Review – A Castle of Thread

This is a medium-length parser game in three acts. As he often does, this author set the story in a world that would be at home in the HP Lovecraft Cthulhu mythos. I’m not sufficiently schooled in the mythos to tell whether this borrows specific elements from the mythos or just its flavor.

[Some spoilers follow beyond this point]

In the story, a discovery has been made in the northern waste lands: a boulder with undecipherable inscriptions in an ancient script. Various kingdoms have sent sages to study the writing. That is the role of the 15-year old protagonist, Polt Kober, who was selected for his exceptional linguistic skills. His journey is divided into a boat ride across a swampland, meeting a contact at an inn, and traveling north to study the boulder.

Danger attends every leg of the journey, which keeps Polt — as well as the player — on their toes. While Polt is often the target of violence, Polt himself is not a fighter. When he is attacked in the boat, his body guard does the fighting. At the inn as well, his bodyguard runs interference, while Polt figures out how to escape. Confronted with a thug near the inn, Polt defends himself using magical dirt rather than his knife. Finally, when all hell breaks loose in the final scene, Polt considers using his family blade, but ultimately, the smart move is to run away at full speed. This seems right for a Lovecraftian hero, who is often framed as an academic or investigator who has no chance going toe to toe with evil in terms of brute force, but can at least save his sanity by relying on special knowledge and knowing when to run.

The presentation of text in this game feels more Old School than the other parser games I have played in this comp; I had more the sense of playing a game than reading a novel. Descriptive room text is shorter and output from most commands is relatively terse. Turn by turn text is often choppy. There are frequent short, repetitive descriptions of the environment. I spent quite a while exploring around on the deck of the swamp boat, for example, and slithering snakes, fisherman, and rundown shacks constantly spooled by in the background, like a low budget Hanna-Barbera cartoon. Near the inn, the sadiki birds were constantly doing this or that, but as far as I can tell, did not really contribute to the story. I think there is a good role for stage business and occasional environmental descriptions to keep the story dynamic, but care needs to be taken to avoid firing off a series of them, each as a standalone paragraph. The same can be said about roving characters, who are described as entering, leaving, or just being in a room.

The author used a surprisingly long list of inform extensions, including Eric Eve’s conversational framework and related modules. Despite that, this game contains a lot of less PC/NPC interaction than some of the other parser games. There are a good number of NPCs and each is primed with a limited set of conversational topics, but the exchanges tend to be short and for the most part not informative. In some cases, the mechanism does not seem to be working correctly, because not all topics disappear after use. Since the author went to the effort of incorporating these modules, it is a shame that they weren’t put to more extensive use and that features such as topic branching were not used.

I’d like to focus on a number of implementation decisions that lowered my scoring in the play and technical categories. I tried to put the more important ones first:

1) Often attempted actions are blocked with a reply along the lines of “you were about to do that, but realize that’s a bad idea”. For example, breaking into the first mate’s cabin when he’s standing next to you. Well, as a player I fully recognize that it’s a bad idea, but I want to see what would happen. Undermining player autonomy in this manner seems like a bad idea, even if it does save the author some writing. In this specific example, why not let Rakton attack me? He’s going to do so anyhow. Might as well just move that event earlier in the time line.

2) No hint or help command. The author did provide a detailed walkthrough, which was appreciated, but looking at the competition this year, some sort of in-game hint system is pretty much expected, even in a relatively short game. I found some of the puzzles unintuitive. In several cases, I had to employ game metalogic: If I could do some, like drop the cat down the stove pipe, I did. Is that something I’d normally do? Is it a good way to clean a stove pipe? No, but there are a limited number of objects and ways to combine them, so it was worth a try.

3) The author did not take the opportunity to develop two prominent NPCs: Venkath, his bodyguard, and Paevana, the alewife. From the start of the story, Polt is told to keep his bodyguard nearby, which makes sense if the bodyguard is to be effective. However, while various other NPCs mill about the boat, Venkath sits like a potted plant. More motion, dialogue, and perhaps the ability to give orders to Venkath (“Venkath, shove the cat down the stove pipe, would you? I need to retrieve a key from the bird’s nest.”) would have given Venkath more depth. A closer relationship between them would also have made the later loss of Venkath more poignant. As for Paevana, she’s there for rescuing and provides an extra set of hands at one point, but she is entirely passive and has not a glimmer of personality. Yet, she sticks to Polt, follows him north to the boulder and is still with him at the end. Is she just doing this because she’s lost everything and has no other choices, or is there some attraction to Polt? Enquiring minds want to know.

4) On the boat, places are described as port and starboard. The author goes to the effort to make sure that the player understands that as the boat is headed north, starboard is east, so it makes sense to be able to use cardinal directions for movement. However, it is a boat and adding port, starboard, fore, and aft as supplemental commands during this scene would have been trivially difficult. Since the descriptions reference nautical directions, it would be expected by the player.

5) Doors and keys. For the sake of play, I would have appreciated more automation with doors and keys. Do I want to use the key for door number one to open door number one? Of course I do. When I say open a door, if I have the key for that door, it should just unlock and open. As author, to make my life simpler, I would also have probably found a way to ditch keys and other useless items between scene transitions using excuses like fumbling them on the gangplank or needing to pack light for a journey.


Story: 5

Voice: 4

Play: 4

Polish: 3

Technical: 5


Preliminary Score: 4.2

Transcript: thread











One thought on “Review – A Castle of Thread”

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.