Arms of God Pro Tips: Complete Walkthrough & Boss Strategies

After about 50 runs and a frankly embarrassing number of deaths to easily dodgeable attacks, here are the things that actually improved my clears. Not the obvious tutorial stuff , the real optimization tricks that separate a 35-minute comfortable clear from a 55-minute death spiral.

Movement Is Your Real Weapon

Damage numbers are fun to watch but positioning is what keeps you alive. Enemy projectiles in this game have consistent travel arcs. They don't track you. They fire toward your current position and then travel in a straight line. If you're always moving, they always miss.

Easy to say, hard to do when there's 40 enemies on screen and your brain is processing a dozen visual inputs. The trick I've settled on: move in a wide figure-8 pattern. It creates a controlled zone in the center of the map where enemy density stays manageable. More importantly, it gives you consistent escape routes in every direction. Moving in circles near walls will get you cornered eventually.

On Steam Deck, the analog stick gives more precise diagonal movement than WASD on keyboard. I actually prefer playing on Deck for this reason, even though mouse aim is technically more accurate for manual aiming mode. The game is Steam Deck verified for good reason.

Weapon Merge Math Nobody Explains

The stat transfer from merging isn't 100 percent. First merge transfers about 60 percent of the donor's stats. Each subsequent merge into the same weapon reduces efficiency slightly.

What this means: don't merge your best donor into a weapon you might replace later. Wait until you've got a weapon you're keeping for the whole run. Those early merges are permanent and the donor is gone forever.

Merging weapons of the same element gives a hidden bonus , the target gets extra projectiles or increased AoE radius. The game doesn't tell you this anywhere in the tutorial or tooltips. I noticed it after accidentally merging two fire weapons and suddenly shooting three projectiles instead of two.

The sixth merge (unlocked in Act 3) resets the efficiency penalty. So your fifth merge in late Act 2 should use a mediocre donor, saving your best donor for the sixth merge where they'll get full value. This is optimization territory but it matters when you're pushing for consistent Act 4 clears.

Blessing Selection Strategy

The triple-choice blessing screen is an actual decision point. I used to pick the highest rarity without reading. Very dumb. A common blessing that adds projectile count is worth more than a legendary blessing that gives 20 percent fire damage when you have zero fire weapons.

The blessing pool is influenced by what weapons you're holding. If all five weapons are fire-element, you'll see more fire-related blessings. This means you can soft-target specific blessings by shaping your weapon loadout early.

S-tier blessings to take every time: projectile count increase, piercing, chain/bounce, movespeed. A-tier: AoE radius, status effect duration, attack speed. Everything else is situational or bad.

Per-Boss Positioning Reference

Act 1 final boss: stay below it. Horizontal sweep attacks miss you there. Projectile ring has its widest gap at the bottom.

Act 2 mid-boss: circle clockwise around the arena. The add spawns favor the left side and circling clockwise keeps them in front of you where your weapons are already firing.

Act 2 final boss: the safe zone during phase 2 bullet hell is slightly left of center. I don't know if this is a bug or intentional design but you can stand there and roughly 60 percent of projectiles miss entirely.

Act 3 mid-boss: apply every status effect you can. The resistance mechanic means each unique status is effectively a damage multiplier. Burn, poison, slow, chain, anything. Stack them all.

Act 3 final boss: during mirror phase, walk directly toward the boss. The reflected projectiles spread behind you. Feels wrong, works every time.

Act 4 final boss: never stand still. Not for half a second. Not even during what looks like a safe phase. Constant lateral movement. The boss can fire a targeted projectile at any moment that one-shots low HP builds.

What I'd Tell My Past Self

Buy movespeed in the Training Ground first. Max it. Every death in my first 20 runs was a positioning death, not a stat death. Being able to outrun enemies is worth more than any HP padding.

The Divine Form isn't just a panic button. Use it proactively when you see a wave that looks problematic. Popping it at 10 percent HP wastes the damage immunity time. Pop it at 80 percent HP when a dangerous wave spawns and you get the full benefit.

And for anyone struggling with the Act 2 difficulty spike: one AoE weapon changes everything. The game is balanced around you having some form of area damage. If all five weapons are single-target, you're fighting the game's design, not the enemies.I should mention one more thing that doesn't fit neatly into any of the sections above. The Early Access build as of June 2026 is remarkably polished for a solo developer project. Dominik Sójka clearly put the time in before launching. I've encountered maybe two minor bugs across 50-plus hours. The frame rate on Steam Deck is locked at 60 even during the most chaotic Act 4 wave sections with every weapon firing and particle effects filling the screen.

The weapon balance isn't perfect , some weapons are clearly better merge bases than others , but that's part of the fun in a roguelite. Figuring out what's strong and what's a trap is half the game. The other half is execution. And the metal soundtrack.

If you're on the fence about buying this in Early Access, I'd say go for it if you like bullet heavens. The content is substantial for the price. Four full Acts, 10 bosses, 10 characters, 60-plus weapons, 100-plus blessings, a meta-progression system that makes every run feel meaningful even when you lose. And the full release with console ports is coming Q4 2026 or Q1 2027. Your Early Access purchase supports a solo developer who clearly cares about their game. That's worth something in my book.