It’s always nice when two independent developers disagree with each other on design stuff.

Vic Davis of Cryptic Comet, home of Armageddon Empires, recently posted a mini-dissertation on the role that events play in strategy games. He compiled a list of the purpose that events can play and how they can be used to challenge the player.

Events punish (your library is destroyed and your tiles all yield -1 resources per turn) or reward (A hero has shown up at your capital and desires to serve you). Even if the distribution of reward vs punishment is skewed heavily towards punishment, people enjoy this.

Michael Akinde, sole designer of the still in development Imperium: Rise of Rome, wrote a counter post to Davis, arguing that event mechanics interfere with good game design.

In short, events tend to reduce the role of the player as an actor in the game. The thing is – as a gamer, I am trying to play a game, but events are things that happen to the player. There is a pretty significant distinction between those two positions. I suspect that there is a good reason that Sid Meier’s famous quote defining a game doesn’t state “A game is a series of interesting things happening to you”.

Here are some (to my mind) obvious things to keep in mind when considering the place of event mechanics in strategy game design.

1) Randomness is a part of every game. Strategy games are, generally speaking, games in which you make decisions focused on forcing the odds to your favor. Attack with overwhelming force to get better combat results, use counters or terrain to negate enemy advantages, plan an alternate route in case your first choice of cards doesn’t appear. Games are, indeed, a series of interesting decisions, but decisions are only interesting in a climate of uncertainty. But this uncertainty should not be rooted in guesswork about things the player has zero control over or influence on.

2) Events only make sense in a game with a long play time. In a 20 minute RTS, an event that threatens your military viability is needlessly cruel since the player has little room to find an alternative. The unspoken rule of the penalty event is that it should distract but not destroy. Events should happen to every actor on the field, not just the human player(s), and provide similar challenges to each.

3) Events should be contextual, responding to choices the player has made or an environment the player is experiencing. In Civ 4, only people with the Slavery civic can get the slave revolt event. Only those with horses in their cultural borders can get quest events centered on chariot or stable production. Though sometimes things just happen (weather events, new resources, etc.) the player should be able to mentally or materially prepare for bad luck. The obverse postulate is that good luck should not be so easily manipulated.

4) The more complex the system, the less important events are. This gives the designer free reign to make a difficult choice. Will events be used to make the game more colorful, since, in the long run they won’t matter? Or will events be left aside so the player can focus on the underlying mechanics? There is no wrong decision here.

5) If events are used, the causes should transparent after a few encounters. I recently went bankrupt by event in a session of Europa Universalis III and I had no idea why since I hadn’t taken a loan in a century, had 800 ducats saved up and had a fair monarch. Only peeking in the event file gave me any clue as to what the hell was going on. Event text should have pointed to my high inflation rate and large number of gold producing cities.

6) Akinde’s strongest point is about the reliance on events to communicate a game’s theme. “[I]f the random events are the only way in which I can distinguish a game about Napoleon from a game about Caesar, then the game engine needs some work.”. Flavor events should be used sparingly, not as a fall back position to communicate what your game is about. “Mad Max joins your army!” is not the best way to tell me that I am playing in an apocalyptic wasteland.

7) However, if events are the chosen means to communicate a theme like a Medieval soap opera, then have controlling the effects of the events become a strategy in and of itself. You can do this by having cause and effect chains in predictable but uncertain directions.