Co-operative Play

In a games industry obsessed with direct competition, co-operative play is generally overlooked. Despite the odd high profile game featuring co-operative modes or elements, (Halo and San Andreas spring to mind), co-operation is at best seen as an optional extra in game design. At a time when we are hurting for alternative game evangelists; for new ways to drive awareness and interest in new games, is co-operative play more valuable than the industry recognises?

The spur for this piece was twofold. Firstly, showing my wife Shadow of the Colossus this weekend lead to us starting a two player game of Ico (my wife was unsurprisingly not enamoured of a game whose play is essentially crippling giant animals, no matter how beautifully rendered). I doubt we’ll play much of it, because the pace of play in Ico is extremely slow, but it was interesting to explore the co-op mode briefly. Secondly, we finally made the time to complete Klonoa 2: Lunatea’s Veil, a game with a highly unusual asymmetric co-op mode.

What is a co-op mode, or a co-operative game? Basically, these terms are generally used to refer to any game situation in which the primary relationship between the players is supportive rather than antagonistic. The players generally work together to achieve goals and competitive elements (if any) are minimised.

The most basic form of this kind of play is the team game. A typical team game is based around each player having the same capabilities; in essence, the game provides multiple avatars, one for each player. One of the first games to support this style of play was Gauntlet, which back in 1985 re-energised arcade revenues with its canny re-imagining of the basic tabletop role-playing clichés in a game which was one of the first ‘pump and play’ games (requiring players to feed it coins to keep playing). 

Mostly, we see two player team games which we could term partner games - like the co-op mode in Halo, the two player rampages in San Andreas, or the old school style of games like Contra (recently revived in Alien Hominid). In a partner game, it is usually possible for the players to play independently, co-operating only when a particularly difficult problem blocks their path. This makes these games easy to play, but it also means that the co-operative elements tend to be sidelined. The players are essentially still playing an agonistic game – they just happen to have an ally.

Some partner games take this further, usually by exploiting some measure of asymmetry to define separate roles. One example is the explicitly co-operative play of Kuri Kuri Mix which can only be played two players working together (or by a single player with exceptional skills!) The two avatars have identical capabilities, but the game world is bifurcated into two parallel paths – each player works down their own path. Switches and puzzles in each single path allow the player in the other path to advance. The game is entertaining and different, but suffered greatly because it could only be enjoyed with two players. (There are in general terms many more players playing alone than playing together, so games which require two players to be played tend to lose out). 

Ico also works as a partner game, and is asymmetric. Here, the asymmetry is in the avatar abilities: Ico is the more capable character, being capable of fighting, climbing, falling and carrying. Yorda, by comparison, is only capable of climbing and carrying. Sadly, the game camera is intolerably unworkable. Whomever plays Yorda faces an exceptionally bumpy ride in terms of frequently not being able to see their avatar on screen.

Team games in general tend to exploit asymmetry in avatar abilities. Gauntlet has its classic roles of warrior, wizard, valkyrie and elf (archer in more recent versions), for instance. Project Eden, which is structured more like a puzzle-driven adventure, has four characters with radically different abilities as the basis of its play. (Given that the adventure style of play seems to be a good fit to introverted play, the four player co-operative mode for this game has seen little use, which is a shame as it is where the game comes into its own). The Conflict games offer similar play but with gameplay based around simple tactical scenarios, and largely devoid of puzzles. 

Also, team games have a tendency to include competitive elements, perhaps because the development team didn’t know how to design a game any other way. In Gauntlet, competition over food is common, and the ‘It’ levels devolve into direct competition. Similarly, The Legend of Zelda: Four Swords has the players competing for rankings at the end of each level. Presumably, there are some players for whom this is a spur to play – it is doubtful that a co-operative game genuinely benefits from being tied so directly to competition.

More unusual forms of co-operative play can be found. In Klonoa 2, the main player controls Klonoa, who has the usual range of abilities for a 2D platform game. The second player controls Popka, and has one and only one ability: they can rocket Klonoa high up in the air. (After doing so, Popka takes a while to recover). This style of play could be considered support play. The main player does all the work, but the second player has the potential to contribute support to the main play. What is especially interesting about the play of this game is that puzzles and challenges which are very tricky to achieve as a single player become easier in support play. Not to mention that the supporting player can quite frequently save the main player from certain death through careful use of their ability. 

