It was in the lesson about American law where I learned about the use of leading questions "breaking down" the action and expanding the time. The narrative trick of describing each discrete element slows time. I realize that it's the same in combat - rolling to hit, to defend, damage, etc...
We came from D&D and have broken combat into these discrete elements, but Wargames and Boardgames have done these abstractions for a long time already. So if I want speedier combat, then I will have to go for more abstraction.
Its all-time and place, what is appropriate given the game and the story. If the story doesn't stress combat then we can abstract the combat more. We look at the 3-4 hours and ask what do we want to be doing? This is where imagining the Story as a Flow Chart - and that Combat is not a binary outcome - but one of many processes that open many possibilities. In Activity On Node, or Precedence Diagraming, or in Flow Charting its Diverging Paths and Converging Paths.
We can imagine the games as a flow chart with a line that begins to enter a Process (Square), that later diverges to decisions/options (diamonds) that diverge into more other Processes. Visually we want to have very unique branching possibilities. I want to imagine the "Inertia" of the Situation as the Main Trunk of Possibilities and the PC's are experiencing the Extremes - if you're visualizing from left to right, imagine a complicated branching trunk that follows a particular path of highest probability - now the PCs are diverging from that path. Either imagining in 3d as a Tree trunk and its many branching possibilities of adventure - or in 2d where it branches out into a thick web of possibilities.
The visual cues that make the outliers special - are Control Limits/Color Bands of Probability and Information. The more Information the PCs have their Branch of Probabilities are farther out the main trunk. At the same time - the farther from the trunk the more they are going into unknown territory. More risks and dangers happen far outside the trunk.
After imagining the game as these probabilities - I can't imagine rolling that many dice or having the bandwidth to think of stories in dice to dice. I'd rather it be shaped by decisions and behaviors. In 3-4 hours - what kind of story would the players have had? Given the Expectation, Conflict, and Resolution curve of emotion and storytelling - what branching decisions got Players into the story?
I guess for me I'm imagining 40% of time establishing the baseline knowledge and decisions, world-building and the setting. 40% resolving the many conflicts in the setting with the PCs and their own goals and expectations. then the 20% of guessing and reshaping the world with unlikely rolls, setbacks, and change of strategy and path.
That means of that the second 40% and the last 20% Players will work with the GM to figure out what kind of world this is. What kind of timeline and history we are making, with our decisions?