First of all, the post-comp release of Winter-Over is now available here. If you’re not interested to hear us rambling about our last 4.5 years of IF-writing, that’s all you really need to know! However, if that does (inexplicably) sound fascinating to you, read on.
Pro-logue
The origin of the idea; Encorm's entry into IF; early attempts to write the game
EJ: The seed of the idea that would become Winter-Over was first planted quite a while ago, while I was writing a thesis on H. P. Lovecraft’s At the Mountains of Madness and the history of Antarctic exploration. Modern Antarctica was pretty much irrelevant to the thesis, but I’ve never met a research rabbit hole I didn’t like (especially when procrastinating), so I ended up reading a lot about it and thinking, “Wow, an Antarctic research station in the winter would be a great setting for a closed-circle murder mystery.”
But although I was starting to get into IF writing at about this time, I knew I didn’t have the technical chops to pull this off (especially since I was using Inform 7 back then and I always knew this was going to have to be an NPC-heavy, conversation-focused game), so I put it on the “someday” shelf and proceeded to leave it there for almost ten years. Of course, partly this was because from 2015 through 2019 I didn’t do any IF writing at all, because I had a very draining job that pretty much killed my creative writing drive.
But I got a new job in late 2019 and decided it was time for me to dive back into IF—especially once the pandemic hit and my other major time-consuming hobby, theatre, was off the table for a while. I decided I was going to make a game for IFComp 2020, and I pulled the old “Antarctic murder mystery” idea off the shelf because I thought it was the kind of thing that would go over well with the IFComp audience.
During my IF-writing hiatus (during which I remained an active player and reviewer) I had started dating Encorm, and I dragged her into the hobby during IFComp 2015 by making her play Birdland. She strongly encouraged me to get back into writing and was heavily involved in brainstorming characters and plot elements for Winter-Over, although at the time the plan was for me to be the sole actual author because she had no writing or coding experience.
NC: Despite having no writing experience at this point and in fact being terrified by the mere idea of creative writing, I was (and still am) also a sucker for a good research rabbit hole. Between that and being critically bored at this point in 2020 I was happy to dive headfirst into brainstorming! I was more or less leading the way on this, actually, not because I was better at it or had more enthusiasm for the project but because EJ was having anxiety about getting back into game dev. I had no reputation to uphold yet so I was free to blithely charge ahead!
We didn’t have much of a plot outline at this point beyond “someone dies and you have to solve it by talking to the other denizens of the base”, so we decided to make up characters and then see if anything followed naturally. I had read about a way to create lots of distinct characters quickly by pulling Tarot cards so that’s what we did – not by using any particular spread, but by pulling one Major and Minor arcana card per character, looking up the meanings of each, and spitballing what kind of person it would describe. Several uninspiring combinations got discarded in the process, but before long we had most of our cast created!
I say most because… none of these nascent characters actually stood out as a villain. Proto-Daniel jumped out very early as someone who would give a lot of people reasons to murder him, and we could generally sort the rest of the cast into suspects and side characters, but we were kind of at an impasse here. At least, until I reread the characteristics we’d picked for Jack, which emphasized that he was a toady and a suck-up. A person like that could be the murderer, maybe, but only if someone else made them do it… and there we had our twist! A few more cards were pulled for Bob, and off we went. With that structure in place the outline came soon after, and by this point (less than a month before the submission deadline) we also had a first draft of available evidence, a full set of character schedules, and most importantly – no actual code written.
EJ: Of course, this was a very ambitious game idea, and I had never done anything remotely that ambitious in Twine before, and indeed pre-hiatus I was still using Twine 1, so a lot had changed in the intervening time and I had to relearn a lot of stuff. Also, we didn’t really manage our time well, and I had undiagnosed and untreated chronic health issues which were affecting my ability to get work done on the game. So ultimately all that was written of Winter-Over at the time was a version of the prologue, after which we realized that there was no way we were getting it done by the deadline and put it on hold. I spent the last couple weeks before the deadline writing What the Bus? (which Encorm also helped brainstorm and came up with the title for) and submitted that instead. (It was not the kind of game that would do well in IFComp and I knew that, but I had told myself I was going to make my big return to writing and it would have felt terrible to then not do it. The game indeed did not do well, but there were aspects of the experience that were positive and I’m ultimately glad I submitted it.)