Of course, many game-literate players might be bored by having so little direct influence on play but there are many benefits. The complexity of control for the supporting player has been reduced to a single control – this is simple enough that anyone could do it. Since the main player can call for help out loud, this allows a second player to contribute to play even if they lack any conventional gaming skills. Additionally, the development cost of adding the second player in this way is extremely minimal. In the case of myself and my wife, we switched off pads quite frequently throughout the game because we both had sufficient gaming skills to play either role, but I believe there is potential for the very young or the very inexperienced to contribute to play through some kind of support play.

This particular example also opens the door to another kind of play, one which is rarely exploited: tutor play. It is often the case when a player comes to a game for the first time that they will be taught to play be a second player. In boardgames, this element of play is practically assumed; Magic: The Gathering expects that its rules will be learned in this way, and explicitly references this in the rulebook. In videogames, the tutorial is still provided as a static single player experience. Consider the benefits of including a two-player tutor mode in which one player demonstrates game actions, and the other player mimics those actions. There can be no doubt that a human player is a better teacher than any static tutorial could ever be. 

Tutor modes could also be designed based around support play. Consider that a driving instructor has a full set of controls, but usually lets the student drive. In case of emergency, the driving instructor can seize control and save the day. Something similar can be implemented in a game context, either directly or indirectly. Consider a platform game in which the tutor player controls a jetpack or bird character who can rescue the main player avatar by the press of a button, or a similar mechanic. In this context, the support play of Klonoa 2 and the Yorda role in Ico could be seen as tutor modes – although this is not how they are presented, they can be used in this way.

I’ve been thinking about ways to leverage the tutor-pupil interaction in game design since Katherine Isbister’s presentation on extraverted play at last year’s GDC. There is even an emotion associated with the success of a student (or offspring) – Ekman refers to it as naches – which can be a highly enjoyable emotional payoff. There is surely more to discover in this corner of play, but we will need new game designs to explore it fully.

It is not yet clear how valuable co-operative play is to a game in commercial terms. There is little doubt that it can contribute to a game’s success (Halo in particular is believed to have supported sales of the original Xbox by attracting new players who first experienced the console’s only killer ap by playing co-op with an existing Xbox owner). It is clear, however, that extraverted play (which includes but extends far beyond co-operative play) can drive sales. The success of the console versions of Dance Dance Revolution which were not predicted to have a significant market demonstrates that meeting new play needs can create new niche markets.

I believe that mass market products will have an increasing duty to support co-operative play in interesting ways. As budgets in the upper market rise to insane degrees, the cost of developing additional modes and features such as co-operative play become trivial next to the total development and marketing budget – and since these games need to pull every trick in the book to court a large enough audience to make back their costs, no such game can afford to exclude some element of co-operative play that could expand the appeal of the game to a wider audience, provide a mechanism to introduce new players to the game, and potentially drive awareness of the game to a much wider audience.

Comments

[Patrick]

On the other hand, for a certain project with a budget under $50,000, this rule still applies. A co-operative mode could be overlayed with eas in Fireball, actually more easily than a competetive multiplayer. Instead of having co-operative puzzles, we resort to the latent layer of puzzle solving that exists in almost all the feilds we've made, the combo maximizers. Think of how fun that could be for people, telling each other where to go and when to pull the trigger, giving way to the satisfaction of a high score niether could have gotten alone.

[Jack Monahan]

Hold on--does the European release of ICO feature a two-player option? Certainly the first I've heard of it. Would be nice if I had a special someone to play it with... or at least one of my siblings around.

A pretty good breakdown of the basic co-op types, as well as the observations about the curious niche the gametype seems to be found in. I've always been a very big fan of games that are best enjoyed as a co-op experience. Even the legendarily bad Daikatana had fairly passable co-op action, which is better than none at all.

Pity your wife doesn't appreciate playing SotC, but it's clear that it really isn't her game I guess. That heartbreak is very much at the core of the game's impact; that essential commentary about the role the player character takes and the devestation he he deals against these "boss monsters." It's still of the most thought provoking and emotional gaming experiences I've ever had, which seems especially resonant now when I seem to have very little desire to play games at all.

Still looking forward to a full break-down of SotC!

[Chris]

Patrick: Sanjit and I have to thrash out what we can afford to do with multiplayer modes in Fireball. If we could only afford to do one multiplayer mode, would you favour a co-op or a competitive mode?

Jack: Ico does indeed have a two player co-op mode in the European release; that's the version I have. I believe you have to complete the game once to access it. Start a second game, then pause the game and change 'number of players' to 2. That's what we did, anyway. :)

[Andrew]

'It is not yet clear how valuable co-operative play is to a game in commercial terms. '

Was wondering if you had considered that many MMO games (in particular the PvE kind) are largely co-operative in nature? This would certainly seem to be a validation of co-operative play in commercial terms?

