Branching Skill Tree

Points arrive with levels and are spent on nodes arranged in dependency chains. Since points are finite, every allocation forecloses another, and the tree's shape communicates that cost visually before the player commits. Good trees make trade-offs legible and let a specialisation feel like a plan; weak ones present an illusion of choice, where most paths converge or a single branch is obviously superior to the rest.

Blizzard North's Diablo in 1996 gave each class a page of skills tied to levels, in a game built around repeated runs whose reason to start again is to build differently. Visual layout mattered as much as the mathematics, since a chart shows a possible future in a way a list of numbers cannot. Its sequel expanded this into the large synergy trees that most later implementations imitate.

Trees spread far beyond role-playing games because they attach a visible plan to any progression, which keeps players oriented across long sessions. Same ubiquity is the complaint: many trees are filler, offering percentage increases that change nothing about play, and a respec button quietly removes the opportunity cost that justified the structure. Designers now argue for smaller trees with fewer, sharper decisions instead.