I initially tried playing this story towards the beginning of the competition, but hit a couple puzzles that seemed to block any further progress. I perseverated on these points for a while, but ran out of options. There is no built-in help or hint system, but I didn’t want to ruin the experience, so I shot off an email to the author along with a transcript showing where I was stuck. She wrote back with a hint that at least let me know that I had not reached a dead-end position. Even more importantly, she pointed me towards an updated version that not only patched a couple trouble areas, but significantly polished the writing and story elements.
From both a technical and writing standpoint, this is a more ambitious work than I have encountered up to this point in IFcomp 2012, so it’s not surprising that it requires more debugging and tweaking than most. I’d have been surprised if releasing this work to the wild did not produce some bug reports.
With the new version, I got somewhat further, but I still could not complete the game without a peek at the walkthrough. That didn’t detract much from the experience, however, so this story still gets a strong recommendation.
I have to invoke spoiler mode now, so more comments after the break.
This story combines three elements that have deep roots in the thirty year history of IF, but the work as a whole does not seem derivative or played out. These elements are: character switching (Suspended, A Mind Forever Voyaging), escape (a staple of not just IF, but casual games in general), and an amusement park setting (Infocom’s Sorcerer, Scott Adam’s Mystery Fun House). The object of the game is to free a monkey trapped in an amusement park, and the player does this by jumping back and forth between different bodies: the monkey itself, a ghost and a robot.
There are two flavors of character switching in IF: the body possession version, and the switching cameras version. I believe that Infocom’s Suspended was the archetype for the body possession version. In that story, there is a central PC who is in a state of suspended animation. To navigate that story, the player must extend his consciousness to take control of various robots, each of which has specific capabilities and limitations, just as the characters in EFS. Figuring out how to exploit the capabilties and deal with the limitiations is a central part of both games. More on that in a bit. A natural extension of the body possession model is any story taking place in a virtual realm, where a player character takes control of a virtual character, whether in a computer model, dream world, parallel reality, etc. This also goes back to early Infocom days: A Mind Forever Voyaging, for example. The common element in these stories is that the player identifies with a persisting, central character and the swapped bodies are just an extension of that character.
The other flavor is camera switching, in which there is no central character. The player moves around from body to body, changing perspective. In most cases, when not under the control of the player, it is presumed that other characters go about their business autonomously. I guess this could be called the SIMs model.
Escape from Summerland follows the SIMs model, then. Characters are introduced one after the other, more or less as they are encountered in the story. At one point, a text box, which is out of the flow of the story itself, provides instruction that the player can switch between characters using an abbreviated command. Similarly, the parser tells the player when it is not possible to switch to certain characters.
The EFS authors were clever in their choice of characters in that they side step a lot of implementation hassles that would add many lines of code, but not really improve upon the story that they are telling. First, none of the characters can communicate with each other. While each character has some stylized internal thought dialogue, no dialogue takes place between characters. There is no need to extensively implement ask, tell, talk, etc., so that whole scope of development is avoided. Similarly, the ghost’s inability to interact physically with objects narrows the scope of his interaction with most world objects; it also requires that the player switch to another character to accomplish any physical feat.
The story is seen largely through the eyes of Amadan, the ghost. As the only human character, his perspective is the most easily understood, and his thoughts provide exposition about the situation and goals for the game. His inability to manipulate items is offset by his utility as an incorporeal scout to scope out otherwise inaccessible areas and to describe places and items as a human would seem them.
The authors have further contained the game’s problem space and created layout-based puzzles by imposing limits on characters: The robot can’t negotiate vertical movement on its own, the monkey won’t go into dark areas, and the robot has limited power. These constraints add to rather than detract from the experience because they make sense for each character. The only authorial direction that feels heavy handed are the compulsions that the ghost feels to not stray from a path towards aiding the escape of the monkey. A larger world is described, particularly within Summerland, but locations outside the main plot are not accessible. Given narrative economy, this is not a major criticisms, and even the ghostly compulsion could be excused as a matter of supernatural forces.
The story is not complicated, but it is well told. The plot amounts to three words: “free the monkey”, but accomplishing this takes significant effort and is entertaining. I had the sense of a larger world and was surprised to realize at the end of the story that I’d only explored about ten locations. I had some difficulty figuring out the setting. At the mention of a caravan and what sounded like a roving circus or carnival, at the outset, I thought this would turn out to be set in the 19th century. Then there is the mention of aerial bombing — a steampunk past? England during the Blitz? That idea went out the window when I encountered the battlebot — now I was thinking late 21st century at earliest. In the game description, Summerland is referred to as a fairy amusement park, so the anachronistic elements might be explained by the amusement park being otherworldly. However, in actually exploring the amusement park, it seemed worldly enough to me — I didn’t see evidence of real magic, just the sort of props that one would expect. Even the Selkie tanks seems to be a mechanical contrivance to entertain the “marks”.
Each character has a distinct and consistent voice: the responsible Amadan, the simple but emotional Joquette, and the analytical but damaged robot. Getting into the heads of non-human characters is a particular strength of IF, and body-swapping capability is specifically built into Inform at a fundamental level. Joquette’s sentences are simple, and reflect primal emotions: usually joy, sometimes fear. Maybe I’m overanalyzing, but she strikes me as having an infantile world view — everything revolves around her and the world is there for her to explore. The emoticons that color her internal dialogue do a great job of conveying both her emotional state and also reflect her preverbal mental state. The robot, on the other hand, has a very alien perspective consistent with its purpose-built firmware. As a battlebot, its perception is informed by a number of sensors, but its interpretation casts everything in terms of potential battle tactics.
Despite seeing the world from three perspectives, the player usually has a good sense of goals and after sufficient monkeying (and roboting, and ghosting) around, can work out how to advance the plot. The one exception to this, for me, was getting the balancing stick out of the cupboard. I don’t think I would have figured that one out on my own, regardless of the amount of time I spent on it. The game is structured so that the player must make frequent use of each body, but the overall structure comes off as natural rather than contrived.
While the authors did what they could in many areas to constrict the scope of required implementation, in other areas that they player might encounter, they went correspondingly deep. In many cases, I was very pleased to find that the authors had anticipated some unlikely player actions: for example — as the ghost, try entering your own dead body; as the robot, try changing perspective to the robot, or at the end, try going in every possible direction. This, and the general lack of bugginess in this programmatically complex game, must reflect multiple rounds of high quality beta testing.
The revised version of the story flows well and as far as I can tell, can’t be put into an unwinnable state. Generally, I wasn’t distracted by spelling or grammar issues, although there are a copule of programmatic items that could be tweaked. The work would have been improved by a context-based hint system.
This work required careful attention to the complexity of the character and object interactions. The story provides a gradual introduction of characters occupied by the player and provides the ability to smoothly change perspectives. As mentioned above, a number of design choices have cleverly limited the complexity of implementation to those requried to tell this story; even so, the player has a lot of freedom in exploring this world, and it’s encouraging that everything works as well as it does.
I think my rating may reflect what I’d like this game to be rather than what it is in its current form. With some additional editing and programming, this story could be a strong contender for the post-comp-comp, regardless of how it does in IFcomp.