Micro Composition

Micro composition is the structure, timing, sequencing, and arrangement of individual elements (or groups of elements) from moment to moment. Micro composition uses the concepts of tempo, rhythm, melody, and harmony (although this can be applied to macro composition as well). This terminology is borrowed from music theory because it works well to describe the structure and flow of elements that occur over time.

Tempo
The tempo from moment to moment (micro-tempo) is measured by the frequency with which events occur or the frequency with which actions must or should be performed. The chart to the right shows a simple example of an experience that just repeats a single event or action at a constant interval--a monotone rhythm that isn't very interesting on its own. Well-constructed changes in tempo will usually feel good, and just changing the tempo is an excellent way to vary the tension of the experience, but that alone is not enough to make it engaging. Ultimately, these events and actions will need to have interesting patterns as well as the correct tempo.

Note that in a game, the tempo can vary just based on play-style (a casual RTS player will have a much slower tempo than a professional RTS player), although this is not true for a game that directly dictates the tempo. The concept of tempo can also be applied at the macro scale (the frequency of segments and episodes in a game) and the meta scale (the frequency of new titles in a series), but is most useful at the micro scale.

Rhythm
A particular repeated pattern of actions and events defines the rhythm (i.e., "the beat") of the experience at the micro level. While the rhythm of a non-interactive experience is usually highly directed and constant, interactive experiences (especially games) are rarely highly directed (especially at the micro level) and do not force an exact rhythm. Instead a game's rhythm is usually judged by what repeated patterns of actions are most effective (or easy to learn), because that is how most players will play the game. This means the rhythm in a game is rarely as constant in practice as it is in theory, but even a rough pattern can result in gameplay that flows much better than just a repeated monotone pattern (or something random with no rhythm at all). The chart to the right shows an example for a game that has a basic rhythm of three quick attacks, followed by two slow ones, that might be found in a brawler or shooter (based on the cool-downs for the attacks or types of enemies encountered).

While a good rhythm is essential to an engaging experience, a rhythm alone is not enough. The nature of the actions and events themselves need to be varied in an interesting way as well. The concept of rhythm can be applied at the macro scale (a repeated pattern of segments or episodes), but rarely is relevant at the meta scale.

Melody
When actions and events are varied in a repeated loop, this is like varying the pitch of the notes in a musical melody. When combined with a good rhythm, you have a melody, which is the core of a compelling moment-to-moment experience. In an interactive experience such as a game, when talking about the "core dynamic" you are generally referring to the melody of the gameplay, which is what makes a particular dynamic feel unique. For example, a character in a combat game might have the jab, block, and uppercut actions. If the jab generally provokes an opponent to counter with a cross, that can be stopped with the block, which will expose the opponent to a devastating uppercut, then you have a simple little melody. This assumes it has a good rhythm (timing/pacing) as well. The chart to the right shows a melody of jab, block, uppercut, move, move, with each action being a different color (the height represents the tension created by that action). This pattern would be repeated as the player moved from opponent to opponent (assuming perfect execution) in a simple brawler. The pattern would never be this regular in a real game, of course (there would be critical hits, combos, different enemies, etc.), but this could be the baseline pattern that additional actions and events would enhance.

Note that actions can vary, even if the input for the action does not. For example, the inputs in Tetris never vary (left, right, rotate), but each different type of piece varies the actions you are taking and creates a melody. Mechanics, components, and space can all be used to create a good melody in an interactive experience; don't just focus on the mechanics.

In a non-realtime game (especially a non-digital game), a real melody only tends to emerge with experienced players who are very focused on the game. The interruptions caused by confusion about rules, uncertainty about what to do next (too many choices in a game, for example), or other distractions destroy the pacing of the game and tends to disrupt any melody that might be there.

Harmony
When multiple melodies occur at the same time, this is called harmony. Making two or more melodies work well together is difficult, but can be very engaging when it does work. Examples of multiple melodies would be an RTS with resource collecting, building, exploration, and combat all occurring at the same time, or a multiplayer game with different roles such as tank, DPS, and healer.

Micro Tension
The individual actions and events at the micro scale create very rapid spikes of tension, since each action or event has its own stakes and anticipation. (How will the protagonist respond to that insult? Will the block be timed right against that attack?) If these quick spikes of micro tension are layered on top of building tension at the macro scale (Will the protagonist escape from her captors? Will you be able to defeat this entire wave of enemies?), the tension curve created would be exactly the form needed to drive engagement effectively.

The chart to the right shows how this might look. Compare the peaks and valleys of the tension on this chart to the those on the tension/engagement curve chart and notice how the pattern is essentially identical. The difference is that this curve is happening moment to moment (time scale of seconds) while the other curve is happening segment to segment or episode to episode, based on the macro composition of the experience.