Sega Genesis / Mega Drive · Any% · 1991
A speedrun of the game that was already about speed — and one whose record barely moves, because thirty years of optimisation have left the route almost nothing to give.
Sonic the Hedgehog is unusual among speedrun targets in that going fast is the game's explicit subject matter. That does not make it easy to run. The 1991 original is a game of momentum and physics, and the fastest routes require the runner to understand exactly how Sonic accelerates, how he interacts with slopes, and precisely where the level geometry can be exploited to skip large sections entirely. The run leans on a handful of well-known but demanding techniques. Zip glitches — which exploit the game's collision handling to launch Sonic through solid terrain at extraordinary speed — are the most spectacular, most famously in Marble Zone and Scrap Brain Zone, where a correctly executed zip can eliminate huge portions of a stage. Beyond the glitches, the run is a test of pure movement optimisation: preserving speed through every jump and slope, avoiding the enemies and hazards that would cost momentum, and threading a path that the level designers never imagined. What makes the category distinctive is how thoroughly it has been solved. Runners spent three decades chipping at the route, and the sub-nine-minute barrier — long treated as the definitive milestone — was finally broken in June 2022, when Eandis completed the game in 8 minutes and 55 seconds. It was the first sub-nine-minute run of Sonic the Hedgehog on the Mega Drive, arriving more than thirty years after the game's release. The community's own assessment is that there is very little left. Sonic runners have generally concluded that the record is close to the theoretical limit of what the route allows, and improvements now come in fractions of a second rather than the leaps that characterised earlier years. It makes for an instructive contrast with games whose routes are still being torn open by new glitch discoveries: Sonic 1 is a category that has essentially reached its floor, where the remaining question is not what to do but how perfectly it can be done.