Froke’s debut asks you to manufacture rubber ducks and throw them into a corporate hole, and then quietly turns into a physics engineering toy worth staying for. The tinkering is better than the premise deserves, and thinner than the atmosphere keeps promising.
The elevator, and my problem with number-go-up
For years I have kept incremental games at a polite distance. My objection was never that they are lazy. It is that the best of them are spreadsheets in costume, and the costume tends to come off somewhere around the 40-minute mark, leaving a graph you have agreed to keep watching out of politeness. Numbers rise. Your hands do nothing. Some people find that soothing. I find it a bit like being read a bank statement.
So a game whose entire pitch is “make rubber ducks, throw them in a hole, watch the money climb” was not going to get much patience from me.
Then it handed me a golf club.
That is the moment Project P.I.T.T. stops being a joke you scroll past on Steam and becomes something with a grip on you. Because the hole is not an abstraction and the ducks are not tokens. They are objects, in a physics simulation, and everything between the crafting bench and the pit is a problem you solve with fans, magnets, panels, and whatever else the terminal is willing to sell you. It is the debut commercial release from Froke, a French solo developer who came to games through audio fiction, photography, and 3D animation, and it arrived on August 19 with more than 115,000 wishlists behind it and a publisher, the Polish outfit Pretty Soon, that clearly knew what it had.
I went all the way through it, both endings included. It is not the game the store page promises. It is better in one direction and thinner in another, and both of those are worth explaining.
What the job actually is
You start in an elevator. You leave the elevator. You are an intern, the facility is enormous and made of concrete, and in the middle of it sits the Maw, a pit that eats what you throw in and pays you for the privilege.
Ducks first. You hand-craft them at a machine, you carry them to the edge, you throw them in. Money appears. That money goes into a terminal that sells you upgrades and parts, and the parts are what turn a job into an engineering problem. Concentric yellow rings around the pit mark your progress; when one fills, the facility advances a phase, the room physically changes, and something new becomes available to manufacture. Ducks give way to cash registers, cash registers to piñatas, and later phases introduce assembly, where products stop being single objects and start being components of stranger ones.
Feeding the Maw in bursts triggers combos, and combos multiply the payout, which is the hinge the whole economy swings on. A steady trickle of ducks earns you a living. A synchronized barrage earns you a factory. Everything you build is, at bottom, an attempt to make more things arrive at the same place at the same time.
That is the loop. It sounds like an idle game described by someone who has never played one, and for the first half hour it plays like one too.
Fans, magnets, and a golf club
The tools are where this thing earns its keep. Fans push. Magnets pull. Panels redirect. Spring pads bounce, pistons shove, launchers fling, and the golf club exists so you can drive a duck across the room into the pit like Bubba Watson at the 18th. There are brooms. There are basketball hoops for putting a cash register through. There are little robot vacuums that scurry around the floor and, in my experience, make trouble.
None of it snaps to a grid. That is the design’s central bet, and it is a bold one. In a normal factory game you place a conveyor and the conveyor works. Here you place a fan at an angle, and then you place another fan to catch what the first one overshot, and then you wedge a panel in to stop the overshoot from clearing the wall entirely, and eventually you have a machine that functions for reasons you only partly understand. Watching that machine run is the best feeling the game has. There is nothing quite like standing back and seeing a hand-built Rube Goldberg mess deliver 12 ducks a second into a hole that wants them.
I will be honest about the cost of that bet, because it is real and it is felt daily. The physics are unpredictable in ways that cross from charming into irritating. Fans decide to launch objects in directions they had not launched them for the previous half hour. Panels clash when they touch and get stuck in each other. Sticky Tools lock a placed tool once its panel meets the floor or another stuck object, which is the fix for drift, and it is also the thing you forget to apply until a robot vacuum has nudged your carefully aimed fan four degrees to the left and ruined a lane. The panel rack removes panels automatically when you interact with it, and there is no way to opt out, which breaks the flow of building more often than any physics quirk ever did.
Every one of those is a small thing. Together they are the difference between a great toy and a very good one.
Phase 3, and the thing that got taken out
Phase 3 deserves its own heading, because for the first day of this game’s life it was the worst decision in it.
The facility changes as the Maw grows, and the change is not cosmetic. Debris falls. Large scripted chunks of the building come down, and where they come down is where your production line happens to be. At launch, the Phase 3 transition did that thoroughly enough to bury hours of work under rubble, with panels trapped underneath and no way to recover them. There was no warning. You crossed a threshold you could not see coming and your factory was gone.
It went down badly enough that it was out of the game within a day. Froke pulled the event on August 20 and said plainly that the workplace is still unsafe but your factory should no longer be punished for being a factory. He was right on both counts, and the speed of it tells you something useful about who is maintaining this thing.