I then decided to enter a more substantial game into ECTOCOMP, which I did manage to pull off. That was Social Lycanthropy Disorder, which was by far the most successful game I had ever made. But because I’m a glass-half-empty kind of person, I homed in on a small handful of complaints and criticisms, including that some people wished the protagonist were less strongly characterized (or at least wished that she had several possible characterizations to choose from, Choice of Games style) and that some people didn’t like the conversation focus and would have preferred more traditional puzzles. (In retrospect, this was less because people don’t like conversation and more because SLD draws heavily on dating sims and visual novels, mechanics-wise, and I didn’t do a good job elucidating those mechanics for an audience not familiar with them, but I didn’t realize that at the time.) I made note of both of these things for my future Game That Was Supposed to Do Well in IFComp, and in particular decided I needed to do something to hone my non-conversation-based puzzle design skills before IFComp rolled around again.
NC: And that’s how Lady Thalia happened! And it was much, much better received than we expected. Which, frankly, was really important – I don’t think we needed that level of praise to keep going, but the unambiguously positive reception was a huge morale boost. (And this was my first attempt actually writing IF, and the fact that nobody felt that my portions stood out was fantastic.) Even before Spring Thing 2021 was over we were really happy with the reception, and super hyped to make our next game!
And then I had a very bad month, capped off by the sudden death of a family member. (I’ll talk about it in a separate post because it is relevant to Winter-Over in the end besides the schedule disruption but is otherwise too personal and depressing for the main post). It goes without saying that this threw all our plans into chaos. And by the time I got my feet back under me (and EJ didn’t need to support me so much) it was almost August and we definitely did not have time to complete Winter-Over before IFComp 2021. So we pivoted instead to Starbreakers, a game born out of our love for escape rooms and puzzle hunts but also out of a desire to make a game that we thought the general IFComp audience would like. Hard emphasis on “thought” here, though, because it didn’t do very well. Part of this was down to over-scoping it and not having time for more than the most basic testing, which is a lesson we did learn, but the other part of this is that we are actually pretty bad at telling what people “want”, in a broad sense.
EJ: So then 2022 was going to be the year that we were going to finish Winter-Over, and we got some preliminary planning done; we sat down and thought about the gameplay based on what we had learned from our efforts so far. In particular, between the failure of Starbreakers and the unexpected success of Lady Thalia (1 and 2, by that point), we ended up moving away from puzzles a little and leaning more into the character- and conversation-based elements. It’s at this point we came up with the relationship tracking and the mechanic of having activities that would both reduce your stress and improve your relationship with a character of your choosing. We took another look at the evidence and the character schedules and made some tweaks to those too, and took a stab at mapping out what information was gated behind what obstacles.
All in all, a good start! We were excited to see where this new direction led us! And then my father needed emergency heart surgery. He’s fine now, but the whole thing was immensely stressful, and I ended up having to stay with him for a while to help out while he was recovering, and this pretty thoroughly derailed our IFComp plans.
In 2023 we put out another Lady Thalia game for Spring Thing and then naturally turned our thoughts to IFComp… but of course you know what’s coming at this point. In this case Encorm had been on a stressful project at work and was having issues with burnout, and when we tried to start working on the game it rapidly became clear that she was struggling with focus and motivation. I could have tried to get Winter-Over together on my own, as had originally been the plan, but at this point we’d been full coauthors on the second and third Lady Thalia games, and Encorm had been so involved in all the plotting and planning for Winter-Over, and it just didn’t feel right for me to solo it anymore.
Game-dev
The development process for IFComp 2024
NC: So after our attempt at entering IFComp 2023 fizzled out, we decided that 2024 was going to be IT. This was going to be OUR YEAR, and we were finally going to make this game! Of course, there’s no such thing as a perfect year and we had a plethora of other commitments. We both have family in far-flung corners of the US (all of whom want to see us for every holiday), I was a bridesmaid in a friend’s wedding, EJ and I both have multiple other hobbies to attend to…
And if that wasn’t enough, we were planning a wedding. For ourselves. Scheduled partway through the IFComp judging period. (Let it never be said that we lack dedication to IF!)
So we decided to skip Spring Thing and kick Lady Thalia 4 down the road by a year in order to get a head start on Winter-Over. And while we didn’t quite manage to start writing before the Thing ended, in retrospect there’s still no way we could have done both without hardcore burning out. So for once we made a good decision! And frankly we did end up burning out anyway, which is why it took us over a month to write this post-mortem. Happy but unmarried IF couples: please learn from our mistakes and space the two things out more.
