My Friend Pedro

Most action games want you to feel powerful. My Friend Pedro wants you to feel like you’re showing off, and that’s a different design goal. It’s developed by DeadToast Entertainment, essentially the studio of Swedish developer Victor Ågren, and published by Devolver Digital, released June 20, 2019 for $19.99. The game grew out of a free browser game Ågren put on Newgrounds years earlier. That origin shows in both directions. The concept is tight and easy to grasp in seconds, and the full release sometimes stretches it further than it can go.

The premise is pure Devolver. You wake up in a butcher’s basement with no memory and a talking banana named Pedro telling you to kill everyone between you and the exit. That’s about as much story as you get. The real point is the systems: slow motion, split aiming that lets you shoot two enemies on opposite sides of a room at the same time, acrobatic dodges, wall-runs, and an environment full of physics objects you can kick, ricochet off, or ride.

The game goes after the same “John Woo as a playable system” fantasy that Max Payne made famous and Hotline Miami reworked as a top-down puzzle. My Friend Pedro moves it to a side-scrolling 2.5D plane, and that turns out to be the smart choice. On a flat plane, your attention can go to the shot itself rather than to camera management.

The reception has been strong, with a Metacritic score of 81. That’s a solid showing for a game that costs twenty dollars and lasts an afternoon. The scores make sense. The game has real flaws, and it still does one thing better than nearly anything else on the market.

• Split Aiming: With a button held, you can lock one arm onto a target and aim the other arm freely. Dual-wielding stops being cosmetic and becomes a real tactical tool, so you can clear two threats in a single jump.
• Focus (Slow Motion): A meter-limited bullet-time ability that stretches the frames you have to make decisions. It’s the scaffolding that holds up the rest of the game.
• Spin Dodge: A mid-air pirouette that makes you nearly untouchable by incoming fire. It looks ridiculous and it’s also essential.
• Physics-Driven Environments: You can kick frying pans into the air and bank bullets off them, ride skateboards and barrels, cut zip-lines, and drop props on enemies. The levels are built for improvisation.
• Scoring and Global Leaderboards: A combo multiplier rewards variety and pace, and every level ends with a ranking. This is where the real game lives once the campaign is over.
• Escalating Set Pieces: Motorcycle chases, falling sequences, and boss encounters break up the standard shooting stages.

The Gameplay Loop

My Friend Pedro is built on a simple loop: enter a room, read the space, execute, then go back and do it better. That last step matters most. A first pass through any stage is usually sloppy. You burn focus too early, take chip damage, and land kills with no style. The game doesn’t punish this much. Checkpoints are generous and health regenerates. But the scoring system keeps telling you that you left points on the table.

Everything depends on split aiming. Mechanically, it separates your two arms into independent tools: one stays on the gunman to your left while you sweep the other across the shotgunner on the catwalk to your right. In practice, it changes how you see a room. You stop thinking in targets and start thinking in pairs. The first time you leap through a window, flip upside-down, and drop two enemies on opposite walls at once, it feels like you’ve learned something rather than just pressed a button. Few action games manage that.

Focus slows the clock so you can actually pull these moves off, and its tuning is the most important design call in the game. The meter drains fast and refills just quickly enough that you rely on it all the time but can never fully lean on it. That constraint keeps encounters tense. Without it, slow motion would become the default speed, and the game would turn into a slideshow.

Creative Combat, and Its Limits

The physics toys are where the game is most charming and also where it runs out of ideas. Kicking a frying pan into the air and bouncing a bullet off it to hit an enemy behind cover is a great trick. So is sending a skateboard across a room and riding it while shooting. These moments deliver exactly the fantasy the game’s marketing describes.

The trouble is that the game introduces most of its best tricks early and then mostly recombines them. Later stages lean more and more on enemy volume and environmental hazards instead of new mechanics. Some detours, especially the surreal late-game sections, swap shooting for platforming and light puzzle work that undercut the pace. They feel more like filler than variety. When the game breaks its own rhythm, you notice how much of its appeal depends on that rhythm.

