Games towards the right ( Shogun 2) are smaller and more focused. Meanwhile, the chart’s X-axis measures scope. Games towards the top ( Attila, Warhammer I, Shogun 2) have a more defined structure, typically ushering the player towards a do-or-die endgame. Structure is measured along the Y-axis of the chart. (How well they succeeded is a different question.) I’ve created the following diagram to illustrate this: Classifying the Total War games by scope and structure. This is why different players prefer different entries in the series - the designers were trying to accomplish different things. There is no one Total War design there are several, differing by structure and scope. They are united by circumstance - and the design of their respective games.
A universe away, a different emperor raises his magic hammer, and beckons his griffon into the skies. In Western Europe, circa 400 AD, a Roman emperor inspects his comitatenses and scholae, the successors to Caesar’s legions. Two different emperors prepare to defend their worlds.