The Problem with Initiative

Comparing game mechanics with reality is the bane of role playing and as a general rule is best avoided. In some cases though those comparisons can yield important insights into how rules could be restructured. This is the case with initiative.

Games being played and moderated by humans obviously need to be turn based on some level. Traditional initiative systems, like that used by d20 games, split up time into discreet units. Each character gets a chance to act usually once each round, and the order in which they act is determined by their speed, reflexes, etc. compiled into an initiative modifier. There is a limit to how much can be done each round very roughly based on reality.

It's easy to come up with situations in which this type of system breaks down. The main problem boils down to the fact that actions take place atomically- either all or nothing. You move 30 ft or not at all, you cast the spell or do nothing, etc.

The sensible alternative is to have actions declared as if atomic (e.g. "I will move to this spot (30 ft away)"), but have time proceed incrementally, second by second or resolved-action by resolved-action. In this way actions suddenly take their true amount of time rather than being crunched into artificial boxes of time. If my character wants to move to a spot 30 ft away, it would never again mean using up 3/4 of their movement and wasting the rest of their turn.

The other advantage is that characters' actions can be changed mid-stream to realistically react to other happenings. The character who has started moving forward and is shot by an arrow from an unseen marksman can duck behind cover rather than finish out the rest of the move. Making such changes could depend upon a character's reflexes (and possibly current level of awareness), more realistically making use of various attributes rather than lumping them all into one initiative score.