varicell.jpeg['Homer in Silicon' is a biweekly GameSetWatch-exclusive column by Emily Short. It looks at storytelling and narrative in games of all flavors, including the casual, indie, and obscurely hobbyist. This week she looks at the idea of the accretive player character in two works of interactive fiction: Adam Cadre's classic "Varicella" and Jon Ingold's recent "Make It Good".]

"Accretive player character" is a term of art in interactive fiction. It refers to a protagonist who has motives and abilities that the player doesn't understand or share the first time he plays through a game, but that he gradually learns over the course of many replayings.

Perhaps the classic example is Adam Cadre's "Varicella", a scheming palace minister in some alternate-reality Italian principality that combines the technology of the modern day with the ethics of Machiavelli and the methodology of the Borgias.

The title character hopes to outmaneuver everyone else and end up with the regency. Varicella is a fastidious man with a protocol fetish -- not the strongest nor the most charismatic nor the most openly ambitious member of the court. But he does happen to know all the weaknesses of all his opponents and he's poised to take advantage of them all.

The first time the player confronts this scenario, he lacks Varicella's keen political understanding, and inevitably loses. The second time, he's probably still not up to snuff either, but loses in a new way, for a new reason. But by the winning runthrough, he has Varicella's role down. He does know exactly what is happening everywhere in the castle, every minute of the day. He does know exactly whom to assassinate, when, and how. The player/protagonist gap has been erased by careful training.

It takes a special kind of game structure to make the accretive player character work. The game has to play through fast, so that the player doesn't get frustrated with the prospect of retracing his steps. It has to be fun enough and deep enough to keep the player coming back despite the obvious frustration of losing many, many times.

It has to give enough feedback and guidance that the player does learn from every losing step, and has a chance to make a new, better plan for the next runthrough. It has to offer significantly different experiences to the naive player and to the accomplished one who understands what he needs to do.

Perhaps the hardest aspect of all: the game has to somehow communicate to the player, Look, it's all right, you won't get this the first time, but that doesn't mean the game is too hard for you. Stay around. You'll learn.

Done right, though, the accretive PC is a hugely powerful device.

Having the player learn by replay allows for a narrative arc that has no tutorial phase, no gradual ramping up; big things can start happening right away. The first several times the player goes through the game are the tutorial. Early playthroughs of "Varicella" involve a lot of wandering around the palace, meeting and getting to know people, spying out secrets. That's fun and interesting, and it lays the groundwork for later-- when "Varicella" has time for none of those things, and instead executes a ruthless plan. The final runthrough is a lean story with no futzing around, no time lost.

The accretive PC also promotes player identification. The better the player understands the protagonist, the more persuasively he can play the role right from the start.

This cuts both ways, because (I also find) if the player has spent hours upon hours working out the perfect plot for "Varicella" to execute, he's less likely to balk at some of the protagonist's more reprehensible actions.

makegood.jpgJon Ingold's recent mystery "Make It Good" (also playable online) takes the same concept of the accretive PC and plays it in new directions. The protagonist is a down-on-his-luck cop, one drink from being kicked off the force, who is called in to investigate a murder.

As the game progresses, it becomes clear that the player is going to have to do more than find evidence. He's also going to have to manipulate the suspects in order to bring about a satisfactory ending.

Gameplay thus goes beyond interrogating suspects about keywords and looking under beds, into the realm of assessing and meddling in the psychology of the non-players. And these are alert characters, not the dull ciphers found in many games. They see what the player is carrying when. They observe how he acts. They notice evidence if it's left in plain sight, and sometimes find it when it isn't. They draw conclusions. They talk to one another, sometimes behind the player's back.

All this makes for very difficult gameplay, and there are times when the implementation doesn't live up to the design challenges as smoothly as it might -- but the process when it works is compelling.

To complicate matters yet further, the protagonist evidently knows, or half-remembers through the boozy haze, some important things. The full accounting of what he thinks is never fully available. While he doesn't exactly lie in his thoughts to the player, he verges on being an unreliable narrator.

It's frustrating, not having a good clear access to what exactly the protagonist really knows. The player is stuck running around doing his best to look out for this character's interests without having full access to his memories or total control over his thoughts.

And yet that too factors very effectively into the characterization: the player is coping with the constraints imposed by the narrative, while the protagonist is struggling with a long-term devotion to cheap whiskey. The player would like to know more. The protagonist would like control over his life.

Gradually those two desires converge. In the winning playthrough, the player-protagonist finally achieves both agency and understanding.

[Emily Short is an interactive fiction author and part of the team behind Inform 7, a language for IF creation. She also maintains a blog on interactive fiction and related topics. She can be reached at emshort AT mindspring DOT com.]