Cooperative games

I recently listened to episode #16 of the “Ludology” podcast, about cooperative games. As usual the discussion between Ryan Sturm and Geoff Englestein was quite interesting. And it made me reflect again on a cooperative game I designed recently and played several times, but which I put aside because it doesn’t work suitably.

The game is about up to four players representing star faring nations defending themselves against the attack of massive war-machine intelligences whose only objective is the destruction of all life. You may have read Fred Saberhagen’s “Berserker” series that depicts such a situation. My game is a wargame where the players are trying to hold off the annihilators long enough that their research will give them an upper hand and enable some of their planets to survive. The very powerful annihilator units are face down for “fog of war”, and their movements are regulated by a deck of cards. Toward the end of the game as the players’ units and mobility improve tremendously thanks to research, they have lost so many planets that they cannot maintain many units. It’s a race to see whether enough planets can be saved and enough annihilators destroyed to save humanity.

A key part of the game is four annihilator bases at the outer edges of the concentric-circle board, bases that produce more annihilators. The players have to eliminate at least some of these bases, and ultimately all of them, while still defending enough planets to support enough units at the end of the game. They have to devote enough economy to research, while still having enough units to slow down the annihilators and destroy the bases. Inexorably the annihilators move inward and destroy human planets. If the humans win a battle, there are always more annihilators coming if the bases are intact. If the humans send strong forces to the bases, will they have enough to slow down the annihilators?

I planned to include a set of event cards to provide more variety but my main purpose in testing was to see whether the idea would work.

And it does work as a model of the situation, but as I said not suitably for a commercial cooperative game. The biggest flaw is that there’s no “momentum” toward a finish. In a close game the balance between the remaining annihilators and remaining humans could be such as to produce a near stalemate. This is poison to a cooperative game. People want the game to come to a tense climax and then end. Yet the theme of the game does not encourage a time limit or a believable way to maintain the momentum. I could introduce a Deus ex machina research track that would automatically save the humans when it was reached, but I don’t much like that idea.

Another flaw is that all of the players are doing the same things. In a good cooperative game you want each player to be doing different things so that one idea or person will not dominate play. If each person is doing different things and is individually accountable you’re less likely to get the situation where one player is essentially persuading everyone else how they should behave. In my game each player is accountable insofar as the success of his nation is concerned, but he’s doing the same kinds of things as every other player.

Another flaw, which is very hard to avoid in the game without intelligent opposition, is that the game is always essentially the same puzzle. In the first few plays a group of players is learning how to solve the puzzle, but when the puzzle doesn’t change then the flaws in group dynamics for solving problems begin to appear. I read long ago, and cannot specify the source, that if a problem is well-understood, for best results you should assign one person to solve it rather than a group. As a trivial example, one person should add a column of numbers, not a committee! If it's a poorly-understood problem, then a group of people will do better than one. The first time people play a co-op the problem is likely to be poorly understood. But after playing some times, the players understand the problem and may begin to lose interest, or to interfere with each other in undesirable ways.

Event cards could help vary the circumstances and increase the amount of plays before this “too well understood” situation occurs. The only way to really solve this is to have intelligent opposition. RPGs are different-than-boardgames cooperative games, and much better ones, because there is intelligent opposition (controlled by the referee). The traitor mechanism works well in otherwise-cooperative boardgames because it provides some intelligent (albeit hidden) opposition.

Computers can be programmed to provide a semblance of intelligent opposition, board and card games cannot. A game with nothing more than card "programming" to provide opposition is definitely a puzzle, and is likely to be solved with just a few plays.

And of course, where there's human opposition, the problem can and likely will change over time, so "this problem is well-understood" does not apply as strongly.

I don’t think people want to play very long cooperative games, either. And my game, even if it had momentum, appears to be 2 to 3 hours

There are ways to avoid one of the big problems of cooperative games, that they are essentially solitaire games and one person of a group can end up being the main player with the others as followers. One method is to limit the communication between the players; another is to include a time limit so that no player has enough time to figure out what he’s going to do and also advise everybody else. In conjunction with different functions for each player this may work pretty well, but we still have the problem that the cooperative “game” is a relatively simple puzzle, not a game, and it will soon be solved.

I have had in mind for many years, and have designed the mechanics for the non-cooperative part, a cooperative role-playing game that does not require a referee. I’ve even tested it a little bit with myself acting as the referee. But I have no confidence that I can devise a set of cards to control the opposition that will be sufficiently interesting.

During the Ludology podcast the guys talked about the necessity that players buy into the theme of the game and not treat it merely as a set of mechanics, not treat Shadows over Camelot as just poker hands. My game, being a wargame-model, doesn't appear to have that problem.

Another technique they discussed was a game where there is only one winner instead of the group winning, yet the group must cooperate sometimes or everyone will lose. The flaw there is a player who regards everyone-losing as just as good as everyone-winning, and so will hold the others to ransom by threatening to cause the loss. I do not care for this technique; in effect with this technique you're providing a form of intelligent opposition from the players themselves, but I prefer the traitor method so that the roles are clear cut, rather than"one wins but everyone can lose".

They also talked about the desirability of co-op games with one way to win but many ways to lose. Again in sync with my theme, there is only one way to lose as the annihilators destroy planet after planet. In a sense it is too linear, even as it models the situation well. But models don't always make good games, as we saw again and again in the SPI era.

Something my game does have is the feeling that you're making progress, because you're researching technologies that help you resist the annihilators, yet things are getting worse and worse as the annihilators move toward the center of the board and destroy planets along the way.

I suppose that, in the end, I am not much attracted to co-op tabletop games without intelligent opposition, because without that opposition the game must be a puzzle, and must lack gameplay depth (though it may have puzzle or story depth, or even model depth).

So the game is "on hold", and likely will remain so forever. Unless I try to rework it so that there is one player for the annihilators, but that kind of "one against many" game must be a nightmare to balance.