One game that bears some similarity to Torch is The Legend of Zelda: Four Swords Adventures for the Nintendo GameCube. We haven’t had the opportunity to play this one yet, but we have studied it a bit to see what they do right and what we could improve upon.

LoZ:4SA takes the classic Zelda formula and turns it into a four-player cooperative experience. The game depends on a special link cable that turns a Game Boy Advance into a controller for the GameCube, granting each player their own local screen, while the television is used as a shared screen for all players. This allowed individual players to wander off from the group, with gameplay transitioning from the TV to their local screen and back.

In terms of hardware setup alone, this might be the closest thing we’ve seen to the game we are creating. It gives us a whole set of multi-screen design decisions to compare our game against.

The first thing one notices when watching videos of the game is how rapid and chaotic it seems. It’s very difficult for one person to follow the action on all five screens. However experience with the game will reveal that each player only ever needs their attention in one place unless they really want to check other screens. When a player character is on the main screen, their local screen only shows a message saying “Look at the TV!” (There are occasional callouts on the main screen to alert a player to a message on their local screen.) When they leave the main screen the only need look at their local screen. Thus, there is not much need for divided attention in this game.

The designers also made an interesting choice by giving each player a “picture-in-picture” bubble on the main screen when their character is out of sight. This gives players a clue to each other’s activities without looking at one another’s screens, and it makes it easy to tell if an ally has gone indoors or is just hiding behind a tree.

The game contains lots of small puzzles and obstacles that require a synchronized effort to overcome, such as rocks that must be pushed as a group or pressure plates for all four players. For now, we’ve decided that these challenges are not appropriate for our game, because we want it to be drop-in drop-out friendly. We may take this design into account for the boss encounters though.

It also features a ranking system after each level that scores players on several game statistics (it’s very transparent about it). It also lets players secretly vote each other up and down according to how nice they were, which influences the final score for each player a bit. It’s a nice touch that encourages friendly play.

While there are many key takeaways from this game for us, it’s a more open experience than we intend to create. Torch is designed with a different feel in mind, so that while we want the game to be fast and fun, the constrained dungeon setting is essential to our experience. That also means limiting the items that we have in the game, and thinking about our controls in a whole different way – Zelda would be very different on a touchscreen. Finally, we want to get the players using both screens as much as possible rather than only moving between one and the other. It’s a tough design challenge – which is why we’re looking for good examples like this one to solve some of the challenges for us.

One topic that came up very early in our research process was Nintendo’s Wii U. The new console (set to release in November) prominently features a large touch screen in the controller, which enables two-screen gaming in the living room. A handful of launch titles are already being advertised, and we’ve been taking a look at how they are using the second screen.

We’ve watched gameplay trailers for ZombiU and Arkham City (trailers may contain mature content). What we’ve seen are a few common categories of interaction with the second screen.

• Mini-map: This is something we’ve seen on the DS quite a bit. It frees up some screen space, but doesn’t add much to gameplay.
• Inventory: It’s fairly common to use the second screen to host menus, which is appropriate for a touch interface. This also allows the player to pull up menus without pausing the game. Usually this would create a divided attention problem, but it’s used to great effect in ZombiU to create a tense situation.
• Look-thru: Several games have demoed a kind of augmented reality where the player holds the gamepad up and “looks through it” to get an alternate view of the TV. This is used for sniper rifles, or scanner views like Metroid Prime’s different visors. This has some potential, but curiously it’s kind of like having one screen again.
• Alternate perspectives: This is a variant on the Inventory interaction, but it feels different. There’s a great moment in the ZombiU trailer when the player is trying to enter a code on a keypad. The keypad comes up on the touch screen, but the TV view reverses to show the slowly approaching zombies. It gives the player useful ambient information on the secondary screen (in this case, the TV) that also increases the tension.

These interactions are interesting, but not as diverse as we expected. We’re wondering what other two-screen interactions will arise during the console’s lifetime.

We also did a quick comparison of the Wii U’s capabilities vs. a phone+TV setup.

Form Factor
Wii U wins this one – their gamepad is designed to be held in two hands for long sessions, where phones are not.

Dual Screens
Wii U’s screen looks pretty good, but it will be hard-pressed to rival recent phones and tablets.

Touch Control
Hard to judge before the system is released, but touch control has gotten very good on phones, and is also very software-dependent. Availability of a stylus might be an advantage of Wii U, but some tablets come with a stylus too.

Front-facing camera and mic
Phones are likely to have better quality.

Multiplayer
This is a major bottleneck for Wii U. It sounds like the system will only support one gamepad with touchscreen at launch (maybe a maximum of two eventually), but in theory we could connect up any number of phones and tablets for multiplayer. This difference is one reason we opted to create a multiplayer game.

That’s what we think about Wii U. We hope to set our game apart with its multiplayer that requires many screens, and a more unique use of the phone’s capabilities.

Yet another game we tried to inform the design of Torch is Dungeons of Fayte by Brent Ellison. Dungeons of Fayte is a co-op dungeon-crawling RPG sim. It’s multiplayer on one screen (and one keyboard), but content-wise it has a lot in common with our design. This game was recommended to us by an EA employee after they saw the direction our design was taking.

Right away, one of our favorite parts of Dungeons of Fayte was its sense of humor, and we discovered that humor seems to work better with other people around. We could take advantage of that by infusing our game with a quirky sense of humor. Dungeons of Fayte also allows you to try different classes that really change how you play the game, and it’s a bounded experience: When the game begins, you are told that the boss will show up in four months. At the end of four months, the boss shows up, and whether you win or lose the game’s ending is tailored to the choices each player made during the game. It’s a cute idea that really pays off.

Sometimes the gameplay can be frustrating though, whether because swords are missing enemies by a single pixel, or because players keep bumping into each other and making it hard to move around. The game is intentionally unforgiving, with very little healing available (until later in the game) and you only collect gold in dungeons, which seemed a little unsatisfying to us.

Since our game has a very casual audience we decided we should go the opposite direction on several of these gameplay issues. We wanted hitboxes to be forgiving so that swords work the way people expect, and we want the punishment for death to be light, like a party game. Finally, we’ve starting thinking about how we can acknowledge the player’s play style, to help create a nice ending for our own experience.

Screenshot from Dungeons of Fayte launch post here.

Another game we looked at while doing research for Torch was At A Distance by Terry Cavanagh. It’s a two-player, two-screen puzzle game that’s free online and doesn’t take very long to play. The players move around two separate spaces in this game, but they really do have to work together to proceed. Figuring out exactly how they work together is the heart of the experience. Where Artemis was an Information Filtering game, we categorize this as a Divided Information game where each player has a piece of the puzzle and they have to fit them together.

At A Distance isn’t a perfect model for us to follow though. The players solve the puzzle together, but it’s actually hard to communicate verbally about what they are seeing – the designer recommends a set-up with the two screens side-by-side so that both players can see both screens, and this is a very different cooperative experience than the one we are creating. The game also requires some simple 3D platforming which, while not always bad, isn’t ideal for an audience that includes non-gamers. Finally, the instruction-less puzzle experience, while fine for At A Distance, is not the right approach for us. We weren’t even sure when the game ended, and those moments of confusion at the end of the experience are not what we want to leave the press with.

So we came away from this game certain that our design needed a clear beginning with a clear objective, and a clear ending rewarding the player for finishing the experience. We also are leaning more and more toward 2D controls and navigation, which may be easier for our audience to grasp.

Screenshot from At A Distance main site here.