Concluding Remarks

That’s it for IFComp 2019. My sincere thanks to all the authors. Writing an interactive fiction game is a massive undertaking, but it’s often difficult to find an audience, or even feedback, for all of that effort. I’ve played through 82 games this year and 76 last year, and I’ve found something enjoyable or compelling or simply unusual in each one.

For reference, here’s the distribution of scores this year:

histogram-scores.png

The mean score is 5.2317. That’s definitely an encouraging sign if you’re trying to find some games to play from the competition; see the menu at the upper right for a list of games sorted by rating. The fraction of games with score at most a given value (without rounding) is shown below:

cumulative-scores.png

The mean and median are thus fairly close. The distribution is definitely different than last year’s; I was trying to smooth out the peaks in the latter, and it looks like I’ve succeeded on that front. Still, the mean and standard deviation of this year’s scores are 5.23 and 1.91 (respectively), which are quite close to last year’s values of 5.17 and 2.11. The precise criteria I’m using to rate games may be ill-defined, but apparently they’re at least fairly consistent.

robotsexpartymurder, by Hanon Ondricek

“robotsexpartymurder” is exactly what the title promises: a game about robots, sex, parties, and murder.

Gameplay: The game is choice-based, with the player clicking links in order to advance. Those links are not just related to the plot development; the character also has an ordinary workload and sleep schedule to maintain, computer systems to interact with, and so on. The game takes place in a cyberpunk dystopia, with the character working for an oppressive, controlling corporation that puts restrictions on those schedules. The dystopian environment is treated more lightly, both in tone and in depth of description, than in most games of that genre, and the game has a stronger and more original premise because of it. 7/10.

Mechanics: The main mechanic of the game is interrogating robots that may be invovled with a murder. It’s a clever idea, and there’s enough flavor and detail in the game to support it. The various maintenance actions required of the player— working, sleeping, eating, etc.— have enough of a connection with the setting to avoid being tedious, and the UI makes them easy to arrange. 7/10.

Presentation: The omnipresent sexual content in the game becomes tedious quickly, and that’s part of the point; those interactions are deliberately somewhat unpleasant. Its characters are well-defined, and their dialogue is strong. The user-interface is generally easy to navigate, with the cavil that the sound icon only toggles sound for a particular page; the option to do so globally is in the settings menu. 9/10.

You might be interested in this game if: You’re interested in the cyberpunk genre and don’t mind sexual content.

Score: 8

ALICE BLUE, by Chris Selmys

ALICE BLUE is an experimental game with an unusual, custom interface.

Gameplay: The game is quite opaque. Descriptions are sparse and unenlightening; most refer to memories of the protagonist without much explanatory detail. It’s not clear how exactly to play the game. Highlighted keywords indicate important objects, and they can be treated as commands themselves. Beyond that, the game is difficult to navigate. Part of that is simply the confusing nature of the game itself, but the fact that unrecognized or invalid commands are simply ignored without any feedback adds to the problem. 2/10.

Mechanics: Although its walkthrough indicates that there are definite goals and puzzles in it, the game is too unclear and difficult to navigate for them to be approachable. There’s a general mechanic dealing with memories, but I was unable to accomplish much with it. 2/10.

Presentation: The system requirements are quite specific, and I was unable to find a machine capable of playing the game in its original format; instead, I used a Windows port kindly provided by a user on the int-fiction forum. While I normally don’t dock points for the difficulty of setting up the game, this particular situation is unusual enough that it at least deserves mentioning. The text of the game is vague to the point of meaninglessness, and it’s not polished; the first line of the game even contains two errors. Gameplay is frustratingly slow, with commands taking multiple seconds to run on my machine. 1/10.

You might be interested in this game if: You’re interested in what an IF engine written as a bash script looks like.

Score: 2

Río Alto: forgotten memories, by Ambrosio

“Río Alto: forgotten memories” is a choice-based game with an elaborate, versatile interface.

Gameplay: The game is large and slow-moving. The exploration itself is enjoyable, and the UI is convenient for that purpose; in particular, visited locations appear in its bottom menu, and clicking on them allows the player to instantly return to them. Despite the size of the world, I ran into an apparent dead-end in my playthrough, with no objects in the current room or inventory available to interact with and no other location available. 5/10.

Mechanics: The main appeal of the game is its interface, which is clever and well-designed. There are tabs for the character’s thoughts, inventory objects, and locations, and interacting with them is generally smooth. There are some unexpected difficulties with navigation, such as some issues with the row of quick-slot entries, but the UI is solid overall. It occupies an unusual niche between the interfaces of a graphical adventure game (e.g., later Sierra games or LucasArts game) and of traditional text-only IF, and I’d be interested in seeing how it applies to other games. 7/10.

Presentation: There are some errors (or mistranslations) in the game, e.g., “The doctor recommends me to take walks in the pine grove near the village to relax and acquire vitality,” or “Is he taking me for a crazy?” The characters and setting are promising, though they could use a bit more detail. 5/10.