EJ: The IFComp deadline being earlier this year felt a little stressful in that we had less time, but on the other hand, it was over a month before the wedding rather than a few weeks. This meant we had to finish before the point where we truly got sucked into the wedding planning vortex, and in retrospect if the deadline had been later I don’t think we would have made it.
While we had a lot going on, we at least didn’t have any unexpected personal crises this time, so we finally were able to make headway. Once we actually got started, we did encounter a few unexpected flies in the ointment, though.
One issue we hadn’t foreseen is that the game was more emotionally difficult to write than our other big projects; maybe some people can write a protagonist going through something terrible without any emotional spillover, but we couldn’t, and it limited how much writing we could do in a single session compared to our previous projects.
Another is that the game is significantly less linear than any of our/my previous projects (except Social Lycanthropy Disorder, which I wrote in a deadline-induced fugue state that I barely remember). It turns out that when I have a bunch of different things that I could work on in any order, rather than having an obvious chronological next scene to go to, I get choice paralysis and just don’t get anything done. I wasted a lot of time by being too embarrassed about this to ask Encorm for help prioritizing, which is itself pretty embarrassing.
So given these two things, we didn’t get things done as far before the deadline as we had really wanted to, but we still did have time for two rounds of playtesting! Which is a first for us.
NC: Luckily in the four(!!!) years since we started Winter-Over I had gotten a bunch of project management experience at my day job, so once EJ finally got up the courage to ask for help ( 
) things progressed pretty smoothly. The game was technically feature-complete by the end of July but was missing a lot of important scenes, but we had a pre-planned vacation at that point so we took what we had and sent it out for our first round of playtesting. The biggest thing we learned from Starbreakers is that we are absolutely crap judges when it comes to difficulty (Lady Thalia we got right mostly by accident), and since it was relatively nonlinear we were worried about the pacing and time limit as well.
Of course, the nature of playtesting means that not only were we only partially correct about what needed attention, our playtesters found a whole bunch of other stuff that needed work as well. The notes system was well received but was broken up into two sections based on playtester feedback here, and almost everything we had thought players might write down (such as character schedules) was added to those sections. The pacing of the main game was surprisingly… fine??? (per majority consensus), but it was also agreed that the prologue was wayyyyyy too long.
(As a side note, this is about when Draconis first approached us about participating in their planned meta-puzzle. Our game was too planned out to introduce new puzzles into our game and the nature of Winter-Over didn’t really gel with a meta-puzzle of our own anyway, but we hadn’t written anything for the rec room yet so hiding clues for Miss Gosling was both easy and very fun! I’m glad we got to participate.)
And then… we caught something on our vacation that knocked us out of commission for a week or two. Oops! But because we’d started so early (by our standards), we could recover and THEN start crunching on the second draft instead of trying to crunch through a stomach bug like we did on Lady Thalia 3. That’s progress, right?
In addition to incorporating all the feedback from the first round of testing, this is when all the various endings and sabotage incidents were written. The third incident in particular was called out by reviewers as lacking interactivity and generally feeling a bit out of whack with the rest of the game – it was one of the last things written before the second round of playtesting, so that’s why. I haven’t touched it for the current post-comp release but I’m not above tweaking it in the future if I have better ideas.
EJ: To be honest, I actually expected the playtesters to complain about how non-interactive the sabotage incidents were, but they didn’t, so because they didn’t seem to mind (and in the absence of any feedback that would help us figure out what kind of interactivity people would want to see there) we left those scenes as they were.
A lot of the feedback we’d gotten on round 1 involved the gameplay being confusing and underexplained and various things being hard to keep track of, so we made a lot of tweaks related to that. We didn’t want to have too much extradiegetic/obviously “gamey” explanation of mechanics, because we felt like it would be bad for the atmosphere, but in addition to the changes to the notes that Encorm mentioned above, we threw in a bunch of more explicit “maybe you should do X” in the narration, a few instances of characters saying “I don’t know this, but so-and-so might,” and that kind of thing. The playtesters agreed that things were much clearer and we thought, “Great, problem solved!” … More on that later.
