47 KiB
SpeedBleed — Game Design Document & Development Roadmap
Working title: SpeedBleed Engine: Godot 4 Genre: First-person combat racing / vehicular FPS Visual Style: PS1-era low-poly / low-resolution 3D Core Fantasy: Drive dangerously fast, manually operate your vehicle and weapons, and kill or outrun your opponents. Target Team: 1–3 developers Primary Platform: PC initially Multiplayer: Future expansion after the single-player game is proven
1. High-Level Game Vision
SpeedBleed is a high-stakes first-person combat racing game inspired by the aggressive arcade racing and vehicular combat games of the PS1 era.
The player is seated inside a manually operated low-poly vehicle. Instead of controlling an abstract car from behind it, the player physically interacts with the cockpit:
- Grab and turn the steering wheel.
- Shift gears manually.
- Operate the handbrake.
- Reach for weapons.
- Manually reload firearms.
- Lean, duck, and protect themselves from incoming fire.
- Manage vehicle damage.
- Choose when to attack and when to concentrate on racing.
- Tune the vehicle around specific driving and combat strategies.
The game should feel dangerous, physical, fast and slightly ridiculous.
The player should frequently have moments such as:
"I'm taking a hairpin at 100 mph, my engine is screaming, I'm steering with one hand, I'm trying to reload with the other, and the guy behind me is shooting through my rear window."
That is the experience the entire design should serve.
2. Core Design Pillars
Everything added to SpeedBleed should support at least one of these pillars.
2.1 Drive Like a Maniac
The driving model should reward learning the vehicle rather than simply holding a racing line.
Important characteristics:
- Weight transfer.
- Tire grip and slip.
- Oversteer and understeer.
- Drifting.
- Braking technique.
- Manual gears.
- Engine braking.
- Handbrake turns.
- Collision recovery.
- Surface changes.
- Damage affecting vehicle behaviour.
The goal is not to create a hardcore simulator.
The goal is an arcade-sim hybrid that is believable enough to feel physical while remaining entertaining.
2.2 Your Hands Are Part of the Gameplay
The cockpit should not merely be decoration.
The player's hands are an important gameplay system.
The player should physically:
- Grab the wheel.
- Shift gears.
- Pull the handbrake.
- Grab weapons.
- Aim weapons.
- Reload.
- Operate weapon mechanisms.
- Interact with the dashboard.
- Potentially activate switches, lights, horn, radio, etc.
This creates the central SpeedBleed problem:
You only have two hands.
If your right hand is holding a revolver, it cannot simultaneously shift gears.
If both hands are on the steering wheel, you cannot reload.
If you are ducking under the dashboard, you cannot see properly.
These limitations should create tension rather than frustration.
2.3 Racing and Combat Are Equally Valid
The game should never become:
"A racing game where you occasionally shoot people."
Nor should it become:
"A shooter where everyone happens to be sitting in cars."
Both systems should matter.
A skilled player should be able to win by:
- Racing aggressively.
- Eliminating opponents.
- Disabling opponents and escaping.
- Combining racing and combat.
- Building a vehicle specifically for one strategy.
2.4 Every Upgrade Has a Cost
Avoid traditional:
Engine Level 1 → Engine Level 2 → Engine Level 3.
Instead, upgrades should create meaningful trade-offs.
Examples:
- More power = more wheelspin.
- More armour = more weight.
- Better suspension = different handling.
- Bigger engine = worse fuel/temperature management.
- Tall gears = easier driving but weaker acceleration.
- Short gears = faster acceleration but more shifting.
- Large brakes = better stopping but potentially more weight.
- Reinforced windshield = protection but worse visibility.
The player should build a car, not simply increase a numerical rating.
2.5 PS1 Aesthetics Are a Feature, Not a Limitation
The visual style should deliberately embrace:
- Low-poly geometry.
- Low-resolution textures.
- Texture warping.
- Vertex wobble.
- Dithering.
- Strong colour palettes.
- Aggressive fog.
- Dark interiors.
- Chunky animation.
- Low-resolution effects.
- Visible polygonal damage.
- Simple but memorable silhouettes.
Do not attempt to hide the limitations.
Make them part of the identity.
3. Feasibility Assessment
SpeedBleed is particularly suitable for a small team because the first-person cockpit dramatically reduces the amount of vehicle exterior and character animation that must be produced.
The player spends most of the game inside one vehicle.
This means development effort can concentrate on:
- One excellent cockpit.
- A small number of exterior car models.
- Modular tracks.
- Weapons.
- Effects.
- AI.
- Systems.
A polished 5–8 car roster is preferable to attempting 30 mediocre cars.