You might be interested in this game if: You’d like to play a game with an original, well-designed interface.

Score: 5

Sugarlawn, by Mike Spivey

“Sugarlawn” is a scavenger hunt set in a Louisiana mansion, with the protagonist trying to explore the mansion while dressed as a chicken to win a prize on a reality television show.

Gameplay: Despite the premise, gameplay is reminiscent of an old-school adventure game, with its focus being exploration and treasure collecting. The game has substantial replay value, both from the puzzles themselves and the optimization problem of getting through the mansion in the minimal amount of time on the in-game (but not real-time) clock 8/10.

Mechanics: As the synopsis above implies, the game doesn’t take itself seriously. Beyond that, though, there is a serious optimization problem for the player to solve: The player has thirty minutes to find the most valuable collection of treaures, with each in-game move taking an appropriate amount of time. Aside from figuring out the routing puzzle, there are also individual puzzles to be solved to obtain certain treasures, as well as a larger puzzles of matching treasures to appropriate locations (cf. the trophy case from Zork I). 9/10.

Presentation: Although it’s fundamentally a puzzlefest, the game is smooth to play throughout and has numerous embellishments (e.g., the dialogue from the television host) to add flavor to the game. The replay value is high, and the style and humor is still fresh after multiple runs through the game. 8/10.

You might be interested in this game if: You like old-school treasure hunts, but would like to play one with more character.

Score: 8

Winter Break at Hogwarts, by Brian Davies

Because the IFComp is a small competition for independent developers making non-commerical games, it’s an ideal venue for games based off other works. While fanfiction is not a genre I’m particularly interested in, I nevertheless found “Winter Break at Hogwarts” to be an appealing puzzlefest set in the Harry Potter universe.

Gameplay: Gameplay begins by picking the main character’s house among the four available. The choice doesn’t appear to be hugely meaningful, but it’s one of a series of entertaining bits of flavor from the Harry Potter series that appears in the game. Professors from the books appear as NPCs, generally with an object to collect or a puzzle for to solve. Locations in the book also appear in the game, and there is a map to help with navigating the sprawling environment. 7/10.

Mechanics: The puzzles are largely solid, though some are simply standard adventure-style ones rather than anything more evocative. Even given its size, the game environment is not compact, and it’s occasionally difficult to navigate with the map and frustrating without it. The game also contains frustrating timers to remind the player when the main character is hungry or cold. Both problems are easy for the main character to solve (as in, for example, “Savoir-Faire”), but the timers don’t contribute anything to game besides annoying the player. 7/10.

Presentation: While the game does recreate the overall feel of the setting of the Harry Potter books, both the in-game environment and the style of play reminded me of the mid- or late-90s style of open-ended puzzlefests set on a college campus (e.g., “GC: A Thrashing Parity Bit of the Mind”). There are some minor issues with unhelpful descriptions for objects (e.g., a number chart that’s simply described with, “You see nothing special about the number chart”) and guess-the-noun issues (e.g., an overcoat that’s described initially as a spare coat, but isn’t recognized as COAT or even SPARE; only SPARE COAT works). Overall, though, the text is solid and fits the tone of the game. 7/10.

You might be interested in this game if: You’re a fan of the Harry Potter series.

Score: 7

Mental Entertainment, by Thomas Hvizdos

In “Mental Entertainment,” the protagonist is charged with evaluating whether three clients have an unhealthy dependence on the virtual reality of its dystopian setting.

Gameplay: Gameplay consists of entering three small virtual-reality settings, having a conversation with the client in each area, and marking each one as healthy or unhealthy. There aren’t any apparent guidelines for judging each person, though, and the dystopian setting isn’t detailed enough to indicate the criteria expected of the narrator or the criteria of interest to the player. Conversations with the NPCs have significant depth, though I didn’t find them extremely illuminating. 5/10.

Mechanics: The bulk of the game is the set of conversations with the three characters, with the player using them to make his or her decision. There’s little explanatory information about the decisions to make and little feedback after making them, though, so the choices feel arbitrary and unsatisfying. Despite the branching of the conversations, there are still many terms the player would reasonably expect to be topic keywords that are not recognized. 4/10.

Presentation: Even though it’s mostly a hub for the three characters’ scenes, the starting room has several embellishments to make it more interesting, including books that give some indication of the setting. The other three settings are less evocative, but the conversations are interesting. 5/10.

You might be interested in this game if: You like dystopian fiction.

Score: 5

Hard Puzzle 4: The Ballad of Bob and Cheryl, by Ade

“Hard Puzzle 4” was actually near the top of my randomly-generated review queue. It’s a clever, well-crafted game, and my reluctance to spoil it with hints or walkthroughs has caused me to push it further down the stack several times.

Gameplay: As the title implies, the game is a difficult puzzlefest with a complicated overarching puzzle. The game strikes a balance between difficulty and approachability; although its main puzzle is very involved, it’s something that the player can approach and experiment with (as in, for example, the babel fish puzzle from Infocom’s Hitchhiker’s Guide to the Galaxy game), instead of something that requires a single flash of insight 7/10.

