A tiny, unusual little wordplay game, scarcely IF. World-model is irrelevant, and the parser only accepts one type of command. You are given words, and have to supply a rhyme for each word. Your score is how many acceptable rhymes you supply within 60 seconds. The framing is that this is a psychotherapy exercise; you're being asked whether you feel emotional state X, and you respond with, 'no, I'm feeling Y'. Y does not always have to be an emotion, since the game recognises a wide range of rhyming words, some of which aren't even adjectives.
On the one hand, gameplay is simple and easy, so much so that it's tempting to think of this as a toy rather than a game; getting a good score might take a few tries, but mechanically it's not difficult. The game tells you the puzzle's parameters straight off. On the other, the framing creates a weird uneasiness about the exercise. A fair amount has been written about how the psychotherapy premise of Eliza tended to make players take it more seriously. Similarly, WDHYF has something of the tension of a word-association game about it; it's a silly exercise, but there's the sense that you're being judged on it. This is particularly true because the act in question is laying claim to emotions, which is something people generally invest a lot in. Laying claim to an emotion can be a vulnerable act of self-expression, and it's also a commitment of sorts, a reinforcement; we don't look kindly on people who fake emotions, and we don't want to be seen as someone whose emotions change rapidly for no reason ('emotionally unstable' is a euphemism for 'crazy'). Saying 'I am happy' or 'I am sad' costs you something.
But because of the constraints you're usually forced to lay claim to a rapid, little-considered series of random, mismatched, nonsensical emotions. So there's a sense -- slight, but disquieting -- of brainwashing-exercise about this: repetition, illogic, pressure, self-obliteration.
The thing that struck me most about House of Fear was the contrast between content and structure. For Leonora Carrington and the world of 1940, surrealism was challenging, avant-garde, politically potent. For IF, surrealism is a stock approach, comfortable and familiar. The strange-but-consistent world built from psychological allegory, the focused set of interaction styles, the house-of-locked-doors map will all be immediately familiar. The disorientation of the protagonist isn't shared by the player.
There is, in fact, an unusually strong division between player and PC in this game. In the archive photographs that illustrate the story, Leonora is always looking out, directly at the camera. The flashbacks that recount her past remind us that we're not involved in her real life. The game's plot, which follows a predictable arc of defeating one's demons and self-actualising, feels like something viewed from the outside. There's the sense that Leonora is already pretty damn self-actualised and that the game's events are just a re-enactment of that. There are two elements that don't quite mesh: a coming-of-age, self-creation story, and the depiction of Leonora as someone who already has a strong sense of her own identity and strengths.
As a piece of design, House feels patchy. There are two major gameplay mechanics, which are easy enough to figure out but aren't smoothly managed. The acquiring-forms mechanic is the more interesting of the two, but it tends to default towards one-key, one-door; the alchemy puzzle, on the other hand, is actually two puzzles but doesn't make this clear. The setting feels fractured, lacking aesthetic unity (this may be because I'm unfamiliar with Carrington's work, which is not reproduced in the game). There's less polish than you'd expect out of a design style this orthodox. And I have a particular personal hatred for the floating-in-a-formless-void opening.
House of Fear succeeds insofar as it makes the real Leonora seem like an artist I should find out more about; it doesn't quite succeed at giving me much sense of her personality beyond Strong Feminist Artist. Extensively researched, it sensibly errs on the side of directly using too little of that research, rather than laboriously spelling everything out. The writing is competent to good, and doesn't become obtrusive even when dealing with quite difficult material; if it has a flaw, it's that it verges at times on the overly earnest.
Implementation is mixed; quite robust in places, sparse and patchy in others. The NPCs, in particular, are a little too inert. But I found it impossible not to like House; at heart, it has interesting subject-matter, decent writing and good gameplay flow everywhere but one point.