A Castle of Thread

I remember playing this author's first few games when they came out, and I had memories of something that was promising but not quite there yet; something idiosyncratic, with a strong narrative sense but technical shortcomings.

As of today, this game seems to be his penultimate-latest game, after a quite prolific burst in a small amount of time. It all comes flooding back to me; the same strengths and the same shortcomings are still there.

While I would not say that, as in some cases, his games are just static fiction with a thin veneer of interactivity "just because" - I don't believe they are - it is true that the best part of his games is the story, the setting, the worldbuilding. Perhaps unavoidably in a game that is more narrative than simulationist, the portion of "A Castle of Thread" that I played through (I did not finish it) consists essentially of a sequence of scenes. Not in the claustrophobic sense of one-room situations, thankfully; these scenes can have you in a few rooms, give you space to talk to a few characters, even have some roaming characters.

But then you realise that, for instance, the characters essentially just respond to whatever comes up when you type TOPICS, and little (or nothing) else (might as well code a conversation tree instead), and their roaming around doesn't really mean much. It feels like an exercise in making people alive by not being in the same place, and giving them random lines of text to describe some sort of pseudo-action. Also, certain topics might wield results that are very specific and include, in their results, a reference to the place you're in... and you might have the choice to ask that character the same thing afterwards, in a different place... and you get the same reply... as though in were in a previous location that you're currently not.

There are objects clearly listed in the room's description that aren't implemented (a sink in an unusual bathroom; a closed window in a bedroom; stuff that does call for some sort of interaction or close examination). You cannot "fill container", you must "get liquid" with the container in your posession. Characters are listed by Inform's default system in very innoportune ways; when you first meet an antagonist, whom you pointedly do not know, Inform very helpfully has already told you "So-and-so is standing here, looking such-and-such". Plus, attempts to make characters seem alive by giving them actions ("So-and-so looks at your haversack") jar when they appear in the middle of, say, a tense fight scene. If, upon finding (Spoiler - click to show)the body of Deviah, you try to search it without examining first, you find yourself in the silly position of allowing another character to come into a room where there is something they should react do, but don't; because you didn't trigger a piece of the story by examining that element. This could easily have been avoided by a simple "before doing anything to ___" line.

So there is simply a lack of care to these details; care which is instead given to the narrative. Indeed, the best parts of the game are the story and the narrative sequences. Simulationism is clearly not the point; but if you try to do something that is more narrative-based, like (Spoiler - click to show)trying to attack the bad guy in the Vulgar Unicorn while he fights with your guardian, you are sensibly rewarded by a reaction that makes sense and is visually stimulating, (Spoiler - click to show)of yourself being thrown into another location, where you find another character cowering. Also, the characters that walk around, and appear to take actions of their own, do serve as scenery that makes the scene more lively. I don't know that it makes it all more "alive", because Inform announces them in its default way all the time (the way that the cat follows you around in the tavern section is so strangely unremarked upon I thought maybe it was a bug), so rather than "living", they seem "lively decorations".

Narrative focus is all well and good, but if the technical aspects don't follow suit, it damages the final result. The game doesn't have to be full simulationist, and maybe trying to be so is part of its problem. A more narrative-focused game should accept its strengths. The issue is that, when the player is stuck and trying to solve a puzzle, that's when the illusion will come apart at the (many) seams. When you don't know what you need to solve a puzzle, and start to experiment, and realise that most everything is cardboard scenery, it damages the experience. Again, there is no need to go full simulationist if one doesn't want to; but there is an art to that, an art to describing just enough to lead the player's attention to, and away from, things. An author who doesn't master this art will have players poking where the author would rather they didn't; and if the author didn't also make provisions to gently dissuade them from that, well, then the result is... a game like this. Not bad, but without the proper care.

Not to mention, some puzzle solutions are a bit strange; I did solve the first "proper" puzzle on my own, but that was because I was just doing "stuff" just because the items were there. Every step of the way I thought "surely this won't work, but lemme try it to see what happens". Not only it did work, it was the solution. Well, it's not a spoiler to say I would never think of unclogging a pipe THAT way!

Ultimately, I did not finish this. At a certain point I had the opportunity to ask a new character about an item in my inventory, and their response renamed that inventory item and gave me a new topic. Let's say that the item was revealed to be a Schnoodligan (my made-up word, not in the game) item, and I was now encouraged to ask who the Schnoodligans are.

Well, I entered the vicious disambiguation cycle that is well known. Every time I asked about the Schnoodligans, the game asked me which I meant; the Schnoodligan item or the Schnoodligans. Nothing I input made a difference (incidently, if I'm not mistaken there's an I7 extension called Numbered Disambiguation or something which would have helped here).

I can only tolerate poor implementation up to a point, and this was that point. My trust in the game had been steadily decreasing; with this, I had no reason to trust it anymore. And without trust in the game, how can a player expect to have any enjoyment? Maybe this should have been static fiction instead, after all. Or possibly a choice-based game, instead of a parser.