Methods for determining combat results in wargame are either human-based or system-based or some combination thereof.

In the purely human-based method (which tends to be seen more in professional military games rather than recreational ones), the referee or game director decides the outcomes based on his judgment and/or on the learning objectives of the exercise; basically, he functions as a controlling deity.

In the purely system-based method, the outcome gets determined by a mechanism which is built into the rules and over which there is no human control. (Rolling dice, by the way, isn’t human control; it’s human action, but as long as nobody’s cheating it’s not human control.)

There are many potential types of combination methods, but generally they start with a mechanism-determined tentative outcome and then submit it to some kind of human-influenced modification.

Now let’s disregard the purely human-based methods and the human-influenced combination systems, and go back to the purely system-based methods that I mentioned. Fundamentally, there are only two types of system methods: deterministic and random. In a deterministic system, the same inputs will always lead to the same pre-determined outcome. There is no chance involved. An example would be a professional game in which casualties are calculated according to tables, with the inputs being (for example) large-scale force ratios. To completely make up an example, a division-sized force of, let’s say 10,000 men will always suffer (let’s say) 2% casualties when confronted with enemy of force of the same size and composition, and will always suffer (let’s say) percentage casualties four times higher when an enemy force of the same composition but twice as large. (These aren’t real numbers; I’m just making the point that in such pure systems, Force X doing battle under Condition Y will always end up with Outcome Z.)

If a system-based method isn’t deterministic, then it’s random. I’m not sure there’s anything located between “deterministic” and “random”; I think it’s more correct to view the choice as a two-step one: deciding whether the system should be deterministic or random, and then (if the choice was to go with “random”) deciding *how random* one wants it to be. It can be random within a very narrow and simple range of possible outcomes (“heads or tails” being as narrow and simple as you can get), or random within a broader and/or more complex range of possibilities. One die versus two dice illustrates the difference between range and complexity. One die (6 outcomes, each with a 16.7% chance) has more range than a coin toss (2 outcomes, each with a 50% chance), but not much more complexity because all the outcomes are equally probable in both cases. Two dice, however, not only increase the range but also the complexity, because the chances for each possible result (ranging from 2 to 12) are distributed along a bell curve rather than a straight line.