Galactic Arms Race

My first honest reaction to Galactic Arms Race was fascination, not excitement about the shooting. The premise here is not just a design choice but a genuine research artifact: the cgNEAT algorithm, born out of the Evolutionary Complexity Research Group at the University of Central Florida, watches which weapons you actually fire and evolve new ones from those preferences. Use a wide-spread particle cannon obsessively and the next drop will nudge toward wider spreads, brighter arcs, faster bursts. Reject a weapon by selling it or ignoring it, and the algorithm factors that out too. No two players grow the same arsenal. For a budget indie from this era, that is not a gimmick, it is a quietly remarkable piece of engineering. The moment-to-moment play sits in familiar top-down space-shooter territory. You pilot a gunship through the Perimeter, a loose galaxy of star systems connected by jump gates, fighting pirates, the alien Xindo, space blobs, and geonoids across escalating difficulty zones. You earn credits, dock at stations to repair and buy mods, level up to a cap of 200 and spend Mod Points across a decent range of ship stats, and slot up to three weapons at a time. Missions mostly amount to clearing enemies or delivering cargo, which is not a knock against the game so much as an honest description of what holds the loop together. The loop itself is the familiar action-RPG rhythm: fly, shoot, collect, upgrade, repeat. The shooting feels clean and the particle effects on weapons are genuinely pretty, which matters more than it sounds when weapons are the entire point. Where the game strains is depth. Outside the weapon evolution system, the quest structure is thin, generic enough that critics have accurately described it as a framework for demonstrating the tech rather than a complete narrative campaign. The solo campaign has a start and a destination, but the journey between is mostly open-ended sandbox flying without much story momentum pulling you forward. Players who need authored purpose to sustain engagement will tap out before the weapon system has time to do its most interesting work, which is a real shame. The system needs hours to show its full character, and the surrounding content does not always give you enough reason to log those hours. Multiplayer supports up to 32 players in both co-op and PVP, and the server browser lets anyone run their own galaxy with a standalone download. The honest caveat in 2025 is that the multiplayer population is effectively zero; finding a live session without coordinating friends in advance is unlikely. Co-op would genuinely improve the experience, turning the sandbox grind into something more social and stretching the weapon evolution across group playstyle data. On your own it stays playable but quieter than intended. Achievement hunters should also note that several achievements are gated behind multiplayer milestones that are now nearly impossible to reach organically. For a certain kind of player, none of that will matter much. If you find procedural content generation genuinely interesting as a design problem, this is one of the earliest commercial games to implement it implicitly from player behavior rather than explicit player input, and that distinction matters. The weapon properties vary across color, firing pattern, spread, speed, and direction, and watching your preferred combat style slowly reflected back at you through new drops has a quiet, almost hypnotic quality that the bigger-budget genre peers never attempt. Go in knowing the solo content is sparse, the multiplayer is dormant, and the campaign is a scaffold rather than a story, and you can find something small but genuinely unusual here. Kai, Scout Team

Runs on Linux but with crashes or issues. Based on 5 ProtonDB community reports.

Be the first to comment on Galactic Arms Race.