Likewise, 3 excellent tracks are preferable to 15 unfinished tracks.
4. Recommended Team Structure
For a 2–3 developer team, use the following division.
Developer 1 — Gameplay / Vehicle / Technical Lead
Primary responsibility:
- Vehicle physics.
- Player controls.
- Cockpit interaction.
- Hand IK.
- Weapon interaction.
- Damage systems.
- AI vehicle integration.
- Game-state systems.
- Save/load.
- Performance.
- Multiplayer architecture later.
This developer should own the playable game loop.
Developer 2 — Art / Vehicles / Tracks
Primary responsibility:
- Car models.
- Cockpits.
- Track construction.
- Track props.
- Weapons.
- Environmental assets.
- Destruction models.
- Texture atlases.
- Materials.
- UI art.
- Garage art.
This developer should own the physical world the player drives through.
Developer 3 — AI / Audio / UI / Effects
If a third developer is available:
- Rival AI.
- Race management.
- Combat AI.
- UI.
- Menus.
- Audio.
- Music integration.
- VFX.
- Camera effects.
- Gameplay feedback.
- Statistics.
- Progression.
- Testing tools.
This developer should own the game's presentation and opponents.
If there are only two developers, these responsibilities should be split between Developers 1 and 2.
5. Core Gameplay Loop
The basic game loop should be:
Choose Car
↓
Choose Loadout
↓
Choose Race
↓
Start Race
↓
Drive + Fight
↓
Damage / Kill / Repair / Adapt
↓
Finish Race
↓
Earn Money / Reputation
↓
Buy Parts / Weapons / Cosmetics
↓
Tune Vehicle
↓
Race Again
Later:
Race
↓
Series / Championship
↓
Rival Drivers
↓
Unlock New Cars
↓
Unlock New Weapons
↓
Harder Tracks
↓
Boss / Championship Events
6. Victory Conditions
Each race should have two primary ways to win.
Racing Victory
Finish in the top three.
This is important because it prevents the game from requiring the player to kill everyone.
A skilled driver can simply survive.
Elimination Victory
Eliminate all rival drivers.
This creates a completely different strategy.
A player who is behind in the race may decide:
"Forget the podium. I'm killing everyone."
This also makes combat builds viable.
Optional Hybrid Objectives
Later race types can introduce additional objectives:
- Finish first.
- Eliminate a specific rival.
- Survive for a set number of laps.
- Protect an allied driver.
- Destroy a target vehicle.
- Finish with minimum vehicle damage.
- Finish without visiting the pit lane.
- Finish with limited ammunition.
- Kill a specific number of opponents.
- Finish while carrying a cargo objective.
7. Control Scheme
7.1 Recommended Method — Direct Mouse-Hand Coordination
This should be the default control model.
Keyboard
| Input | Function |
|---|---|
| W | Throttle |
| S | Brake / Reverse |
| A / D | Optional steering assistance / secondary controls |
| Space | Duck |
| Shift | Handbrake |
| R | Reload / weapon mechanism |
| 1–4 | Weapon selection / quick slots |
| Esc | Pause |
The exact keyboard mapping should remain configurable.
Mouse
The mouse controls the player's physical hand interaction.
Left Hand
The left hand primarily controls:
- Steering wheel.
- Emergency controls.
- Potentially handbrake depending on vehicle layout.
Right Hand
The right hand primarily controls:
- Gear stick.
- Weapon.
- Dashboard controls.
- Reloading.
The player can click and drag cockpit objects.
8. Cockpit Interaction System
The cockpit should be built around interaction zones.
Zone 1 — Steering
The wheel supports:
- One-hand steering.
- Two-hand steering.
- Mouse-driven rotation.
- Automatic return-to-centre.
- Physical steering animation.
Two hands on the wheel should provide improved control at high speeds.
One-handed steering should feel less stable.
This is primarily a gameplay consequence rather than a simulation requirement.
Zone 2 — Centre Console
Potential interactions:
- Gear stick.
- Handbrake.
- Ignition.
- Radio.
- Headlights.
- Hazard lights.
- Dashboard switches.
- Weapon selector.
- Vehicle systems.
Most of these can initially be cosmetic.
They can become gameplay systems later.
Zone 3 — Weapon Storage
Weapon locations can vary between vehicles.
Examples:
- Door holster.
- Passenger seat.
- Dashboard mount.
- Centre console.
- Floorboard.
- Rifle rack behind the seat.
This creates subtle differences between cars.
9. Ducking
Ducking is one of the game's signature mechanics.
Holding the duck button:
- Lowers the player's head.
- Moves the camera below the windshield.