Mechanics: The puzzles (or components of the main puzzle) are legitimately difficult, and solving them is satisfying accordingly. The game provides debugging verbs to the player, and my expectation on starting the game was that they would be the single mechanic behind the game’s solution; it’s actually much more complicated than that. 8/10.

Presentation: The game is a puzzlefest with strong puzzles, and it’s worth the time it takes to solve. Using the debugging verbs is a clunkier than it needs to be (e.g., handling the exit command), but gameplay is largely smooth. The one significant problem I found was a crash caused by commanding a character to take a certain action with the stool in the game. (Although it’s related to the ostensible debugging verbs of the game, I don’t think it was deliberate.) 6/10.

You might be interested in this game if: You’ve played the other games in the Hard Puzzle series.

Score: 7

Island in the Storm, by JSMaika

Making an IF parser is a large and difficult programming task. “Island in the Storm” uses a custom system, IntFicPy, that’s far more promising than most I’ve encountered, and I look forward to the author’s continuing development on it.

Gameplay: The game is a conventional, puzzle-heavy adventure. Gameplay is solid, but the main focus of the game is showing off its engine. Unlike most custom parsers, the author’s system is up to the task. The game is reasonably complex, and the system has the necessary features to handle the game’s complications successfully. “Island in the Storm” is not as ambitious or polished as, say, “Curses” as a proof of concept for the system, but it’s an enjoyable game with a solid implementation. 6/10.

Mechanics: The puzzles are generally straightforward adventure-game ones, but the distinguishing feature of the game is its engine. Unlike most custom systems, the one in this game is solid. The UI is attractive and functional, and the parser is easy to interact with. Furthermore, it’s based on Python, which obviates the need for many users to learn a new, specialized language.

The author’s system is so appealing because there’s still no completely satisfactory IF programming language designed for users who are already familiar with general programming. Inform 6 is a solid, well-designed language, but it’s constrained by its association with the Z-Machine format, which is no longer relevant to modern IF and imposes unnecessary restrictions on the code. Furthermore, the lowest-level routines inside it are spaghetti code that’s difficult to overhaul. (In fairness, that’s in part because of the restrictions of the language, e.g., a restriction on the number of local variables in a function.) Inform 7 has a wonderful IDE and a clever rule-based system, but those features are marred by an arbitrary, poorly-designed syntax and inadequate documentation. (There’s quite a lot of documentation, but it’s largely a series of examples; it’s hard to find an explanation of how exactly the system works. If your problem doesn’t match an example, you’re out of luck. Compare that to the masterfully-written Inform Designer’s Manual.) A well-designed, easily extendable, polished language or engine suited to IF would be an incredibly useful for developers. Python is a particularly good base for the system, since it’s a familiar, well-designed language that’s often used in machine learning. 9/10.

Presentation: There are some typos and grammatical errors in the game (e.g., “A board has been lain across the chasm.”), and some of the descriptions are extremely terse (e.g., the description of one key is is simply, “There is a key here.”). While the parser is promising, there are some useful features conspicuously missing from it. There is no UNDO command, for example (or, at least, I didn’t find the term or abbreviation used for it), and pronouns can’t be used. The parser also doesn’t recognize groups of objects, whether with commas or the word AND. Thus, for example, neither TAKE NOTEBOOK, KEY nor TAKE NOTEBOOK AND KEY are recognized; the player must take the objects individually. 6/10.

You might be interested in this game if: You’d like to see a proof of concept for the author’s system.

Score: 7

Citizen of Nowhere, by Luke A. Jones

“Citizen of Nowhere” is a lightly comedic, parser-based game reminiscent of old-school puzzle games from the 1980s.

Gameplay: The style of the game is one familiar from games in the 1980s like Zork I or The Pawn: an emphasis in the game on puzzles, with the plot being desultory; a sprawling, crazy quilt world containing many otherwise unremarkable hallways of junctions; a juxtaposition of lazy-medieval fantasy elements (e.g., giants and ogres) with modern machinery; text that’s lightly comedic and contains several puns; and so on. It’s an interesting style game of that isn’t common anymore, and it’s refreshing to see such a different kind of game. 5/10.

Mechanics: Most of the puzzles involve gathering inventory items and using them in other places. There are some variations, though, such as the word puzzles posed by one NPC carrying a useful object. As is common in its genre, the NPCs in the game are either obstacles to be defeated or the equivalent of vending machines. 5/10.

Presentation: The light tone of the game fits is genre, and puns are rampant in the text. The text also contains many instances of missing puncutation (e.g. a missing period and line break in the response to READ LETTER) and capitalization in the text. There are some implementation issues, including the parser’s not recognizing FLAGS (as opposed to the singular FLAG) in a puzzle involving manipulating several of them, and the response to READ MAP is simply, “Fill in map description here.” There are many NPCs in the game, but none of them is particularly memorable or well-characterized. 4/10.

You might be interested in this game if: You like puns.

Score: 5