In our efforts to nail down the correct length for the prologue, we did end up with a sort of round 2.5 of playtesting for just that section, but otherwise round 2 generated minor tweaks rather than big changes, and our playtesters generally were pretty enthusiastic about the (mostly-)finished product.
IF-Comp
Our experience during the comp and thoughts on placing 10th
NC: Finally, IFComp opened. We made it, right? …Nah, now we had to survive the mortifying ordeal of being known. But our playtesters were really positive about the game, so how bad could it be?
Well, I don’t want to complain about this too much given that we did place highly but honestly the first two weeks were rough. Luck of the draw meant the first few reviews that rolled in were pretty middling – not bad, per se, but certainly less positive on the game than what we’d heard from playtesters. Eventually this evened out and the more positive reviews rolled in, but we were still caught off guard by the breadth of opinions on the game. Some of that is probably due to the game’s complexity? Previous stuff we’ve written isn’t simple, per se, but had a lot less going on so people generally either liked what we were doing and had suggestions on how to do it better, or they didn’t like what we were doing and it wasn’t for them. Meanwhile, every reviewer’s experience with Winter-Over seemed to be different: how much the game engrossed them, whether they liked the writing or not, where they got stuck, and so on and so forth.
EJ: One of the things a lot of people, even ones who mostly liked the game, complained about was, guess what, gameplay elements being confusing, underexplained, and/or hard to keep track of. How??? I wondered. We specifically solicited feedback on this and worked on it and in the end our playtesters agreed it was fixed! … Of course, most of our playtesters had by that time played through the complete game at least once. Lesson learned: if we get this type of feedback on gameplay in the future, we should solicit a few brand-new testers for round 2.
(Plus also, our playtesters were all fans of our/my previous work, so we’d accidentally self-selected for people who weren’t bothered by our usual flaws, presumably including my unreasonable aversion to extradiegetic explanations of game mechanics and my tendency to make direction and feedback a little too ambiguous, which people have complained about in many of my solo works as well. Sorry for continually failing to learn any lessons about that!)
NC: Another thing that came up again and again across reviews was a general lukewarm feeling about the cast. The atmospheric writing was consistently praised, but people also consistently felt that the interactions with the broader station staff felt transactional and generally weren’t as into them as they were with side characters from our previous work. To be egotistical for a minute – we have not one but two XYZZYs for NPC writing, and we hadn’t done anything that different for this game compared to previous, so what gives? Had people mysteriously stopped appreciating our genius, or were we frauds all along and people finally caught on?
But on further consideration, we realized that actually we did do something differently, which was to have a thinly characterized protagonist instead. The magic of good side characters is frequently in their interactions with the protagonist, so if the protagonist is barely there then of course your interactions are going to feel mechanical! If you’re aware of this you can construct them differently so that they stand on their own, but we, uh… weren’t. So changing nothing from Lady Thalia here came back to bite us, because she is a VERY strongly characterized PC and of course her interactions with NPCs will be different! We could have also just chosen to have another strong PC here, and the reason we didn’t is because we assumed back in 2020 that the IFComp audience didn’t want that. Which is wrong. We’ve challenged a lot of our other assumptions from back then about what makes a game popular in the Comp, but that one stuck through the process despite everything else we’ve learned since then. The real moral of the story here, and one that we have theoretically learned three times over, is that at least for us, tailoring our games to what we think the audience wants gives us worse results than just doing what we feel like and polishing it well. Hopefully the third time’s the charm?
And in the end, of course, it turns out that a lot of this was reviewers doing what they do best: taking a good game and analyzing what could have elevated it into a great game. Also what we were doing here was pretty ambitious and definitely opened us up to more avenues of criticism than our previous games! So the critical gauntlet felt pretty harrowing but in the end, we both did well and got a lot of useful feedback to carry forward.
EJ: Of course, this isn’t to say that there’s nothing wrong with the NPC writing in and of itself or that there aren’t too many of them for how short the game is. I think if we’d cut the cast down considerably, it would have made the mystery a little underwhelming and would also have made it harder to suspend disbelief regarding how the station operates, but I do think maybe the game should have been longer to accommodate the cast size. But then again, without punching up either the NPCs or the PC substantially, that would probably just have resulted in another half hour of players going “who are any of these people and why should I care?”