- Reduces headshot vulnerability.
- Makes incoming fire sound louder.
- Restricts visibility.
- Prevents or limits some cockpit interactions.
- Makes high-speed driving significantly harder.
The player should be thinking:
"Do I stay up and risk getting shot, or duck and drive almost blind?"
10. Weapon System
Weapons should be deliberately tactile.
The game should not treat every gun as:
Aim → click → reload.
Instead, different weapons should have different manipulation requirements.
10.1 Service Pistol
Role: Reliable general-purpose weapon.
Characteristics:
- Moderate damage.
- Moderate recoil.
- Good accuracy.
- Magazine-fed.
- Fast reload.
- Effective against exposed drivers.
- Limited vehicle penetration.
Manual actions:
- Grab pistol.
- Aim.
- Fire.
- Remove empty magazine.
- Insert magazine.
- Rack slide if necessary.
This should be the first weapon implemented.
10.2 Revolver
Role: Heavy single-shot damage.
Characteristics:
- Very high per-shot damage.
- Six-round cylinder.
- Slow reload.
- Excellent intimidation factor.
- Strong penetration.
- Significant recoil.
Manual interactions:
- Open cylinder.
- Eject shells.
- Load rounds.
- Close cylinder.
- Fire.
A speedloader could later provide a faster reload option.
10.3 Sawed-Off Shotgun
Role: Close-range destruction.
Characteristics:
- Two shells.
- Massive close-range damage.
- Poor range.
- Strong recoil.
- Excellent against drivers at close range.
- Can damage vehicle components.
Manual sequence:
- Break open.
- Eject shells.
- Insert shells.
- Close.
- Fire.
This weapon should make passing at close range extremely dangerous.
10.4 Compact SMG
Role: Suppression.
Characteristics:
- Large magazine.
- High rate of fire.
- High recoil.
- Lower individual bullet damage.
- Excellent against exposed drivers.
- Can damage tyres and engine components.
The player should be able to fire it one-handed, but doing so should be inaccurate.
10.5 Hunting Rifle
Role: Precision weapon.
Characteristics:
- Extremely accurate.
- Slow reload.
- Powerful.
- Excellent at long range.
- Difficult to operate while driving.
The rifle encourages players to find moments of stability.
10.6 Future Experimental Weapons
These should be considered after the core game works.
Possible concepts:
- Flare gun.
- Nail gun.
- Homemade machine pistol.
- Heavy revolver.
- Lever-action rifle.
- Short rifle.
- Incendiary weapon.
- EMP-style fictional weapon.
- Vehicle-mounted weapon.
- Harpoon/grappling weapon.
- Deployable road hazard.
Not every weapon needs to be realistic.
SpeedBleed can gradually become more absurd as progression advances.
11. Weapon Damage Model
Avoid a single "enemy HP" value.
Drivers and vehicles should have separate damage states.
Driver
- Head.
- Torso.
- Arms.
- Potentially legs.
Vehicle
- Engine.
- Wheels.
- Brakes.
- Suspension.
- Fuel/coolant system.
- Body panels.
- Windshield.
- Lights.
This allows interesting situations.
For example:
A player may destroy an opponent's engine without killing the driver.
Or:
A driver may be wounded but still have a perfectly functioning vehicle.
12. Vehicle Damage
Damage should visibly and mechanically affect the car.
Potential states:
Light Damage
- Scratches.
- Cracked glass.
- Bent panels.
Medium Damage
- Smoke.
- Reduced engine performance.
- Steering imbalance.
- Suspension problems.
Heavy Damage
- Blown tyre.
- Severe overheating.
- Engine misfire.
- Broken lights.
- Major steering problems.
Critical Damage
- Engine failure.
- Vehicle becomes disabled.
- Driver remains alive.
This gives the player a reason to target different parts of an enemy vehicle.
13. Car Roster
Cars should have strong identities rather than simply being faster versions of each other.
13.1 The Street Rat
Small lightweight hatchback.
Strengths:
- Excellent acceleration.
- Very agile.
- Easy to control.
- Small target.
Weaknesses:
- Low armour.
- Low weapon capacity.
- Poor stability when hit.
Ideal starter car.
13.2 The Bruiser
Large American-style muscle car.
Strengths:
- Huge engine.
- Excellent straight-line speed.
- Heavy ramming damage.
- Large cockpit.
Weaknesses:
- Poor cornering.
- Heavy.
- High fuel/temperature demands.
13.3 The Widow
Lightweight rear-wheel-drive sports coupe.
Strengths:
- Excellent drifting.
- High top speed.
- Excellent cornering.
