original scope), and resigned ourselves to how much work it was going to take. We don’t have any regrets because we made a game that gave its characters and themes space.



What we did: We actually developed the game around the title rather than vice versa. Despite that, the name doesn’t reflect the gameplay, so as the IFComp deadline approached, we rigorously picked it apart and tried to come up with something more descriptive. However, nothing was as fun or intriguing as “Funicular Simulator 2021”. As first-time entrants, we decided to use a title that grabbed people’s attention, even if that came at the expense of describing the game.

Did it work? The title was divisive – some reviewers thought it was intriguing, and it drew them to the game, while others reasonably thought it was too incongruous. Neither reaction was a surprise to us, and we’d choose the title again.



What we did: Design, theme and genre: As mentioned above, we built the game around the title. We brainstormed “funicular”, and most of our associations were about novelty and recreation: riding a funicular to get to a tourist site or a spectacular view. We therefore put something of interest at the summit of our mountain – initially ruins, rather than crystals. Then we asked ourselves “ok, but why would people visit it?”, and came up with so many answers that we realised the game would be about the different ways people ascribe meaning to things.

It proved a really interesting way to approach game design. In a lot of AAA games, interpretation doesn’t feature much – what matters are the player’s actions. In more ambitious IF games, players might take actions and interpret the world around them, but for Funicular Simulator 2021, we knew the player wouldn’t be taking many actions because they’re just passengers. Consequently, in order to give them meaningful choices, we let them define their character through their reactions to the other characters, and their interpretations of the world around them: the mountain, the crystals, and the aurora.

Because we wanted to honour the player’s interpretation of the world, there is a fair amount of ambiguity written into the game. We referred to this as “dreaminess”, and it guided a lot of our design decisions.

Defining the genre was something we only had to do when we finally submitted the game to IFComp, and it was a struggle. We settled on “dreamy mystery”, because there is a mystery at the heart of the game – what is up with this mountain? – but it’s not actually one players can definitively solve. Funicular Simulator 2021 is also a dating sim and a time loop game, but both of those elements felt, to us, more like mechanical and structural necessities rather than descriptions of what the game is really about.

Did it work? For the most part, yes. Several reviewers appreciated the meditative nature of the game, and said it made them think about how we ascribe meaning to things.

However, some players had an understandable urge to solve the “mystery”, and were frustrated that there was no definitive explanation for the mystical events, and that they couldn’t resolve the time loop. This is a fundamental tension at the core of Funicular Simulator 2021. The game is by no means a failure, but there is dissonance between gameplay and theme.

What we did – Dating sim: We decided Funicular Simulator 2021 would be a dating sim for a number of reasons:

  • We like dating sims and play a lot of them; we felt equipped to write one.
  • There’s something romantic about funiculars and their associations with holidays and picturesque European towns.
  • Practically, a dating sim would be relatively easy for two people to work on. We could each write two characters and work on them separately.

We were clear that the “dates” would be about connection rather than romance. The player can romance two of the characters, but a romantic connection isn’t necessary to connect with either of them, or have an ending where they end up “with” that character. Ideologically, we also wanted to give the NPCs agency, and not treat them as prizes. Broadening the concept of connection (and not making romance an option for two of the four NPCs) was the best way of doing this.

Did it work? Yes and no. “Dating sim” provided a really solid structure with which we could explore the theme of how everyone interprets the world differently. Also, players had pleasingly strong responses to the various characters, and could express those reactions through gameplay because they were free to either forge connections or back off without missing out on content. They can also connect with multiple NPCs without diminishing the value of the other relationships.

However each date only lasts for about 10 “real” minutes (though it’s implied that more fictional time passes). That’s not long to build a connection with someone, especially given how much significance each date can potentially have to the NPCs.

Moreover, there’s a sense in which the “mystery” aspect of the game is at odds with it being a dating sim. There’s really only one way players can learn about the aurora and the crystals, and that’s by talking to the NPCs. This means that they’re arguably sources of information, rather than real companions. While we succeeded in not making the NPCs “prizes” in and of themselves, they definitely have a meta gameplay/expositional value that possibly detracts from their agency.

