Forge and Fight!

Forge and Fight! is a physics-based multiplayer arena brawler from Swedish indie studio Flamebait Games, built entirely around one loop: assemble a weapon from a pool of parts, bring it into an arena, and beat everyone else into the floor with it. The part library covers a wide spread of playstyles. Want to hack and slash with circular saw blades? Fine. Prefer to incinerate opponents with a flamethrower attachment, electrocute them with a tesla coil contraption, or poison them from range? Also fine. The forging system rewards experimentation: stacking parts of the same category together unlocks set bonuses that range from flat damage boosts to entirely new special abilities like Charge, Fire Nova, or Bear Trap. That bonus layer is the closest thing to a build theory here, and it gives players an actual reason to think before they click parts together rather than just gluing everything sharp onto a stick. The two main pillars are the online arena and a completely open sandbox mode. In the arena you face off against up to seven other players across modes including Turf War (zone control), Brawl (first team to three eliminations), and Last Team Standing. Each round surfaces random upgrade choices, adding a light draft element that keeps weapon evolution happening mid-match and stops any single loadout from dominating every game. The sandbox mode drops all restrictions entirely, lets you test absurd creations against bot waves, and also hosts mini-games including football and cannon target practice. It functions as a low-pressure lab before you take anything into a live lobby, which is genuinely useful given how non-obvious some part synergies are. Now for the honest accounting. Movement has been called out as sluggish for a brawler, and ranged weapons use a directional line rather than a traditional crosshair, which takes some adjustment. More critically, this is a multiplayer-first game that launched into Early Access with a small player base, and finding populated lobbies has been inconsistent since day one. The bot support in sandbox helps, but bots are not a substitute for the chaos of eight human players all wielding equally deranged contraptions. The cosmetic progression system earns experience and scraps from matches to unlock character customization, and there are no pay-to-win mechanics anywhere in that economy, which is worth noting. I know what you're thinking: strategy guy reviewing a party brawler? Hear me out. The weapon-building system is closer to a draft or build-order puzzle than it first appears. The set-bonus mechanic means that random part availability each round creates genuine resource allocation decisions. You're not just picking what looks dangerous, you're calculating whether your current parts cross the threshold to unlock a bonus, whether a pivot to a hybrid melee-ranged build is viable with what's on offer, and whether your opponent's visible weapon is telegraphing a specific counter-build. It's lightweight decision theory, not Paradox-grade, but it's there. The sandbox also has real value as a controlled testing environment, and for newcomers the lack of a strict meta means almost any experimental build is viable at casual skill levels. Flamebait Games previously shipped Passpartout: The Starving Artist and Verlet Swing, both small, creative indie titles with distinctive mechanics, so this studio knows how to build around a single interesting idea. Whether the player population has held up since launch is the real unknown, and that single variable matters more here than any design consideration. Diego, Scout Team