So, last time, I was talking about the complexity of actual combat–particularly space combat–and ended by allowing how it was understandable that games simplify it. Good board game design means simplifying as much as possible while keeping it interesting, while good computer game design can allow tremendous complexity but needs to still be accessible so that a player can puzzle out good choices from bad.
Take a game like Risk: Military battles on the scale modeled in risk are vastly complex things as well, but reduced to the players rolling (no more than) three dice at once. Fair enough. That keeps things at the strategic level. First edition Dungeons and Dragons took the entirety of martial combat and reduced it one 20-sided die roll–though with substantial rules behind that. 4th Edition D&D is basically geared toward detailed, tactical combat (incidentally signaling its demise as a storytelling medium).
But 4th Edition D&D illustrates perfectly the struggle of the game designer: You want to allow free action on the one hand, but on the other you have to make actions which are compatible in order to compare them meaningfully, which in any system, requires restricting actions. 4th Edition answers the question of Batman vs. Superman by saying, “Superman–but Batman shouldn’t encounter Superman until he’s leveled up and has comparable powers”.
Yet everyone knows that whenever Batman and Superman clash in comic book fiction, Batman wins. Because the alternative is boring and predictable. In fact, Superman almost always loses to every good guy. He has to: He’s invulnerable. That’s almost a literary necessity. It is a literary necessity that the battle be difficult, or you have no story. (Intriguingly, combat gaming must also be difficult, or you have no game!)
But can you have a game system that can make it possible for the Batman to win, simply by being smart? Can you have a game system that allows Frodo to defeat the big boss?
This feeds into the starship combat discussion, at a slightly different angle: Can you have a game system that allows for a wide variety of variants so that creativity is endlessly rewarded?
Let’s look at StarDock’s highly regarded Galactic Civilizations II, a game that I’ve played far more than I’ve actually enjoyed (for complex reasons best left for another time). GalCiv–a space-based form of Civilization–actually goes back to the OS/2 days, but in its latest incarnation has the option to “act out” space battles.
GalCiv, much like Civilization 4, takes a “rock-scissors-paper” approach to combat. But before we get to that, let’s look at what the next step might be past “roll a die/higher score wins”.
Obviously, the “higher die roll” is boring. It works in Risk because Risk is a strategic game, and the tactical simplicity gives just enough randomity to make the strategic game something more than a decision tree.
If we go back to Civ 1, we see the next level of complexity (or one potential level of complexity). In Civ 1, (if I recall correctly) units had an attack strength and a defense strength. Comparing the attacker’s attack to the defender’s defense gave the probability of success. Let’s say any unit has a 50% chance of successfully destroying another. Now, let’s say the attacker has an attack value of 6 while the defender has a defense value of 4. We can do the math like this:
6/4 = %150 * 50% = 75%
So, an attacker has a 75% chance of success. The problem with this approach is that it gets dull, fast. If the attacker has a value of 8:
8/4 = %200 * 50% = 100%
So, in any situation where the attacker’s attack value is double the defender’s defense value, the attacker always wins. This isn’t the way Civ actually worked, as it was famous for having the occasional stone-age spearman beat the occasional modern tank.
A better approach is to improve the odds without ever making them zero. So, if we have attackers with varying attack values of 6, 8 and 800, we could calculate the chance of failure as:
½ * 4/6 = 1/3
½ * 4/8 = ¼
½ * 4/800 = 1/400
Not bad. This sort of system allows us the opportunity to reflect a wide scale, and it works pretty well for a strategic game where technology allows creation of units with higher attack and defense. This single dimension of attack/defense allows us to evoke a battle between swordsman (high attack) and spearman (high defense). It also creates the user-infuriating situation where a spearman–maybe one time in 400–beats a tank.
The next level of complexity, for Civ games, was to add hitpoints. Instead of one “die roll” to determine the fate of a confrontation, one die roll determined whether a point of damage was done. So, where our little spearman before had a one in 400 chance to destroy the tank, now it has a one in 400 chance to cause a point of damage. And the spearman might have one hitpoint where the tank has 20, and his chances of success go down to 1 in 8000.
If we go back to our Batman v. Superman battle, we can see Batman’s in serious trouble. Superman is at least a tank, and Batman doesn’t even carry a spear.
One of the best implementations of attack/defense/hitpoints in tactical game was the combat sequences in the Heroes of Might & Magic series. In these classic games, a stack of monsters would attack another stack, but instead of chance to hit, the monsters had a range of damage they would cause, and the attack and defense values were used to scale that damage. The stack itself functioned as another layer of hitpoints, only one that didn’t regenerate after a battle.
This gave Heroes tremendous opporutnities to provide interesting combat. For example, the genie (in HOMM I) had the power to halve a stack (which was huge, and overpowered, but not dull). The top troll in any stack could regenerate (I think). Ghosts (in HOMM I & II) added to their stack for every creature in another stack they killed. And so on.
But HOMM added something else to the mix we haven’t talked to, because the combat was highly tactical, and not the sort-of strategic/tactical mix of the Civ games: A combat field. In HOMM, the combat field was like a chess board, and the two sides were placed on opposing sides of the battlefield.
So, the first thing an attacking monster had to do was to close on an opponent it wanted to attack. This offered huge tactical possibilities: Range attackers didn’t have to close (but in version 3, they did half damage if their targets were far away or had cover); flyers could cross the board in a single swoosh (in version 1, in later versions, that was only true of some flyers); regular melee attackers had to walk across the board before they could attack. A range attacker also did half damage if a melee attacker closed on it, which gave you the tactic of barricading your ranger attackers with melee attackers.
When you factor in speed of movement and range attacks, poor Batman is toast. Clark fries him from space with x-ray vision.
HOMM, being tactical, also had multiple types of troops engaged at once, which gives us another new attribute to consider: Initiative. (Who goes first?)
Now we have attack, defense, hitpoints, speed, range and initiative. And we’re still dealing with very broad abstractions. We also create the possibility of ships that can attack and not be attacked. (A ship with a high enough speed and a long enough range attack can attack and stay out of range of a slower ship.)
If a system can’t produce a disastrously unforeseen outcome, it’s probably not very interesting.
Next time we’ll look at how we could build a space combat system with these ideas in mind.