Weaknesses:
- Fragile.
- Difficult to control.
- Limited armour.
13.4 The Mule
Old utility pickup.
Strengths:
- Heavy.
- Durable.
- Excellent off-road performance.
- Large cargo capacity.
Weaknesses:
- Slow.
- High centre of gravity.
13.5 The Coffin
Old luxury sedan.
Strengths:
- Excellent armour potential.
- Stable.
- Large interior.
- Lots of upgrade space.
Weaknesses:
- Slow acceleration.
- Poor agility.
13.6 The Razor
Lightweight exotic/kit sports car.
Strengths:
- Extremely fast.
- Excellent brakes.
- Excellent handling.
Weaknesses:
- Expensive.
- Extremely fragile.
- Very limited armour.
13.7 Heavy Special
A ridiculous heavyweight vehicle.
Potentially:
- Armoured van.
- Old police interceptor.
- Taxi.
- Delivery truck.
- Funeral car.
- Hearse.
The roster should contain memorable silhouettes rather than generic supercars.
14. Vehicle Classes
Instead of simply assigning numerical tiers:
Class A — Light
Fast and fragile.
Class B — Balanced
General-purpose.
Class C — Heavy
Slow but durable.
Class D — Specialist
Unusual vehicles with unique characteristics.
Example:
A rally car might be fantastic on dirt but poor on asphalt.
This encourages track-specific builds.
15. Non-Linear Vehicle Tuning
Tuning should be one of SpeedBleed's major long-term systems.
Differential
Welded / Spool
- Aggressive oversteer.
- Excellent drifting.
- Strong throttle response.
- Difficult cornering.
Limited Slip
- Balanced traction.
- Predictable corner exits.
- Good all-round option.
Open Differential
- Forgiving.
- Poor power delivery.
- Useful for inexperienced drivers.
Suspension
Track Coilovers
- Fast response.
- Low body roll.
- Poor over rough terrain.
Rally Suspension
- Large travel.
- Handles debris and kerbs.
- More body movement.
Heavy Springs
- Supports armour.
- Resistant to impacts.
- Poor handling.
Gearbox
Short 5-Speed
- Excellent acceleration.
- Constant shifting.
Tall 4-Speed
- Easy to operate.
- Strong cruising performance.
- Poor acceleration.
Sequential Dogbox
- Very fast shifts.
- Requires skill.
- Potential mechanical failure if abused.
Armour
Lightweight Panels
- Maximum performance.
- Minimal protection.
Reinforced Doors
- Better side protection.
- Increased weight.
Armoured Windshield
- Protects the driver.
- Reduces visibility.
Roll Cage
- Improves crash survivability.
- Adds weight.
16. Other Tuning Categories
Future tuning should include:
- Engine.
- Intake.
- Exhaust.
- Cooling.
- Brakes.
- Tyres.
- Wheels.
- Steering rack.
- Weight distribution.
- Fuel tank.
- Battery.
- Armour.
- Glass.
- Differential.
- Gearbox.
- Suspension.
- Ride height.
- Brake bias.
The important rule:
No upgrade should simply be "better."
Every upgrade should create a new compromise.
17. Track Design Philosophy
Tracks should be designed specifically around the combat system.
A traditional racing game track can simply ask:
"Can you take this corner quickly?"
SpeedBleed should also ask:
"Can you take this corner while somebody is shooting at you?"
Track design should therefore include:
- Long sightlines.
- Blind corners.
- Cover.
- Narrow sections.
- Wide passing zones.
- Dangerous barriers.
- Jumps.
- Dirt sections.
- Alternate routes.
- Tunnels.
- Bridges.
- Chicanes.
- Pit lanes.
- Areas where cars can ram each other.
18. Track Concepts
Track 01 — Dead Mile
A rough industrial circuit.
Environment:
- Warehouses.
- Shipping containers.
- Railway crossings.
- Concrete barriers.
- Loading yards.
- Rusted machinery.
Gameplay:
- Beginner-friendly.
- Large sightlines.
- Several long straights.
- Good place to learn shooting while driving.
Track 02 — Blacktop County
Rural American-style road circuit.
Environment:
- Forest.
- Farmhouses.
- Power lines.
- Gas stations.
- Bridges.
- Dirt roads.
Gameplay:
- Fast.
- Long corners.
- Excellent for high-speed combat.
Track 03 — Redline City
Urban circuit.
Environment:
- Apartment blocks.
- Parking garages.
- Neon signs.
- Construction sites.
- Alleys.
Gameplay:
- Tight.
- Lots of ambush opportunities.
- Short sightlines.
- Heavy traffic-style obstacle placement.
