In my last post, I talked about the unique challenges involved in designing two-player games. This time, I’d like to discuss the challenges of games for small groups; three or four players is very common, but most of what I’m going to say applies to games with more players as well, except that games for larger groups have additional problems that I’ll cover in a separate post.
As I said in the previous post, there are objective differences between games for two and games for groups. Firstly, whereas “perfect play” is guaranteed to exist for two-player games (even ones with chance and/or hidden information), the same is not true for games with more players. This itself stems from the other main difference, which is that cooperation is possible when there are more than two players. Thus, unless the game has minimal interaction between players, a player’s chances of victory are not solely determined by his own strategic choices, but those of his opponents.
This is mixed blessing. On the one hand, it makes balancing a game somewhat easier, as the players who are trailing will tend to gang up on the leader, bringing the game back into balance on their own. Similarly, whereas in a two-player game it is entirely up to the designer to ensure that no single-minded strategy can win, the players themselves can often cooperate to render an overly simplistic strategy unviable by taking advantage of its predictability.
But the first difficulty in designing multiplayer games arises when this seemingly positive phenomenon is allowed to go too far. The end of a game can drag on much too long if it’s too easy for players to bring the leader down. This can be solved through the use of an alternate game end condition, but often this only changes the nature of the problem; if the underdogs cooperate against the leader throughout the game, then everyone’s positions end up being roughly equal coming into the final turn(s), rendering the early game almost meaningless.
There are two main ways to actually solve the gang-up problem. The first is to cut down on the degree of interaction between players, but it’s important not to cut down so far that the game feels like “multiplayer solitaire,” a common criticism. Ideally, unless the intention is specifically to create a political game, a player’s choices should have more effect on herself than on her opponents, while still leaving a significant decision to be made about who to help or hurt in the process of furthering one’s own interests. The second solution is to have a certain amount of hidden information in the game; if it’s unclear who is actually winning, it’s much harder for players to gang up on them; on the other hand, not all games lend themselves to this. Small World and Puerto Rico, for instance, both keep players’ victory points hidden… but since VPs are never earned secretly in either of these games, experienced players can easily keep count; thus, it only actually solves the problem for players who are not taking the game very seriously.
If a multiplayer game successfully avoids or corrects for the problem of excessive self-balancing, however, a new problem arises. If players can successfully pull out ahead without being dragged back by a coalition of their opponents, then by the same token, players must be able to fall behind. In a two-player game, this isn’t as much of an issue, as a forgone game can always be resigned by the losing player. This is not the generally the case in a game with three or more players, however, because a game’s rules often assume that the number of players will be constant throughout. It may in fact be impossible for the other players to continue the game if one trailing player drops out, but even if not, it will often affect the outcome.
Thus, players are typically expected to play multiplayer games out to the bitter end, even if they themselves have no chance of victory, an experience that can be boring, frustrating, or even humiliating. And since games are social activities, if one player is not having a good time, that will often diminish the experience for everyone else. In the worst case, there may even be the potential for kingmaking, where a player who has realized the game is hopeless can still attempt to help a friend win or, more likely, seek revenge against the player he blames for his own spoiled chances.
Games featuring player elimination don’t suffer from this to as great an extent, as players generally continue to have a chance so long as they survive, and losing players are eliminated and can no longer interfere with others’ chances. And, since the games are designed to have an ever-diminishing number of players, players can usually drop out any time they like. However, a large percentage of the gaming public despises player elimination as a mechanic, so this shouldn’t be the default solution.
The problem can also be partly solved through hidden information, though not quite as readily as the hyper-balancing problem. Although players don’t know how their opponents are doing, they know how they themselves are doing, and can usually tell when they’re losing. So this alone is not usually enough.
Really, the only solid way to ensure that trailing players remain interested in a game is to design the mechanics to allow for comebacks, whether through brilliant strategy, dumb luck, or both. Dominant Species does this really well in my opinion, with many opportunities to score massive points towards the end of the game, so that, although the points scored on earlier turns are still relevant, it’s rare for anyone’s position to be entirely hopeless.
Putting all this together, we find that the biggest challenge for three- and four-player games is not so much how to balance the game, but how much to balance it, and how much that balance should depend on the factors of luck, individual skill, and player cooperation.