What remains is the correct version of the idea. Later phases still drop debris, still reshape the room, still expect you to rebuild rather than preserve. That is fine. Attachment to a single layout is the wrong instinct here, and the game is better for insisting on it. Stronger fans can even shove some of the fallen debris into the pit, which is the sort of detail that makes me trust a designer. There is a difference between a game that asks you to adapt and a game that deletes your homework, and Project P.I.T.T. now sits on the right side of it. Anyone who bounced off it in launch week on this specific point should know the reason is gone.
Who is watching, and what the ducks are for
The company is watching. Cameras track you from the pillars, and the game notices if you start climbing where you should not. Break the cameras, interact with the lights at the ends of the support pillars, and the facility starts giving up things it was not advertising: catwalks above the production floor, a fuse box that wants three fuses you have to find, a weighing platform that will not open until you get 200kg onto it, a device that sits stubbornly at 0% no matter what normal product you feed it. There is a figure that occasionally appears behind the observation window. There are two endings, and the second one is a genuine hunt.
This is the best-kept part of the game and the part the store page undersells. The secret route is real design, not an achievement list, and the 200kg puzzle in particular made me look at my own factory as a set of masses rather than as a throughput machine.
It is also where I have to be straight about what this game is not. The atmosphere writes checks the writing never quite cashes. Brutalist architecture, safety posters, harsh sodium lighting, a total absence of other people, an employer you never identify: all of that promises a slow-building dread that the game never really delivers. The horror stays as texture. If you go in expecting the facility to turn on you, expecting a revelation about the Maw that reframes the last few hours, you will finish the second ending feeling like the mystery was set dressing with a lock on it rather than a story. The questions the environment raises are better than the answers it eventually gives.
I do not think that sinks it. I do think it explains why some people finish and feel slightly short-changed. The ending is a door, not a payoff.
Concrete, cameras, and the PSX filter
Low-poly geometry, chunky dithered textures, and the deliberate vertex wobble of a 1997 console. It is a well-executed version of a style that has become extremely crowded, and what saves it from being another PSX pastiche is the architecture. The facility is genuinely oppressive: too much concrete, too few windows, lighting that is functional rather than kind. It looks like a place built by a company that filled in a compliance form about it. That is the joke, and the art direction tells it better than any line of text in the game does.
Requirements are modest. Windows 10 minimum with an i3-6100 or FX-6300, 4GB of RAM, and a GTX 750 Ti or RX 460; Windows 11 recommended with an i5-8400 or Ryzen 5 2600, 8GB of RAM, and a GTX 1060 or RX 580. Install size is around 1GB. There are native macOS and Linux builds, with Apple silicon covered from the M1 up.
The thing worth knowing is that the visuals are never what costs you frames. The simulation is. Every duck, register, and piñata on the floor is a rigid body, and a late-game facility running a fully automated line has a great many of them in motion at once. Performance in this game is a function of how much junk you let pile up, and a tidy factory that keeps objects moving through the pit rather than accumulating in corners runs noticeably better than a messy one. That is an unusual optimization lever to hand a player and it fits the game perfectly. It is Steam Deck Verified, which I would trust for the first two-thirds and treat with mild caution once your throughput gets serious.
Quacking into the void
There is no music. Not a sparse score, not ambient drones: none.
The case for that choice is easy to make. The facility is meant to feel abandoned by everyone except the entity paying you, room tone and machine noise are the honest soundtrack of a job like this, and the mass quacking of several hundred rubber ducks tumbling into a pit is genuinely funny the first dozen times. The manufacturing machines have satisfying mechanical chatter. The Maw itself makes a sound you stop noticing and then, later, notice again.
The case against it is that this is the single biggest missed opportunity in the game, and it is directly connected to the horror that never lands. The facility wants to be unnerving. Silence alone does not do that. A composer with a synthesizer and a sense of restraint would have carried the dread the environment art keeps setting up and the writing keeps declining to pay off. As it stands, most people will do what I did after the third hour and put their own music on, which is not something you should be able to say about a game trading this heavily on atmosphere.
Verdict
Project P.I.T.T. is a very good physics toy wearing the costume of a bad joke, and the toy is what you should buy it for. The tools are expressive, the combo economy gives your contraptions a reason to be efficient rather than merely funny, the facility looks and feels like a place that hates you, and the hidden route through the catwalks and the fuse box is smarter than anything the marketing hints at. For a first commercial game from one person, that is a real achievement, and the speed with which the Phase 3 disaster was recognized and removed says the support will hold.
What keeps it out of the 8s is that it never becomes the thing its own atmosphere keeps promising. The dread stays decorative, the ending explains less than the environment implied, there is no score to carry any of it, and the physics that make the good moments great also produce a steady drip of stuck panels and rogue fans that no amount of charm fully excuses.
Buy it if you like building machines with your hands and can enjoy one that works for reasons you cannot entirely account for. Buy it if the corporate-void gag makes you laugh, because the game commits to it completely and never once winks. Skip it if you want the horror the screenshots imply, or if you need an automation game with clean placement and predictable outputs, because this one hands you a golf club and a bottomless hole and considers the briefing complete.
