I had a debate with another user on a board game forum recently about whether it’s harder to design two-player games or multiplayer ones. He felt two-players games were easier and I felt multiplayer games were. In the end, we decided that our disagreement stemmed from our differing definitions of what constitutes a “successful” design, which is another question altogether. But it got me thinking about just how different a two-player game is from a multiplayer one.
There are two main differences: the first has to do with solvability, and the second with the way players interact. Both lead to a situation where the designer is forced to make a choice between alienating one type of audience or another, or attempting to strike a balance, which is easier said than done, and necessitates making certain compromises.
The solvability issue is a wholly objective one: according to game theory any two-player game is solvable, while multiplayer games are not, in general. Although we’re well familiar with this idea for zero-chance, perfect information games (such as checkers, Go and chess), it actually extends to games with an element of chance, simultaneous moves and/or secret information (such as poker), the only difference being that the definition of optimal play has to be expanded to include mixed strategies.
Thus, whereas in a multiplayer game, a seemingly “perfect” strategy can often be thwarted by the other players’ combined efforts, there is no such safety net for a designer of a two-player game; it is certain that a perfect strategy does exist, and if it is too easy to find (whether exactly or approximately), the game’s appeal will fade for players once they’ve mastered it. On the other hand, any good game needs to allow its players some ability to identify good and bad strategies when they see them; if the game’s complexity is such that players can’t identify good and bad moves even in hindsight, then there is no way for them to improve, and the game will feel arbitrary and frustrating. Thus, it is much harder in two-player games to strike a balance between making it easy for players to assess their options, and making it hard to decide which of those options is the best.
The second difference is that, objectively speaking, there is no distinction between positive and negative play. Assuming that the game is guaranteed to have a winner and a loser, anything one does to reduce the opponent’s chance of winning automatically increases one’s own, and vice versa. This is in stark difference to a game with even as few as three players, in which hurting an opponent does not necessarily improve one’s own situation, and sometimes improving one’s own odds has the incidental effect of helping one of the opponents as well. As an abstract example, a move that loses 2 points for oneself and 3 for an opponent is equivalent to a relative gain of 1 point in a two-player game… whereas in a multiplayer game, it puts both you and the targeted opponent at a disadvantage relative to another player who did not lose anything.
This means that two-player games are inherently “nastier” and more confrontational than multiplayer games. This is most easily seen when looking at high-level play in games that work well one-on-one as well as multiplayer, such as Ticket to Ride and Carcassonne. Both these games include the possibility of negative play, i.e. moves that deny scoring opportunities to the opponent instead of creating them for oneself. In both cases, these strategies are employed constantly in a one-on-one game, and more situationally in a multiplayer session, when either the points at stake are too great to be ignored, when the player has nothing better to do, or when the player thinks only one opponent is a realistic threat.
While this isn’t a problem for some players, it does create an atmosphere of antagonism, which is often a problem for couples, who are probably the biggest market for two-player games. Fortunately, a game’s theme can create the illusion of difference between positive and negative play in two-player games, and mask the fact that one player’s win inevitably comes at the expense of the other’s loss. Unfortunately, it’s generally the case that non-interactive mechanics are often the ones that feel “nice,” while interactive ones feel “mean.” Thus, many games that seek to be “couple-friendly” end up coming off as “two-player solitaire” by the same token.
Thus, we have two axes to think about when we’re designing two-player games:
Unclear Strategy <------> Clear Strategy
Light Interaction <-----> Heavy Interaction
If we look at historical two-player games, chess being a prime example, they mostly avoid these two issues by simply choosing one extreme on each scale and thus appealing only to a very specific audience; for instance, chess is clearly very difficult to solve, but also hard to learn, and about as mean-spirited a game as exists. Thus, only people with appropriate temperaments play it. At the opposite end of the spectrum are games like Parcheesi, whose designs appeal only to children and the most casual adult players.
Of course, it’s still possible to pick one side and embrace it, but in the modern world of commercial games, if you’re intentionally going to restrict your audience to a tiny section of the market, the game had better be extremely good. Rather, the modern ideal is to strike a balance on both these two axes, so as to appeal to as broad an audience as possible, and create a game that hardcore and casual players can enjoy together, that parents can play with their kids, that can be enjoyed by first-time players, yet played in tournaments, and so on.
But in the end, any attempt to move towards one end of one of these axes results in a corresponding loss of the advantages of the other. There is no escaping from this need to compromise when designing two-player games. By contrast, multiplayer games circumvent these issues by allowing the players themselves to cooperate to thwart “perfect strategies,” and allowing room for positive interactions between player as well as negative ones. On the other hand, they come with their own share of problems, which I’ll talk a bit about in the next post.