At Runic Games, if it isn’t randomized, they don’t even want to hear about it. The office restrooms change sexes several times per day, each employee takes a new desk at random intervals, and the water cooler is just as likely to spit out boiling oil as H2O. It’s the same way my posts bounce between fact and fiction without so much as a kind word of warning. In Torchlight, the team further iterated on the concept of randomized dungeons with great success. Now, with Torchlight II, Runic is experimenting with an idea rarely explored: random bits of story.

You’ll still get the randomized dungeons you love, and, if you’ve followed the game, you may already know you’ll be getting outdoor areas as well which will also be randomly generated. In the early days of procedurally generated dungeons, the engine built the dungeon one tile at a time. These days, designers build large chunks which the engine pieces together to form a cohesive area to explore.

The size of chunks fed into the engine has been growing and this sequel with its outdoor areas brings a significant increase in the number and variety of pieces composing a single area.“…the largest dungeon we had in Torchlight was maybe six or maybe seven chunk types and that was in Palace towards the end of the game and that was a very linear progression but it was still a pretty lengthy one. For the outdoor areas forTorchlight II, I think right now we’re averaging around 14 chunks for one of the bigger outdoor areas,” says Patrick Blank, the game’s lead level designer.

The outdoor areas are a natural evolution from the fully underground questing in the first game. It’s a great change of pace for the player to be able to get out into the open from the cramped catacombs. Runic’s founders made the same jump going from the first Diablo to the second back when they were Blizzard North. However, quest randomization in the Diablo series was… well, it barely existed.

In Torchlight II, you will find random events along with other random quests tied to, what Runic calls “landmark areas.” The random events sound very much like those of Red Dead Redemption. JD Wiker, Torchlight II’s resident wordsmith, told me, “An event quest could basically just be a little piece of art asset that just happens to be sitting there like a burned out wagon or something.” This burned out wagon would then have a task associated with it — perhaps finding the bandits who destroyed the wagon or returning a ring found in the rubble to its rightful owner.

The landmark areas, on the other hand, are larger set pieces. JD’s example is the cemetery. “So, if a zone has a cemetery landmark in it, it’s because there are quests available that are associated with that landmark. If you’re following the story, you should get the associated quest. And, if you’re a thorough player, who takes a lot of side quests, you’ll probably get an unrelated quest or two that goes to the same place.”

NPC motives for assigning a quest also have some variance.  JD provides an example. “(The NPC) will have you go take care of a task that’s going to be the same because that’s the event task, but each time you encounter him, he’ll have different quest text. He may say, ‘Go fetch me the big rubber ball.’ I’m thinking of that because I’m looking at (Runic Minister of Culture, Wonder Russell)’s Dog, Falcor. He may say, ‘Go fetch me Falcor’s big red rubber ball,’ one time because he was playing with Falcor and he needs Falcor to come back. The next time he might say, ‘Go fetch me the big rubber ball because Falcor’s old rubber ball is worn out and I need a new one.’ It’s the same task, it’s just that he’ll have a different text of why he wants you to go on that particular task.”

It’s reminiscent of recent Rockstar games in which the dialog is slightly different on subsequent repeats of the same task. Although, in those cases, the scenario is the same each time. Anyone who has had to repeat missions in GTA 4 can attest this is a welcome addition to that game, and, in a game that begs to be replayed like Torchlight II, the feature will be welcome to players.

Just as dungeons remain constant in a single game, so do quests. It is only in starting a new character that the procedural content generation engine works its magic. JD explains, “…it’s not going to be that you’ll run into this guy and each time you run into him he’ll say, ’Oh, by the way, I need another rubber ball.’ It’s you’re going to run into him, and he’s gonna say, ‘Get me a rubber ball,’ and you’ll go get it and he’ll be done with you. He’ll have nothing else to talk to you about the rest of the time except to say maybe, ‘Hey, thanks again for picking up my red rubber ball,’ but when you play the game again or play with a different character or whatever — if you’re playing multiplayer and someone else approaches him, they might get completely different text when they talk to him.”

These new additions are exciting, but it made me wish there was some way to incorporate some of these ideas into the main story. Wouldn’t it be cool to have a videogame that was tailoring a story to your action just as a GM might in a game of tabltop D&D? When asked about the prospect, JD has some great insight. “I think it’s theoretically possible. I think that it shifts the workload onto the person who’s generating that system you’re describing. Instead of the level designers being able to plug things in where they want them or to set things up so that the random bits mesh together well, you’re putting a lot of this onto the writer so that no matter which direction the characters go or which random event pops in, it will all eventually tie back in again at the end. If that’s what you’re describing, I think it’s still a lot of work; it just shifts who does that work.”

Gaming is in a transitional period in which narrative is taking on a much larger role than it has in the past. That usually comes in the form of crafted stories. In the wake of this change, it’s refreshing to see a developer like Torchlight continue to push in other directions, namely that of procedural content generation, all the while still conveying an overarching story. “There is an overarching storyline and it is not randomized,” says JD. “So, everytime you play, you will be on that main storyline quest if you follow it all the way through, but there will be side quests and quests revolving around landmarks and event quests and quests that involve the lore of the world.” There is an undeniable appeal of a singular story created by a writer, but the real draw of the game is the way it randomizes bits of story. Random elements in games have been around since Rogue hit in the 1980s, but the approach still has merit. With Torchlight II’s refined approach to randomization, everything old is new again.

