A shooter game maker where feel is the thing you tune

Every shooter template in every engine gives you a gun that fires. Almost none of them give you a gun that feels good, because feel is a stack of small decisions — mouse acceleration, weapon sway, the delay between click and shot, how long the enemy staggers, whether the reload can be cancelled — and the correct value for each depends on the game around it.

This is why "generate me an FPS" is not a satisfying answer to anything. You get something that technically shoots and then you have to fix all of it anyway.

What actually helps is making each of those numbers cheap to change and fast to test. That is the shape of the loop here: you say what is wrong, it changes, you play it again.

Shooter levels are architecture. Sightlines, cover spacing, the height difference that makes one approach better than another, the corner that stops being safe once the enemy learns to flank. None of that is describable in a sentence, and it should not have to be.

In the builder you place geometry, terrain, water and buildings directly and walk through what you made. That is the hand-built half of the app, and for this genre it is the half that decides whether the game is any good.

It runs on your machine. Stage Engine is a desktop app for Mac and Windows with Godot bundled inside, so pressing play gives you your own framerate. That matters for a shooter more than for most genres, because input latency is part of the feel you are trying to judge.

Weapons, ammunition, damage falloff, hit detection, enemy states, spawn logic, cover selection, scoring. Every one of those is code, and none of them are decisions you make by writing them — you make them by playing.

So you describe them. The shotgun should be lethal inside five metres and useless past ten. Enemies should stop pushing when two of them go down. The last round in a magazine should do the same damage as the first, because the alternative is annoying. Director 1 writes the GDScript in your project and you go and shoot things.

You never open a script unless you want to. Not writing the code does not mean not having an opinion about whether the rifle is too forgiving.

A weapon is a set of numbers. An enemy is a small program that has to be readable from twenty metres away while it is trying to kill you. Players do not experience enemy AI as intelligence; they experience it as legibility. Can I tell what it is about to do?

Practically that means the useful notes are behavioural rather than technical. It should telegraph the charge. It should not shoot the instant it sees me. It should give up if I break line of sight for four seconds. Those go straight into the loop.

Getting this right takes several passes. Budget for that, and start with three enemies that behave differently rather than twelve that behave the same.

After a few hours you are the best player of your own shooter, by a wide margin, and every fight will feel too easy. This happens to everyone and it is the main reason first-time difficulty tuning is wrong.

The cure is other people. Watch one person play, without helping them, and write down every place they stopped or died. That list is worth more than a week of your own opinions, because they are meeting the game the way a player will.

It also tends to reveal that the problem is legibility rather than difficulty. Players rarely say a fight was too hard; they say they did not see where the shot came from. Those are different fixes and only the second one is worth making.

Download and install on Mac or Windows. The project is a normal Godot project on your disk, because Godot is what is underneath and it is open source. Nothing is stored in a format that only one company can open.

Building needs an internet connection — Director 1, asset generation and world data come over the network. Playing what you built does not.

Free to download with a free plan, Indie $10 a month, Pro $50. The paid plans buy usage, not a better engine.