DemonApologist's IFComp 2024 Responses

58 | HEBE

58 | HEBE
by: Marina Diagourta

Progress:

  • I timed out at the 2-hour mark, having not completed this game. I scored 60 points, solving four of the puzzles. There were two remaining puzzles that I had been working on for a long time but could not solve: I could not find an obol to give to Charon to access the Underworld; I could not solve the puzzle involving scales in Thebes and might have even softlocked it.

Things I Appreciated:

  • I just recently watched a show called Kaos on Netflix, so I was more interested in Greek mythology than I normally would be. I enjoyed the use of Greek terminology (things like amphora; obol) that I had to learn in order to advance puzzles. I liked the feeling that I was in an unfamiliar cultural location, but where the protagonist is comfortable and isn’t going to spell out for herself exactly what an “obol” is, so I need to do that work for myself. It made it a more active process, and I quickly started looking up things on Wikipedia/Google etc for assistance. As annoying as it can be, not having a walkthrough or in-game definitions of terms does make me approach them differently: like, no one else is going to help you get through this, you’d better get serious. There is something empowering about solving a puzzle under those conditions, even if it is ultimately not ideal.

  • I liked the gating of puzzles, where they were mostly nonlinear and can be done in any order. When I was stuck on something, I would just go work on something else, until the end when the only puzzles I had left were the ones I couldn’t solve. Being allowed to explore more and gain experience before returning rather than requiring me to solve them in an exact order allowed me to make more progress than I think I otherwise would have.

Feedback/Recommendations/Questions:

  • I encountered probably dozens of weird bugs and errors. I’ll try to describe these in terms of categories of errors. Feel free to look at the attached transcript if you want more information of how I flailed through this game.

  • (1) Minor implementation errors: The first main category is minor implementation errors, where clearly there was just a slight mixup in the code. For instance, in the location Agora of Thebes, you are told that one of the exits is “Gates of Thebes (S).” However, there is no exit from the south out of this location, you have to guess that it’s actually southwest. While annoying, this seems like a minor error to me because the text could just be updated to “Gates of Thebes (SW)” and solve this issue. Another example of this would be in the “investigate” command. This command (which I wish I had found earlier) lists out interactable objects in the room. However, this is not always accurate. In the location Heroon of Kadmos, investigate fails to list the “inscription” as one of the interactable objects, even though the inscription is mentioned in the room text.

  • (2) Chat GPT description carnage: This game features Chat GPT produced/edited text at different locations. Often, things in the (for a parser game, somewhat long) descriptions generated appear to be things that you could interact with, but the game hasn’t implemented them. Why tell me that there are mortars and pestles on a counter, but then deny the existence of the mortars and pestles and the counter if I try to interact with them? This creates a conflicting relationship between the reader and the descriptions. On one hand, I wanted to tune these out since I was so often rebuffed from trying to examine/make sense of objects in the the descriptions. But, on the other hand, in a parser game, to solve puzzles you are required to pick out things in the descriptions to focus on to find solutions. So I found myself trying to do both things at once, in an unproductive way.

  • (3) Location Permeability: I’m sure there’s a more technical term for this kind of error, but I’m describing it as “location permeability.” By this I mean, when you take actions at a certain location, objects at other locations permeate what the reader imagines is a conceptual wall separating that location from the others to interact in a jarring way. For example: the description in the Secret Garden location says that “[your] eyes would be immediately drawn to a statue of an incomprehensibly beautiful man.” But attempting to “x statue” here calls up the description of the Aphrodite statue in the previous room. The game didn’t recognize that the Aphrodite statue wasn’t here, and this was a significant barrier for me in solving this room because it took me a long time to return to trying to look at the man specifically, which held the instructions to the puzzle that I had been failing to solve and flailing around with. Another example of this, from the same room, is if you try to “extinguish candles” in this room, the game says “You can’t reach into Prytaneion Hall. This doesn’t make any sense.” (On that, the game and I agree.) The game seems to be recalling the torch puzzle from another room, rather than realizing that I am trying to extinguish the candles right here in front of me. The final example of this came when I was desperately trying to solve the scale puzzle in Thebes. I had resorted to placing different objects on the six pans to test them. One of these objects was a water lily. I placed it on the scale without issue. The issue came when I tried picking it up. “Take water lily” summons the text from the apparently super-cursed Secret Garden location where you’re only allotted a single lily: “It would be a shame to remove any more of those beautiful flowers. Frogs could be sleeping in them!” This meant that the water lily was irretrievably stuck on the pan.

  • Okay, so enough about the errors. I also want to give some feedback on the map and puzzle design. Of the four puzzles that I solved, they were all self contained and solvable within that room. That is, you navigate the world, enter the temple, get the puzzle instructions, and then everything you need is in that room (or in the “help” verb list which is essential reading to make progress). The question that leaves me with is: what is the purpose of the rest of the map? There are many interconnected city/port/gate/agora locations that have been added to the game, but almost none of these locations have anything that you can do in them other than move to a different location. When I couldn’t solve puzzles (like the obol puzzle, which appears to be some kind of social puzzle where I need to find a coin somewhere in the world to give to Charon), I was torn as to what to do. The grim task of going through every overworld location and trying to poke at the uninteractable descriptions in case one of them miraculously produced a coin seemed like a bad strategy, given that all other puzzles seem locked to what’s immediately nearby. Yet, that is what I was contemplating doing. So I think the number of locations could be greatly focused down to match the scope of the world. A smaller world with a higher proportion of locations being puzzle sets feels richer and more satisfying than a sprawling map that is difficult to find things to interact with.

What I learned about IF writing/game design:

  • Using Chat GPT to generate descriptions, which then adds in objects that can’t be interacted with, is deeply frustrating for players. Parser game room descriptions need to be precisely tailored to not lead players to interact with illusory items. This requires a lot of specificity of detail, and efficiency in not overstating what is present, which is the opposite of what Chat GPT seems to provide (loose, generic, and vague text).

  • This is a valuable lesson in scope. I just think about all the effort that must have gone to programming all the locations that are severely underutilized. As a player, I’d rather have a short game with three beautifully implemented and thoroughly vetted/tested puzzles, than a sprawling game with six or seven messy puzzles that are hit and miss. So as an author, the advice I take away from that is to continually check back in with the scope of the project and the amount of time I have left to submit it (in a case like this where there was a specific deadline), to see if I need to prune some of my planned branches/sections.

Quote:

  • “The key to passage lies from the lightest touch to the heaviest step.” (I read these words dozens of times trying to make sense of what to do, so they are kind of burned in my memory.)

Lasting Memorable Moment:

  • I really thought I was never going to solve the Poseidon temple season puzzle within the 12-turn time limit, I was so relieved that I had actually solved it after making so many botched guesses.

DemonApologist_Hebe.txt (226.5 KB)

6 Likes