What we did – Time loop: We decided on a loop and grow structure early on, mainly as a way to encourage players to ride the funicular multiple times in order to go on all the dates. Time travel only became an explicit theme when we were brainstorming what phenomenon the scientist character, Meena, should be studying. Since the game already looped round, making the time loop part of the fiction made sense.

To make the time loop narratively interesting, we decided that aspects of the story should subtly change in subsequent loops, and that the player should be able to make reference to experiences they’d had in earlier loops. However, it was very difficult to make this satisfying without having development spin (further) out of control. We tried to strike a balance between making players feel as if the world and their character’s perception of it was subtly changing with each loop without giving ourselves huge amounts of additional work to do. For example, did we want to give each date wildly different outcomes for each loop, given that the chances of any one player playing each date multiple times was relatively small?

The other tricky thing was deciding on the fictional reason behind the time loop – we didn’t want to overstep any of the NPC’s (or the player’s) interpretations of how this worked. In the end, we gave each NPC a different explanation for it, which contributes to the theme of interpretation.

We made what we thought were enough tweaks to the time loop so that players felt like the world was subtly evolving. We also leaned hard on the “dreamy” aesthetic by implying that the player character only had a vague sense that they’d ridden the funicular before – we had a line at the beginning of each loop which appeared to make it explicit, only for the text to change to something more non-committal when players actually clicked on it.

Did it work? Again, yes and no.

On a practical level, we struck a good balance between evolving the gameplay with each loop, and keeping development manageable.

In terms of execution, we were mostly successful. We hoped the dreaminess would help players not be too disappointed that their character couldn’t say, unequivocally, “we’re in a time loop!”, but this didn’t satisfy all players. This is partly because we’d created dissonance between the player (who could remember that they’d ridden the funicular before), and their character (who couldn’t). In some IF games, this dissonance is part of the narrative, but in Funicular Simulator 2021 we were trying to make the player character a blank avatar through which the player could express themselves. Consequently, not letting them express their fundamental understanding of the gameworld was frustrating to some players.

Overall though, the time loop structure was a good way of tying together the practical problems of the game – how to encourage players to play through multiple dates – with the thematic ones. We could maybe have alleviated the dissonance by giving the player a way to address the time loop directly, but that wouldn’t have resolved the central tension of the game.



What we did – organisation: The starting point for each character was “how would this person interpret the mountain?” Once we’d come up with four answers we really liked (scientific, aesthetic, religious, ALIENS), we took two each and built characters, then pitched them to each other. Then we workshopped them together in order to nail down their desires, backstory, aesthetic, personality and star sign, before separating them again to actually build the dates. For a long time we worked on separate versions of the game that we kept swapping so we could edit and query each other’s work.

Did it work? Yes, it was the best aspect of co-writing the game. All our best ideas came from either the workshopping sessions, or by looking at each other’s drafts and adding to them (“yes, and…”).

Coming up with two characters each meant we could go down rabbit holes that interested us about each worldview etc, and this gave each character a strong core. Our process of frequent swapping meant that we were quite rigorous – it’s much easier to assess someone else’s work than your own. It also meant that the overall result felt like a cohesive whole, rather than 4 separate dates written by 2 different people.

What we did – consistency: Keeping the world consistent was a faff – we had to make sure that each thing happened at the same point in each date. When we started writing the game, we actually had a lot more elements on the mountain (and a corkboard full of post-its about them), but we cut these right down to what was most important: the first appearance of the crystals, the aurora, the summit.

Did it work? We think so – or at least none of the reviewers caught any inconsistencies. A huge risk of the time loop structure would have been for us to miss something (e.g. have the crystals appear much later in one date than others) and have players think it part of the time loop, rather than an oversight on our part. We managed not to do this.

What we did – endings: We wanted to give each NPC a unique ending that validated their worldview, without negating the others in the process. The writing was impressionistic and surreal, and hinted at the truth of each NPC’s worldview while still giving the player room to interpret what they were experiencing.