Anyway: I said earlier that this was supposed to be A Game That Would Do Well In IFComp, but by “well” I really mean I was hoping to place 20th and just barely squeak into the awards stream. That already would have been a huge improvement over our previous serious IFComp attempt and a really big accomplishment in general. For a while when the first few reviews were coming in and all of them were like “ehhhh, major parts of this just did not land for me,” I thought that was out of reach, but as the Comp went on and we got some more positive comments I started to feel hopeful about it again. About a week before the voting deadline I actually had a dream that it placed 15th. (This was also like two days before our wedding, which, you’d think I’d be dreaming about that, but….) In the IFComp predictions thread on the forums, a few people mentioned they thought it would place top 10, but I didn’t really believe it was a possibility. I still wasn’t even especially confident that it was going to make it into the top 20 until the moment I got the email inviting me to the top-20 Zoom call. We weren’t able to join because we weren’t at home, but we did watch the stream, without sound, on Encorm’s phone (the autocaptions were mostly okay) and stared in mounting disbelief as places 20 through 11 were announced and none of them were our game. So, uh, shoutout to those people on the forums, who had a more clear-eyed assessment of Winter-Over’s prospects than I did!
However, it was slightly disappointing that my dreams turned out not to be prophetic after all. I really thought perhaps I was getting some mystical revelations there.
Post-Script
Final thoughts
NC: As of this posting we’ve also (finally) released the post-comp version of Winter-Over. We’ve fixed bugs, added some quality of life features, and generally just polished it up a bit. It’s not a major overhaul (if we started that we’d never get anything else done) but we’re not ruling out coming back to it sometime in the future. The most major thing we’ve done is add a 3-day delay to starting the phone puzzle – it was supposed to be both optional and somewhat difficult to finish within the time limit, but we clearly miscalculated how attractive it was and how long it would take to complete. Pushing it back should encourage players to explore the station and get familiar with the mechanics before they start hyperfocusing on this one element.
EJ: We talked a bit about extending the game to give the cast a little more breathing room, but ultimately we didn’t have it in us. I have to admit that at this point I’ve been looking at the game for long enough that I’m sick of it and I can’t tell if anything in it is any good anymore, so at the very least I need a bit of time off before I do anything else with it, I think.
NC: Same same! I have never hated a game more than when I was doing the final playtesting for the post-comp release. Two months off wasn’t nearly enough and I think I at least will need a full year of not looking at this cursed thing to stand it long enough to make substantial changes.
In the meantime, we’re going to start the next Lady Thalia game (targeting Spring Thing ‘25). And then after that… who knows? Winter-Over has been in the background and motivating most of what we’ve done in IF, one way or another, for the last four years. Having it finished is frankly really weird! It’s good to have it done, and it’s great to see how well it did, but in a way it was also our guiding star for a while. Lady Thalia 4 is going to be fun but it’s also kind of going to be “more of the same” (in a good way! hopefully!) and not having a big project to work towards and hone our skills for is new. I’m sure something will come up to fill the void, and I can’t wait to see what it is.
EJ’s personal thoughts
... what it says above
There’s a lot of stuff that went into Winter-Over that was personal for me, but I have a hard time talking about personal stuff, so I’m not going to go as in-depth as Encorm is about to—but I felt like I should say something, because I don’t want to give the impression that Encorm was pouring her heart into this while I was just contributing Neat Antarctica Facts.
I am the eldest child in my family, and there’s a decent-sized age gap between me and my siblings, so I ended up feeling a lot of responsibility for them. I used to routinely have nightmares about them being injured or dying in ways that I could have—should have—prevented. (In real life they’re all alive and reasonably well, to be clear, but of course bad things have happened to them, since that’s just part of life—and in some ways I still feel like I ought to be protecting them from the ills of the world.)
Daniel is also a little bit based on my younger brother as he was in 2020, but it took us so long to complete the game that he’s now doing much better both socially and in terms of generally having his life together. Still, if he ever sees this: hey, bro, I’m sorry for murdering you in a text adventure game.
A lot of the PC’s experience in the game is also drawing on my experiences of having to deal with crises while I was kind of already Not Doing Well beforehand (physically and/or mentally). Of course, having to deal with a crisis then makes everything worse, but I had to keep going because there were things I had to do. Some players expected the PC’s issues with winter-over syndrome to be more of a plot device, and I see how genre conventions set that expectation up, but it honestly did not occur to me while I was writing because I was just thinking about the emotional experience of trying to get shit done in spite of mental illness using you as a chew toy.