Following our previous prototyping endeavor, which involved conveying the feeling of exploring and searching, we decided to try and adapt the combat mechanic in Shadow of the Colossus into the format of a card game.

We kept the game two-player to reflect the conflict between hero and colossus. In addition, we weighted the rules so that the colossus would have more cards but fewer opportunities to use them, and the protagonist would have fewer cards but could cycle through them quickly and play them at a higher rate.

The idea was to establish an even playing field that demanded different tactics from either party: in the digital game, the colossi are strong but ponderous, while the player is comparatively lithe and agile.

In our first iteration, the "hero" was dealt a hand of five cards and a deck of 13, and the colossus began with a hand of ten and a deck of 24. The colossus would put the game into play by placing five different cards in numerical succession (though not necessarily consecutively) on the playing field. Each player would then attempt to complete a straight using the cards on the field in addition to their hands.

For example, if the colossus put down King, Queen, ten, four, two, then the players would simultaneously rush to complete the straight from King to two. The protagonist had the option of playing any card he liked (as long as that card had not already been played), but the colossus could only play cards that directly preceded or followed a card already in play.

In this version of the game, whichever player completed the straight would win all the cards in the straight, and would then shuffle them into his deck. The overall winner would be the first person to acquire all 52 cards in this manner. The cards belonging to a player thus represented a slew of concepts - health (since you would lose cards if you lost a round), elevation (since losing was also a metaphor for falling off a colossus), and advantage (since gaining all the cards would equal a victory).

When Matt played through our game, he questioned the idea of having the cards represent all of these gameplay aspects. By trying to contain so much meaning in the cards, we ended up obscuring some of the concepts that we were trying to illustrate. Clearly having more cards in your deck was better and meant that you were winning (the "advantage" aspect), but the relation to climbing or elevation didn't come across. The cards became a "winningness meter," as Matt termed it: not a reflection of any particular aspect of Shadow of the Colossus, simply a task in which performing well translated into winning.

In the second version of the game, we decided to include a separate "health" mechanic that would be closer to that in the digital version of Shadow of the Colossus. Since the card game phase of battle focused on the climbing mechanic, we developed another phase for the action of actually dealing damage to the opposing player.

The first person to complete the straight would roll a d10 and bet on the outcome. If he rolled higher than the number bet, he would deal damage equal to the bet. But if he rolled lower, he would take damage equal to the bet.

This rule was an attempt to recreate the game experience of taking a risk when stabbing a colossus. In the game, an action button is pressed and held to "charge" strength for your stab. A quick press causes only a very weak attack, but the longer the button is held, the greater the risk of being thrown from the colossus.

We also changed the rules of the card game that represented the climbing phase. In this version, the players battled by playing on different straights: the protagonist would first put down the lowest card in his starting hand and build upwards, and the colossus would put down his highest card and do the opposite.

In addition, play was no longer turn-based: players could put down and draw cards as soon as they saw the opportunity to do so, without waiting for the other player to play. Finally, the acquiring-cards-to-win mechanic was discarded--players no longer had separate decks, but drew from a central draw pile, and the first to complete their straight got a shot at dealing damage to his foe in the die-rolling phase. The loser was the first player to have his health (represented by a stack of poker chips) totally depleted.

An anonymous play-tester was brought in at this point to get a fresh perspective on the game. After a round or so, the tester pointed out that even though a real-time system made the game feel more like a frantic struggle, Shadow of the Colossus also has many aspects that are turn-based. Climbing a colossus often falls into a pattern of struggling up the colossus as fast as you can for a few seconds, then pausing and hanging on for your life while the colossus tries to shake you off. When it's done shaking, you can continue climbing and the process repeats - essentially, you and the colossus take turns. Curiously enough, our game was inadvertently turn-based as both players would politely refrain from taking a card if their opponent was also reaching for the draw pile.

Our solution to this issue was to have the players draw from either of two central draw piles, but discard into their own discard pile. This prevented any lulls in play in which neither player could do anything. Players were also allowed to draw from their opponent's discard pile, motivating each to keep track of his opponent's progress and only discard cards the opponent couldn't use. By forcing players to remain aware of the enemy's behavior, the game felt more like an actual physical struggle.

Matt played through the final iteration of our game and was happy to see that we had made our level of abstraction more concrete and specific. He was also interested in some of the elements that we had eliminated by trying to streamline the game. For example, we lost the clumsiness generated by two players trying to desperately draw cards from the same deck. Having just one deck made the game awkward, but the emotional experience was truer to the spirit of the videogame. Making the card game more urgent also removed any traces of what Matt called the "epic tedium" of scaling a colossus.

Our biggest conflict was between accurately recreating the videogame and making a card game that was upbeat and consistent.