Consequential Branching Endgame

Decisions are recorded as flags and totals during play, then read at a branch point late in the story. What changes is structural: which levels exist, who fights beside you, which characters are alive, what the last confrontation is even about. That costs more than a slideshow epilogue, because it means building content most players will never see, and it forces writers to keep every thread resolvable from several incompatible states.

BioWare shipped Star Wars: Knights of the Old Republic in 2003 with two substantially different final acts, driven by alignment and companion standing built up over dozens of hours. Doing so made the morality score mechanically honest: a number that had mostly bought powers now determined the shape of the ending. Working inside a licence helped, since it supplied two destinations audiences already understood, which kept the branching to a manageable width.

Once endings could genuinely diverge, players began treating earlier choices as investments, which raised the stakes of ordinary conversations and created a lasting expectation that games remember. Critics note how often the divergence is narrower than advertised, with branches reconverging on the same boss fight under different lighting. Studios have answered by branching smaller and more often, or by paying for genuinely different final hours and accepting that most players see one.