Track 04 — Rust Valley
Industrial mountain circuit.
Environment:
- Quarry.
- Mines.
- Conveyor belts.
- Industrial facilities.
- Cliff roads.
Gameplay:
- Elevation changes.
- Dangerous drops.
- Narrow roads.
Track 05 — The Wastes
Desert circuit.
Environment:
- Sand.
- Abandoned military structures.
- Scrapyards.
- Dry lake beds.
Gameplay:
- Large open spaces.
- Long-range weapons become valuable.
- Off-road shortcuts.
Track 06 — Hellring
Late-game purpose-built death circuit.
Environment:
- Stadium.
- Concrete walls.
- Arena structures.
- Giant signs.
- Spectators.
- Floodlights.
Gameplay:
- Short laps.
- Constant combat.
- Extremely aggressive racing.
19. Track Hazards
Potential hazards:
- Oil slicks.
- Gravel.
- Puddles.
- Broken glass.
- Debris.
- Construction barriers.
- Train crossings.
- Falling objects.
- Explosive barrels.
- Collapsing structures.
- Moving traffic.
- Roadblocks.
- Narrow bridges.
Hazards should generally create opportunities rather than simply punish the player.
20. Alternative Routes
Tracks should contain optional routes.
For example:
MAIN ROAD
|
+---- SAFE ROUTE --------+
| |
+---- SHORTCUT -----------+
|
HIGH RISK
LOW VISIBILITY
A shortcut might:
- Save 5 seconds.
- Destroy a tyre.
- Force a jump.
- Pass through an industrial building.
- Put the player directly beside enemies.
This creates strategic racing.
21. Pit Stop System
The pit lane should not simply be:
Drive into pit → car instantly repairs.
Instead, the pit stop can become a brief tactical decision.
Potential services:
- Repair vehicle.
- Heal driver.
- Replace tyres.
- Refill ammunition.
- Cool engine.
- Repair weapon.
- Adjust damage systems.
For the initial implementation, automate most of this.
Later, introduce optional manual interactions.
22. Rival AI
AI should have personalities.
Instead of every rival being identical, define archetypes.
The Racer
- Rarely shoots.
- Takes optimal racing lines.
- Prioritises podium finish.
The Killer
- Constantly targets drivers.
- Ignores racing lines when necessary.
The Bully
- Rams opponents.
- Uses heavy cars.
The Sniper
- Maintains distance.
- Uses accurate weapons.
The Coward
- Runs when damaged.
- Avoids combat.
The Maniac
- Takes shortcuts.
- Rams everything.
- Makes unpredictable decisions.
23. AI Decision Making
AI should consider:
Race Position
+
Vehicle Health
+
Driver Health
+
Ammo
+
Enemy Distance
+
Weapon Range
+
Corner Difficulty
+
Current Strategy
=
Next Action
Potential actions:
- Accelerate.
- Brake.
- Overtake.
- Defend.
- Ram.
- Shoot.
- Duck.
- Take shortcut.
- Enter pit.
- Retreat.
- Target leader.
This creates much more interesting opponents than simply giving every AI a perfect racing line.
24. Rival Personalities and Rivalry System
Named rivals can become recurring characters.
Example:
"The Undertaker"
Vehicle: Black hearse Weapon: Revolver Personality: Calm, intimidating Strategy: Heavy braking + precision shots
"Lucky"
Vehicle: Small hatchback Weapon: SMG Personality: Reckless Strategy: Constant overtaking and close-range fire
"Big Red"
Vehicle: Muscle car Weapon: Shotgun Personality: Aggressive Strategy: Ram first, shoot second
RivalsI'm taking your should remember previous encounters.
Potential dialogue:
"You again."
"That car looks worse than last time."
"I'm taking your head this time."
This gives the campaign personality without requiring elaborate cutscenes.
25. Progression
Progression should reward both racing and combat.
Potential currencies:
Cash
Used for:
- Car parts.
- Weapons.
- Repairs.
- Ammunition.
Reputation
Used to:
- Unlock races.
- Challenge rivals.
- Enter higher-risk events.
Kill Count / Infamy
Optional secondary statistic used for:
- Cosmetics.
- Special challenges.
- Unlocking violent-themed events.
26. Garage
The garage should be the player's main progression hub.
Sections:
GARAGE
├── Cars
├── Engine
├── Suspension
├── Drivetrain
├── Armour
├── Weapons
├── Ammunition
├── Cosmetics
└── Test Track
The player should be able to inspect their vehicle in first-person.
The garage itself can eventually become an explorable environment.
27. Weapon Shop
The weapon shop should be deliberately stylised.