The automobile is a mystery to me. I know what one is and I have a very basic understanding of how one works (the emphasis here is on the word “very”), but my understanding is not nearly nuanced enough to begin to fix one or to figure out why it might be broken. In general, I get around this by doing heavy research before buying a car to be sure I get the most reliable one available. Unfortunately, even the most reliable cars inevitably break and need fixing — fixing I can’t begin to deliver. It’s at this point I usually lament my lack of knowledge and start digging into my savings account making out the check to my friendly and trustworthy (albeit overpriced) mechanic.

In the case of machines we come to rely on, there are definite advantages to knowing as much as possible. In the case of videogames, it usually helps me to have much of the inner workings hidden behind a thick curtain of mystery and wonder. I was reminded of this sad fact soon after beginning my third attempt repaying my father’s loan in recent indie darling Recettear.

In order for you, dear reader, to understand fully, I need to tell you a bit about the game — give you some much-needed context. In Recettear, you manage an item shop to earn money used to repay the gargantuan debt left behind by your father. If you make all the payments, you win the game. If you miss one, you restart the game with your current level and inventory intact. No one wins on the first try. At least, no one wins on the first try without help.

By the third go ’round, I was ready to put the game to bed. It’s enjoyable, but I can only take so much of theGroundhog Day-ness of it (read: not a Dead Rising fan). With about 20 hours logged and two attempts blown, I resorted to searching out strategies online. I found some excellent information… and by “excellent,” I mean “thorough”… and by “thorough,” I mean “terrible.”

The game has been available for a matter of weeks. This is the only fact you need in order to understand that the Internet has already discovered every quirk, every eccentricity, and every bit of minutiae anyone could ever glean from it without combing through the source code line-by-line. The Internet is working on that as we speak.

Said strategies told me exactly how to make customers happy with the precise selling prices necessary to gain their trust. Then, they told me when and how to take advantage of this newfound trust to gain maximum profits. That sounds familiar. Gain their trust, then take full advantage when they express gratitude. Hmm. I felt like I was back in sales training at Comcast. This is not a good thing.

Up to this point, the game felt really dynamic. The shoppers obviously had different thresholds for how much markup they could tolerate, but I had yet to discover exactly what those were. I had noticed the little hearts which indicated the shopper was very happy with the price, but I didn’t really even know how (or if) this affected the game. It was the simpler time in the relationship before you discovered she pees in the shower and likes erotic fan-fiction based on The Simpsons. I was enjoying the process of slowly uncovering the game’s mysteries.

This new information armed me with more than enough to complete the game on my next attempt… but I never did. Now that I knew how simplistic the game’s wants and desires were, it was really more joke than game. It was immature like a baby — a simple little bundle of wants and needs without the tools to properly convey them. Sadly, my love affair with Recettear had ended.

The game’s designers are not to blame for this. I accept full responsibility for ruining the game for myself. Every mechanic has to have limits. The game is working inside a box made by its developer. Aside from the odd patch or update, that box doesn’t really change once the game ships. Even the most sophisticated systems with multiple logic branches and if/then statements nested deeply enough to accommodate the largest of fowl still has limits. (Get it? You see, I’m using “nested” as a pun here, and that’s what a bird… nevermind.)

Every game has these limitations. We human beings live every day in a world of almost infinite possibility. We want our games to capture as much of that possibility as they can… and they do. Sadly, that isn’t much.

Each choice and branching decision we can make in a game is a developer sitting in front of a computer saying to himself, “Maybe we should let the player do this cool thing!” He codes for a while, and — presto! — we have a facsimile of a tiny choice we might make in real life available to the character, our connection to the virtual world.

Games are sold based on what they have rather than what they don’t. Being proper enthusiasts, we gorge ourselves on all the information leading up to release. We bring home the new game excited about the possibilities. To return to the box metaphor, once we have this box inside our home, it’s hard not to see the box’s dimensions, regardless of how sexy the box is.

The walls of the box are little disappointments, little things we try that didn’t quite work as we had hoped; they are options we would have had in life (if life were like the world of the game) that we now discover are missing.

Some games are worse than others. Perhaps they show you an obstacle that would be easy to jump over, but they didn’t give you a jump button. They show you buildings which are meant to be real, but you cannot enter and explore. Some games shine their light elsewhere and are less concerned with possibilities.

I was never bothered by my inability to jump in Mass Effect. OK, that’s not entirely true… there were probably times that was a little annoying, but since the game is focused on conversation and narrative, it wasn’t really a big deal. Now, when I wasn’t able to say exactly what I wanted to say in conversation… that is another issue entirely.

In games like Recettear where much of the focus is on one or two mechanics meant to compose the fun of the game, the limitations are front-and-center. Haggling in life is dynamic, interesting, and, if you have the proper disposition, fun! I can make an offer and explain why, playing to their sympathies. They may refuse but later call me up and say they decided to accept. The motives of both parties are complicated as are their criteria for an acceptable offer.

On the surface, Recettear appears to have captured many of these traits. As I said, I played through more than 20 hours and was enthralled. Of course, in the back of my mind I always realize there are limitations. They don’t have an army of dudes (and dudettes) sitting around haggling with me on the back-end.

Part of me just wants to forget. I know a computer program has more limitations than life, but I want to get lost in the simulation. My only hope is that the simulation is deep enough that I never actually butt heads with the limits. Ignorance is sometimes awesome.

Once those limits are apparent, having been revealed to you either through play of the game or from above (i.e. the Internet), the game loses something. In a game like Recettear where a big piece of the fun is in haggling with shoppers and trying to figure out what makes them tick, losing that sense of mystique is damning. The game is about that discovery. I hope, for your sake, you acquire that omniscience through play. If so, you have gotten from the game what the designers intended. Take the information when its handed to you on a silver platter, and — trust me on this — you’re only hurting yourself.