Methodology and limits

We include a mechanic when it is a discrete, describable thing a game does that other games later did too, and when we can name a title where it first shipped in recognizable form. Cosmetic differences, engine features with no effect on play, and one-off gimmicks with no descendants are excluded. That leaves 168 entries across fifteen families. Families are navigational conveniences — colour and grouping — not taxonomic claims; several mechanics could sit in two of them, and a few sit awkwardly in one.

Origin means first notable shipped implementation, not invention. We assign it to the earliest release we can document where the mechanic is legible as itself rather than as an accident, and we accept that this is partly a judgment call every time. Prototypes, unreleased builds, regional releases, and mainframe software that never shipped commercially all complicate it. Where a mechanic was popularized by a later game, we credit the earlier one and say so: Over-the-Shoulder Aim is dated to kill.switch in 2003, not to Resident Evil 4; Dual-Analog Move + Look to Alien Resurrection in 2000, not to Halo 2.

Parents must predate their children. No feature links to a parent released in the same year or later, and the graph is acyclic by construction. Three link types carry the argument. Derives connects a mechanic to the earlier mechanic it revises, and is the backbone of the feature lens — the claim is that the later idea is unintelligible without the earlier one. Introduces connects a game to the mechanic it originated, one per mechanic. Adopts connects a game to a mechanic it uses significantly without originating; these are illustrative and deliberately incomplete, since a complete adoption list for hit points would include most of the medium.

Known limitations, stated plainly. Coverage is biased toward Western PC and Japanese console releases, because those are the traditions best documented in English. Arcade-era Japan is thin, and several mechanics we credit to home computers likely appeared earlier in cabinets we cannot verify. PC strategy is underweighted relative to its influence, and mobile is represented mainly through business models rather than through interaction design. Twelve entries have no parent, which is more a statement about our sources than about spontaneous invention. Nothing after 2020 is included, since recency makes descent unreadable. Above all, first claims in games history are contested as a matter of routine; treat every origin here as a well-sourced position rather than a settled fact, and tell us when we are wrong.