The weapons have the same problem. The arsenal covers the standard set: dual pistols, dual SMGs, a shotgun, an assault rifle, and a sniper rifle. Each one handles distinctly, but there’s no progression, no modifiers, and no reason to prefer one gun over another beyond ammo and range. That’s defensible in a score-attack game, where consistency helps mastery. It still leaves the middle third of the campaign feeling flat.

Difficulty and Onboarding

Onboarding is efficient. Pedro teaches each mechanic in short tutorial rooms, and the early levels add new abilities at a steady pace without stalling on explanation. The default difficulty is forgiving, maybe too forgiving for players who just want to see the ending, since brute force can carry you through most encounters. Harder difficulty settings unlock later, and that’s where enemy damage forces you to actually use your full toolkit.

The learning curve for split aiming is real, and some players will bounce off it. Coordinating two independent aim points while moving, jumping, and managing a meter is a lot of work for your hands. The payoff is worth it, but the game could do more to ease players into it.

Narrative

The story is the weakest part. Pedro the banana is a funny idea, and the game knows it, but the writing never gets past the joke. A late-game turn toward surrealism and self-referential commentary tries to say something about why you’re doing all this killing. It comes across as more confused than profound. You won’t play My Friend Pedro for the plot, and the game would have been better off trimming it.

Replayability

This is the strongest reason to buy it. The campaign is short, but the leaderboards and per-level rankings turn each stage into an optimization puzzle. Chaining kills, varying your methods, and finishing fast all feed the multiplier. If that kind of chase hooks you, the game’s real lifespan is many times longer than its credits suggest.

DeadToast’s art direction is clean and functional rather than striking. Environments include warehouses, sewers, rooftops, and nightclubs, all rendered in a crisp, slightly flat 3D style on a 2D plane. They’re readable, which matters more here than beauty. You can always tell where enemies are, what’s interactive, and where bullets are flying. Pedro’s bright yellow and the bursts of red during combat do most of the visual work.

What stands out is the motion. Character animation, especially the flips, spins, and ragdoll physics on defeated enemies, sells the game’s choreography. The slow-motion effect is applied with restraint, so the game never looks muddy. The pounding electronic soundtrack fits the pace well, even if it rarely sticks with you afterward.

The UI is minimal: a health indicator, a focus meter, an ammo count, and a score multiplier. That’s all the screen needs.

As a 2D game of this scope, it’s light on hardware, and the design rewards a steady frame rate: split aiming is a precision tool, and it benefits from the smoothest presentation a system can give it.

Steam (PC): This is the definitive version. You can play with mouse and keyboard, and the mouse gives the most precise aim on your primary arm, but split aiming on a mouse requires a modifier key, which feels less natural than twin sticks. Most players will find a controller the better fit. Two analog sticks map almost perfectly onto two independent arms.

Nintendo Switch: The game suits the Switch’s short-session strengths. Levels are short, checkpoints are frequent, and in handheld mode it works well as a quick commute game. The trade-offs are technical: a 2.5D game with heavy slow-motion effects, ragdolls, and dense particle work asks more of the Switch than of a gaming PC, so handheld play is the compromise pick for players who care most about precision. Joy-Con thumbsticks handle split aiming less well than a Pro Controller, so serious leaderboard players should dock the system or use a better pad.

My Friend Pedro is a strong mechanic with a good game built around it. Split aiming, focus, and the spin dodge together make for some of the more satisfying shooting the genre has produced, and DeadToast tuned that core carefully. When the game sticks to it, with rooms full of enemies, physics props, and a multiplier to protect, it works.

It’s held back by everything around that core: a story that doesn’t earn its time, a second half that recycles its best ideas, and detours that misread what players came for. Those problems are real, and they matter. We recommend it to the right audience: score-attack players and action fans who will live in the leaderboards, especially at a sale price.