Sunday, November 18, 2007

Design Flaws to Watch out for in Multiplayer Games

Digital game development students aren't used to thinking about the consequences of games involving more than two opposing interests, because most electronic games include only two sides, one often a machine opponent. Several problems named by boardgamers can occur when there are three or more sides in a game. Many of these are much more likely to occur when the victory condition amounts to "wipe out the opposition":
• Turtling
• Leader bashing
• Sandbagging
• Kingmaking (petty diplomacy problem)

Turtling occurs when a player sits back and builds up strength while others expend theirs. This can often be seen in multi-player online RTS games. When there are more than two sides, a player can hang back, building up bases and technology, while he lets other players slaughter one another's forces. Then he comes out and cleans up the remainder.

A general solution is to use a different victory condition. E.g., capture of certain locations as the means of victory forces players to come out of their shells. Giving points for destoying the opposition also encourages aggression rather than turtling.

Another solution is to provide economic incentives to be aggressive. This often involves capturing economically valuable areas, so that a successful aggressive player can build up forces faster than the turtle.

Leader bashing tends to happen in games without much hidden information, that is, it must be clear who the leader is. Then the other players gang up on the leader. ("Of course", many would say, why wouldn't one try to weaken the leader?) If it isn't clear who the leader is, this is less likely to occur. If it is hard for some players, at least, to affect the leader in any given situation, then there will be less leader bashing, as those players will distract the ones who can affect the leader.

Sandbagging is often a consequence of leader bashing. A player will try to get himself in second or third place, rather than first, so that when the first place player is bashed, the sandbagger can swoop in for the win. Timing, obviously, is quite important here.

The solution to sandbagging is to reduce leader-bashing to a reasonable level.

Kingmaking is a consequence of what R. Wayne Schmittberger calls the "petty diplomacy problem". Where there are three interests, and one recognizes that they/he cannot win the game, that loser may be able to determine which of the other two wins. Even if the game is being played by more than three, it will often come down to three major interests. More generally, if a losing player can determine who wins, you have kingmaking in play.

One way to avoid this is to structure the game so that a player cannot be sure he is going to lose until it's too late for him to become a kingmaker. Of course, some players believe kingmaking is the "wrong way to play", that every player should try to win no matter what. But designers cannot rely on players to be self-governing in this way.

Another way to avoid kingmaking is to make it too hard for a player to use all his capability against another to prevent that other from winning. As a simple example, in a race it's usually hard for a losing player to have much effect on the leading players.


Now here are some alternatives to a victory condition of "kill everyone else". These help mitigate some of the problems we've been discussing. These are:
• economies (especially zero-sum)
• points
• missions

Economies. Players receive more assets as the game progresses, in accordance with some rules relating to locations or resources, not merely to a table of additional appearances. If a player plays well, he will earn more new assets than if he plays badly.

In a zero-sum game, each player's gain is another player's loss. The classic game Diplomacy is the best example of this. There are 34 "supply center" locations on the board. A player gets one unit (army or fleet) per center. If a player takes another's center, the first is going to increase his forces, while the second will lose forces, at the next building period.

Points. Players earn points for certain events or achievements. This could be capture of certain locations, destruction of enemy assets, holding certain places at given times, and so forth. In a wargame, a player could be wiped out, yet if he's done enough beforehand he can still have the most points to win the game. In general, where points are concerned the game does not continue until all but one player is wiped out. Either there will be a time limit or a point limit.

E.g., in my "light wargame" Britannia, players receive points for holding areas, occupying areas during a certain period, for dominating regions (king of England), for forcing nations to submit, and even for killing enemy units. A nation may be wiped out in the course of the game, but each player controls several, and the points that defunct nation earned still count. Points are based on historical performance, and are accumulated at different paces, so the current score is not a good gauge of who is actually winning the game.

Missions. This is a form of points because the mission involves completion of particular goals, but when a mission is completed the game is over, so no point record is needed. A mission can be as simple as capturing certain cities, or much more complex. Occasionally the missions are hidden, that is, you don't know which mission your opponent is trying to fulfill.


Now let's take Risk as an example. Risk is not a particularly good game, but a great many people have played it, and it exhibits most of our design flaws.

In Risk the object is to completely wipe out all competition. It uses economy to try to avoid the four problems. You get extra armies at the start of your turn if you hold an entire continent, to provide an economic incentive to attack. There is also card acquisition: you must take a territory in a turn in order to get a card, and matched sets of three cards gain you large numbers of armies. You also get armies according to the number of territories you hold. If you turtle or sandbag you get fewer new armies than your competitors. In fact, it's typical for players to attack as much as they can until they're out of spare armies, in order to limit how many territories their opponents control (and consequently how many new armies the opponents get).

There is certainly leader-bashing, but some players may not have forces near enough to the leader to do any damage. You are often better off wiping out a weak power rather than attacking the strongest, because when you wipe out an opponent, you get his cards, and if you can make another set you get more armies (in increasing numbers) with which to immediately continue attacking.

Kingmaking is also quite limited, as by the time a player realizes he's a goner, he doesn't have enough force to do much damage to one of the leaders.

Despite all this, a couple decades after the original English edition of Risk was published, "Mission Cards" were added to the mix. Each player receives one with a mission unknown to his opponents. A mission might be something like "Control Asia" (the largest continent). Hence a player can win the game, by completing his mission, long before he wipes out all opposition. Unfortunately, the mission cards aren't modified by the number of players, so some may be much easier to achieve than others in certain situations.

(Another well-known board wargame, Axis&Allies, is two sides even when there are five players (Germany and Japan on one side, Britain, US, and Russia on the other), hence not subject to these problems.)

No comments:

"Always do right--this will gratify some and astonish the rest."Mark Twain
"A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away." Antoine de Saint-Exup'ery

"Not everything that can be counted counts, and not everything that counts can be counted." Albert Einstein

"Make everything as simple as possible, but not simpler." Albert Einstein

"The worst form of inequality is to try to make unequal things equal." -- Aristotle