[Patrick]

Its not so much which I prefer, but what the target audience will prefer, and what is actually more robust in terms of re-play value. I say go Co-op. It would be even less painful to implement than competetive, as you'd just accomadate a second I/O path and avatar object, and let the level designs stretch on their own.

That Fireball lends itself more naturally to co-operation than competition is probably a testament to your design philosophy.

[Chris]

Patrick: you make a good case here. My only concern is that many of our fields don't lend themselves to co-op. As soon as I get clear of the book edits, I'm going to have to draft some paperwork for multiplayer in Fireball. I'll run it by the list as soon as it's drafted. Thanks again!

Andrew: I see the case you're making here, but the social dynamics of MM worlds are highly complex, and it would be premature to assign the appeal to co-op play above, say, community membership. I personally suspect the appeal of MMOGs lies in part in the way it provides a safe environments for introverted players to explore being extroverted. Like so much in games right now, more research is needed!

Also, demonstrating the commercial advantages of online play could be taken as evidence in support of online play, instead of co-op play, so it could be a double edged sword.

I did overlook it when writing this short piece though - thanks for mentioning it!

[RodeoClown]

Hi Chris,
another possibility for the multiplayer in fireball is to have hot-seat like in Tony Hawk, where you can play 'HORSE'; one player runs through the level and gets a score, the next player does the same level and has to beat the score. If they don't beat the score, they get a letter (H for instance). Then the players swap, and the second player sets the score to beat. First player to get HORSE loses (or you can specify any word). This is also good if you have only the one controller as you can swap controllers for multiplayer OR if you have more controllers, one each.

You could make it based on ASH, combo size or on speed (or selectable).

If you can add something like this it would be cool - then it's not direct multiplayer and it's not simultaneously competitive.

I love co-op stuff though, so it should have priority :)

Maybe in co-op you can boost the other player - by letting them rise with you if you jump - then they can jump off your 'head'.

Maybe transfer heat - if you touch a hotter fireball you both then have that heat.

Maybe slams can boost the other player's heat level (although this would be prone to problems - just keep slamming till you are both white hot...)

my 3c

-Ian

[Chris]

Ian:

Yes, hotseat play is my fallback position if all else fails. Once upon a time, that was what a multiplayer game was. :)

Your idea to allow the two fireballs to assist each other to jump higher in Co-op is highly tempting! The thing I liked about Klonoa 2 was how a second player could make the difficult challenges that little bit easier. I have to think carefully about the best way to implement such a system.

Thanks for the input!

[RodeoClown]

I think the boosting each other would come from just making the fireballs solid to each other. One can jump and as they float down the other player can just jump on top of them and can then jump higher. easy boosting and probably comes for free.

[Chris]

That should work. It might be hard for players to do this in a hurry, though (although I bet they'd get good at it pretty fast).

I was wondering about an 'invisible cord' solution, where the players can fuse their two fireballs into a pair of mutually linked fireballs (they would not be allowed more than a certain distance from each other). This might require an extra control (for linking and unliking fireballs). Your solution might be more elegant.

On the other hand, if split screen modes prove not to be an option, we might want to automatically link two fireballs into one shared entity so we can keep the (single) game camera focussed on the pair at all times. It would be different. :)

[Adam]

I hope it's okay to comment on something this old...

You say that no big-budget game can afford to not have co-op:
Oblivion has to be one of the most insanely-budgetted games out there. (Just hearing about some of the lengths to which they went makes me think, "Lord of the Rings".) However, the time required (which is something I think you failed to consider in your post) and cost for developing a co-op mode (or any other play-type) would certainly not be trivial. In fact, the developers' main reason for not including multiplayer is the length of time it would take and the degradation of quality that the single-player mode would suffer.

Just thought I'd point that out.

[Chris]

Hi Adam! Feel free to comment on anything here, regardless of age.

You're right, of course - it's not always cheap to add a co-op mode. The point I was making is that in the upper market, development budgets are escalating to insane degrees, and the cost of incorporating a co-op mode to a game is becoming less and less prohibitive as these budgets rise.

(Incidentally, in terms of 'time' I would assert that in modern games development, time and money are interchangeable on a well managed project: time costs money; money buys time).

However, it certainly is the case that certain styles of game do not lend themselves to a cheap co-op mode - in particular, cRPGs like Oblivion. These games *could* be designed with co-op in mind, but the decision would have to be made from the outset, and it would have implications for the design which, sadly, most developers are too conservative to explore.