After several weeks in development hell, I've finally finished my game! (Full patch notes can be found here.) As such, it is now time to write a postmortem devlog! Needless to say, this might contain a bit of spoilers for the game, so you might wanna turn back and finish it first if you haven't already done so.

As you can probably already guess, I overscoped during MGGJ11. Given the unfavourable circumstances I found myself in at around the jam time, I had wanted to make something relatively simple (so that I could actually complete it), while also trying new things. I saw the "push up" theme, and thought I could perhaps do a stat-raising sim, with push-ups as one of the training activities. Though I've played quite a few stat-raising sims, I've never actually made one myself before this. I do have a fair bit of experience writing visual novels though, and figured it shouldn't be that much harder, right?

I have never been so wrong. While there is some overlap between visual novels and stat-raisers, there was a lot more I had to do; namely, testing a variety of stat builds, as well as writing for both the success and fail scenario every time there is a stat check, which adds up to quite a bit. The whole thing was about ~13k words by the time I had submitted, way more than the 2-3k+ I'm used to writing for my previous jam games, and it wasn't even complete yet. It had only one of the planned routes, as well as a bug in the date/time system, and was way less polished than I'd like. I'm not proud. :(

I had actually updated the game shortly after the jam ended, to hotfix the date/time bug, as well as add Sophia's route (since I happened to be finished with that). But I still had much to do. I still needed to write Justin's route. There's also Teddy's route, which had to be cut from the jam release to meet the deadline. To be honest, I was initially considering just releasing without it, but that didn't feel right. After all, I already had a route which favoured each of the stats other than Temperance, as well as a certain sigil containing the phrases that were the theme / achievement names of each planned route.

Eventually, I finished Teddy's route, but then it felt more polished than the rest of the routes, probably because I had to add a couple of extra scenes to account for the events of the final week. To fix this, I decided to go back and update all routes to have unique lunch break events for each day of the final week, available if you've seen all of their lunch break events in the previous weeks. Hopefully, this should make all routes feel more or less equal. I've also went ahead and added a post-credits scene for not being on any route, so now you're guaranteed a post-credits scene no matter what.

In addition to finishing the rest of the game, there were a few issues that I wanted to address based on comments and feedback received during the jam. First is the chocolate golem, which it seems quite a few players struggled with. I had not intended for it to be too hard, given how early it was in the game, hence the inclusion of partial / saving stat checks, in case one were to fail the initial check. However, it seems it was not enough. In particular, a mixed Prudence / Temperance build could potentially get softlocked - the Prudence choice (ask for help) did not have any partial / saving checks, and the Temperance choice (keep fighting) had only a Fortitude save. Likewise, a mixed build dividing their training amongst all three stats equally might be unable to pass any of the checks.

I didn't want to just nerf the numbers, since I had a formulaic method for determining most of the check requirements (okay, it's really usually just "day # = requirement"), so I decided to rework the fight instead. I've since changed the choice to keep fighting to include an additional Prudence check if the initial Temperance check fails. Passing the aforementioned Prudence check will allow you to end the fight prematurely, a partial success offers a reduced Temperance check, and failing either the Prudence or reduced Temperance check will move on to the existing Fortitude check. I've also reduced the Fortitude requirement for a partial success when attempting to eat the golem, to account for all-stat mixed builds. With this, it should be possible to win with almost any possible combination of stats.

Also, since the checks required for each choice might not be immediately apparent, I've added a new hints system, which can be enabled or disabled in the options. If enabled, it gives hints about what each choice does or requires, and shows a 'warning' icon should a choice result in a bad end (e.g. if you can't meet the stat checks.) Hopefully, this should make things easier for people who aren't too familiar with stat-raising sims.

Another issue was that training stats often felt repetitive. To fix this, I've added more unique training events at certain stat milestones. Mechanically, this changes nothing, but should hopefully make the stat-raising less repetitive. As an added bonus, some of the events also hint at what checks you might be able to pass. Though, since this only applies to stat gains from training at home, there might be slight differences as to when they trigger if you gain stats via other means. For instance, if you had 0 Prudence, and gained +1 from remedial lessons, bringing you to 1, then the event for reaching 1 Prudence via training will instead occur at 2 Prudence, the 2-Prudence event will be pushed back to 3 Prudence, and so on until you reach the point where there's a gap in the milestones, after which all events should trigger at the appropriate milestones. I've also went ahead and updated the 'Play Video Games' option with a couple more games, so every use in a single playthrough should be unique, assuming you only use it when not already at full Sanity.

Overall, while I may not be too proud of the game, especially during the initial release, it has certainly been a valuable learning experience. There are definitely things I could've done better, in particular with the pacing and stat-raising; but honestly, at this point I'd rather not change the game too much, so as to avoid retconning anything or breaking existing saves.

Speaking of which, one thing I had feared when I submitted this project to MGGJ11 was that I might have to break save compatibility in the final release - a known issue with Python (which Ren'Py is based on) is that attempting to reference a variable that hasn't been defined (for instance, if it's added in a newer version of the game, so old saves might not have it) might result in an error. Fortunately, my research has lead me to discover Ren'Py's 'default' statement, which allows one to specify a default value for variables if they haven't been previously defined. In hindsight, I should probably have been using this to begin with instead of declaring my variables at the start of the game or before they're needed... Well, it's a lesson learned for next time I guess.