Did it work? The endings have the same tension as the rest of the game in that while we generally encouraged the player’s own interpretations of the world, they can only choose endings that represent the NPCs’ understanding of events (unless they don’t choose to spend an eternity with an NPC, in which case they become the funicular). Also, some reviewers felt the writing was too oblique.

However, the endings were one of our favourite bits of the game. They had to strike a difficult balance between feeling definitive and satisfying, while still giving players room for their own interpretations. Because of that, no one design choice would have worked for everyone… and the choice we went for worked for us.



What we did – Twine: We decided to use Twine because of the potential languages/platforms, we had the most combined experience in it.

Did it work: Quite early on, we realised that much of the game might actually have been easier to write in ink, because it’s more streamlined and a little easier for two people to collaborate on. However, we made the most of Twine’s tools (see below).

What we did – minigames: Each date has a piece of unique interaction in it: palm reading, a flower game, a snack-based Q&A, and a device that measures time-travel particles.

Did it work? Broadly yes – it was texturally interesting to have a unique interaction in each date so players didn’t get bored of just clicking links. The least successful was probably the snack-based Q&A in Ray’s date, which felt a bit lawn-mowery (the flower game in Sofia’s date is also lawn-mowery, but the design did a better job of concealing it).

The most interesting was Meena’s chronon device, but this was also the most challenging to implement: it is slightly laggy because it uses a bit of a Twine hack in order to work. We spent a long time finding the balance between making it a puzzle and preventing players getting stuck or frustrated, but there’s probably even more we could have done to avoid frustration.



What we did – design: Neither of us has much design experience (and we wanted the writing to stand out), so we took a “less is more” approach. We chose a pastel palette to convey the dreamy, romantic quality of the game, then used CSS gradients and simple animations to simulate the progress of time from early evening to sunset, and eventually the appearance of the aurora. This helped to signal the time loop: the abrupt transition from the night sky to light blue let the player know that something uncanny had happened.

Did it work? Yes, the design was one of the most universally praised elements of the game. A few rather generously suggested that the game is part of a larger movement for Twine games away from maximalist graphics to a more understated style, but this wasn’t intentional on our part.

What we did – cover image: Again, neither of us is a designer, but we’re both fans of a good Photoshop job. We brainstormed what we wanted the cover art to convey: what a funicular is (it’s not obvious!), the fantastical elements of the game, the dreamy aesthetic, and the fact that it is partly a dating sim.

The funicular in the cover art is actually a Lisbon tram, recoloured from bright yellow and white, to blue and brass.

Did it work? Yes, we were both incredibly pleased with the final result. It’s polished, and easy to parse while still being quite complex – the geode is actually reflected in the windows of the funicular carriage.

It also conveys the fantastical nature of the game, and the dreaminess works as a foil to the tongue-in-cheek title – it lets players know that there’s more to this game than meets the eye. The one element that’s arguably missing is a sign that it is a dating sim: while the font choice and romantic glow of the crystals subtly hints at this aspect of the game, we eventually relegated it in favour of playing up the fantasy and mystery, and ultimately creating an image that we loved aesthetically.


Thank you again to the organisers, to everyone who played and rated, and to all the people who wrote such thoughtful reviews. Making this game and entering it in IFComp has been an overwhelmingly positive experience, and we really appreciate everyone’s hard work.

Mary and Tom
:roller_coaster: :mountain_snow:


Thanks for sharing all this! I was actually about to revisit Funicular Simulator for an IFDB review & this is extra timely for me to see and take into account.

More generally, anyone who feels it’s too late to write up an IFComp postmortem? Trust me, it isn’t. You aren’t too late, you didn’t miss deadlines, etc.

There are a lot of people who would just have liked to hear more about entry X, or maybe didn’t have any entry they were specifically following, but they become interested with something to read. Don’t worry if it is too long or short, either.


Thank you!

And I totally agree – I’m always interested in post-mortems, no matter when they come out. (I also think they can be really helpful to write even if you don’t post them – we were able to articulate and understand a lot of things as a result of writing the above)