Potential categories:
- Pistols.
- Revolvers.
- Shotguns.
- SMGs.
- Rifles.
- Experimental weapons.
Each weapon should display:
- Damage.
- Accuracy.
- Range.
- Recoil.
- Magazine/capacity.
- Reload complexity.
- Penetration.
- Weight.
But the most important stat should be:
How difficult is this weapon to operate while driving?
28. Game Modes
Quick Race
Choose:
- Car.
- Weapon.
- Track.
- Number of opponents.
- Laps.
Immediately start.
This should be the first fully playable mode.
Elimination
No requirement to finish first.
Kill every rival.
Time Attack
No combat.
Pure driving.
Useful for testing vehicle builds.
Survival
Waves of increasingly aggressive opponents.
Rival Race
One-on-one race against a named character.
Championship
Series of races where:
- Money accumulates.
- Damage carries over.
- Rivals develop.
- Vehicle setup becomes increasingly important.
29. PS1 Rendering Pipeline
Target resolution:
- 320×240.
- Optional 400×240 widescreen.
Use nearest-neighbour scaling.
Potential rendering effects:
- Vertex snapping.
- Vertex jitter.
- Affine-style texture distortion.
- Low-resolution render target.
- Colour quantisation.
- Ordered dithering.
- Fog.
- Screen-space noise.
- Pixelated particles.
Do not overuse post-processing.
The game should look like an actual low-poly game rather than a modern game with a "PS1 filter."
30. Art Direction
Cars should generally target:
- Approximately 300–1,500 polygons depending on importance.
- Low-resolution texture atlases.
- Strong silhouettes.
- Minimal interior geometry.
- Deliberate texture seams.
- Hand-painted details.
Track props can be extremely simple.
Examples:
- 20–50 polygon barrels.
- Low-poly tyres.
- Simple boxes.
- Billboard signs.
- Repeated modular buildings.
Reuse is encouraged.
A good PS1 environment can be built from surprisingly few assets.
31. Cockpit Art Direction
Cockpits are disproportionately important because the player spends most of the game inside them.
Each cockpit should have:
- Distinct dashboard.
- Steering wheel.
- Gear stick.
- Instrument cluster.
- Weapon storage.
- Windshield.
- Side windows.
- Door.
- Interior lighting.
Small details create personality:
- Air fresheners.
- Cigarette packets.
- Receipts.
- Stickers.
- Dashboard toys.
- Old radios.
- Racing numbers.
- Cracked gauges.
- Warning lights.
These details should be low-poly and cheap.
32. Audio Direction
Audio should strongly communicate speed and danger.
Important sounds:
- Engine idle.
- Engine acceleration.
- Gear changes.
- Clutch slip.
- Tyre squeal.
- Suspension impacts.
- Collision.
- Wind.
- Bullet impacts.
- Glass breaking.
- Weapon mechanisms.
- Shell ejection.
- Cockpit rattling.
- Damage alarms.
The cockpit should sound different from the outside world.
33. Music
The soundtrack should combine:
- Breakbeat.
- Industrial techno.
- Drum and bass.
- Distorted electronic music.
- Aggressive synths.
- Dark ambient sections.
Music should ideally respond to gameplay.
For example:
Normal Racing
↓
Combat Begins
↓
Music Intensifies
↓
Vehicle Critical
↓
Distorted / High-intensity layer
34. UI
UI should feel like late-90s/early-2000s racing software.
Possible design language:
- CRT-style typography.
- Green/red monochrome displays.
- Chunky meters.
- Pixel icons.
- Industrial warning labels.
- Cheap-looking sponsor graphics.
- Track-map displays.
Avoid modern minimalist UI.
35. Recommended Development Phases
The following order is the recommended production sequence.
The important distinction is that phases describe project milestones, while developer workstreams describe who should be working on what at the same time.
Phase 0 — Pre-Production & Technical Planning
Goal
Establish the technical foundations before significant content production begins.
Developer 1
- Create Godot project.
- Establish project architecture.
- Set input system.
- Establish scene conventions.
- Establish reusable vehicle interfaces.
- Establish damage interfaces.
- Establish save-data structure.
Developer 2
- Establish Blender/Sprytile workflow.
- Create PS1 texture atlas.
- Establish polygon budgets.
- Create placeholder car.
- Create placeholder cockpit.
- Create modular track pieces.
Developer 3
- Establish audio pipeline.
- Create placeholder UI.
- Create temporary music system.
- Create debug HUD.
- Establish AI design document.
Milestone
The project opens, runs, renders correctly, and has agreed technical conventions.
Phase 1 — PS1 Rendering & Prototype Assets
Goal
Get the game looking like SpeedBleed before building the entire game.
Developer 1
- Low-resolution viewport.
- Vertex jitter.
- Dithering.
- Texture filtering.
- Fog.
- Camera system.
Developer 2
Create:
- Placeholder vehicle.
- Cockpit.
- Track.
- Weapon.
- Basic props.
Developer 3
Create:
- Temporary UI.
- Speedometer.
- Gear indicator.
- Damage display.
- Placeholder sound effects.
Milestone
A player can sit inside a crude SpeedBleed cockpit on a crude PS1 track.
Phase 2 — Vehicle Physics
Goal
Make driving fun before adding combat.
Developer 1
Build:
- Suspension.
- Wheel grip.
- Steering.
- Engine.
- Torque curve.
- Gearbox.
- Braking.
- Handbrake.
- Reverse.
- Damage hooks.
Developer 2
Provide:
- Vehicle collision meshes.
- Wheels.
- Suspension visualisation.
- First proper car model.
Developer 3
Build:
- Engine audio.
- Tyre audio.
- Skid effects.
- Speed effects.
- Debug telemetry UI.
Milestone
The player can drive a complete test vehicle around a figure-eight track and it feels good.
Do not proceed until this is fun.
Phase 3 — Cockpit & Two-Hand Interaction
Goal
Implement the feature that makes SpeedBleed different from conventional racing games.
Developer 1
Build:
- Hand state machines.
- Mouse raycasting.
- Interaction system.
- Steering interaction.
- Gear interaction.
- Hand IK.
- Hand ownership rules.
Developer 2
Build:
- Proper cockpit.
- Steering wheel.
- Gear stick.
- Dashboard.
- Weapon holster.
- Hands/arms.
Developer 3
Build:
- Interaction UI.
- Hand prompts.
- Interaction sounds.
- Camera feedback.
Milestone
The player can drive using physical cockpit controls.
Phase 4 — Driving Systems & Ducking
Goal
Complete the core driving interaction.
Implement:
- Manual transmission.
- Clutch.
- Gear grinding.
- Steering assist options.
- Handbrake.
- Ducking.
- Camera movement.
- G-force effects.
- Collision reactions.
Milestone
The player can race around a track while manually managing:
- Steering.
- Gears.
- Braking.
- Ducking.
Phase 5 — First Weapon & Combat Prototype
Goal
Prove that shooting while driving is actually fun.
Start with one pistol.
Do not build four or five guns yet.
Implement:
- Weapon pickup.
- Aim.
- Fire.
- Recoil.
- Magazine.
- Reload.
- Slide operation.
- Hitscan.
- Driver hitbox.
- Vehicle hitbox.
- Damage.
- Death.
Developer 1
Weapon framework and combat logic.
Developer 2
Pistol model and animations.
Developer 3
Gun sounds, hit effects and UI.
Milestone
Player vs stationary enemy vehicle.
If this is not fun, stop and improve it before continuing.
Phase 6 — Combat-Racing Prototype
Goal
Combine driving and shooting.
Create:
- One complete track.
- Player vehicle.
- 2–4 enemy vehicles.
- Basic racing positions.
- Basic weapons.
- Driver health.
- Vehicle health.
- Kill condition.
- Podium condition.
Milestone
A complete 5–10 minute playable race exists.
This is the first vertical slice.
Phase 7 — Track, Race & Pit Systems
Goal
Turn the prototype into a real racing game.
Implement:
- Checkpoints.
- Lap counting.
- Race positions.
- Respawning.
- Wrong-way detection.
- Pit lane.
- Repair.
- Driver healing.
- Ammunition.
- Race results.
- Cash rewards.
Developer 2
Build the first production-quality track.
Developer 3
Build race UI and results screens.
Phase 8 — Rival AI
Goal
Create opponents that feel like actual drivers.
Implement:
- Waypoint navigation.
- Racing lines.
- Overtaking.
- Braking.
- Collision avoidance.
- Combat targeting.
- Shooting.
- Ducking.
- Damage reactions.
- Pit behaviour.
Then add personality archetypes.
Milestone
An entire race can be played without developer-controlled enemies.
Phase 9 — Vehicle Roster & Tuning
Goal
Turn the prototype into a system-driven game.
Create the first 4–6 cars.
Each car should have a strong identity.
Implement:
- Differential.
- Suspension.
- Gearbox.
- Engine.
- Brakes.
- Tyres.
- Armour.
- Weight.
- Cooling.
Milestone
Two players using different cars and setups should genuinely approach the same track differently.
Phase 10 — Weapons & Garage
Goal
Expand player choice.
Add:
- Revolver.
- Shotgun.
- SMG.
- Optional rifle.
Build:
- Weapon shop.
- Garage.
- Upgrade shop.
- Ammunition.
- Loadout management.
The player can now create a meaningful build.
Phase 11 — Progression & Content
Goal
Turn the game loop into a complete progression system.
Implement:
- Cash.
- Reputation.
- Unlocks.
- Car unlocks.
- Weapon unlocks.
- Rival unlocks.
- Championships.
- Statistics.
- Challenges.
- Cosmetics.
Potential statistics:
- Races won.
- Races lost.
- Driver kills.
- Vehicle kills.
- Longest drift.
- Highest speed.
- Most expensive repair.
- Most damage survived.
- Favourite weapon.
- Favourite vehicle.
- Total distance driven.
Phase 12 — Full Content Production
Goal
Create the actual game rather than the prototype.
Target approximately:
Vehicles
6–8 cars.
Weapons
4–6 meaningful weapons.
Tracks
5–7 tracks.
Rivals
8–12 named drivers.
Game Modes
3–5 polished modes.
Quality is more important than quantity.
Phase 13 — Polish & Playtesting
This should be a substantial phase rather than a final week of bug fixing.
Test:
- Driving feel.
- Steering sensitivity.
- Gear difficulty.
- Weapon balance.
- Ducking.
- AI difficulty.
- Race length.
- Track readability.
- Damage.
- Tuning.
- Economy.
- Progression.
- Performance.
- Controller support.
- Mouse sensitivity.
- Accessibility.
Phase 14 — Release Candidate
Final work:
- Save corruption testing.
- Crash testing.
- Performance optimisation.
- Graphics options.
- Audio options.
- Control remapping.
- Resolution scaling.
- Controller support.
- Tutorial.
- First-time-user experience.
- Settings.
- Credits.
- Final branding.
Future Phase — Multiplayer
Multiplayer should not be developed until the single-player game has a stable, enjoyable combat-racing loop.
Potential architecture:
- Godot high-level multiplayer.
- ENet.
- Steam networking / GodotSteam where appropriate.
- Host-authoritative racing simulation initially.
- Server validation of combat events.
- Synchronised vehicle transforms.
- Synchronised damage states.
- Synchronised weapon actions.
Network:
- Car position.
- Rotation.
- Velocity where required.
- Lap/checkpoint state.
- Damage.
- Driver state.
- Weapon state.
- Shot events.
- Race results.
Do not initially synchronise every individual hand-bone transform.
The hands can use local animation while the important gameplay state is synchronised.
36. Recommended Developer Work Order
This is the most important practical section for the team.
Developer 1 — Gameplay Lead
Work approximately in this order:
- Project architecture.
- Input system.
- Vehicle physics.
- Camera.
- Cockpit interaction.
- Hand state machines.
- Gearbox.
- Ducking.
- Weapon framework.
- Damage framework.
- Race system.
- AI integration.
- Tuning.
- Garage.
- Progression.
- Save/load.
- Optimisation.
- Multiplayer foundations.
The rule for Developer 1:
Do not build content-specific systems when a reusable system can be built instead.
Developer 2 — Art Lead
Work approximately in this order:
- PS1 art pipeline.
- Placeholder cockpit.
- Placeholder car.
- Placeholder weapon.
- Modular track kit.
- Production player car.
- Production cockpit.
- First track.
- First weapon.
- Additional cars.
- Additional weapons.
- Track environments.
- Damage models.
- Garage environment.
- UI assets.
- Remaining tracks.
- Cosmetics.
The rule for Developer 2:
Build reusable kits wherever possible.
One good building module should create an concrete barrier should be usable in ten locations.
One good building module should create an entire street.
One good texture atlas should support dozens of objects.
Developer 3 — AI / Audio / UI Lead
Work approximately in this order:
- Debug HUD.
- Placeholder audio.
- Engine audio.
- Weapon audio.
- Basic race UI.
- Basic AI.
- Racing AI.
- Combat AI.
- Rival personalities.
- Menu systems.
- Garage UI.
- Progression UI.
- Music.
- VFX.
- Accessibility.
- Polish and QA.
The rule for Developer 3:
Prioritise feedback.
Every important action should have:
- Visual feedback.
- Audio feedback.
- Mechanical feedback.
37. Parallel Development Strategy
The team should avoid waiting for each other.
For example:
Developer 1
Vehicle Physics ────────────────┐
↓
Combat Prototype
↑
Developer 2 │
Car + Cockpit ─────────────────┤
│
Developer 3 │
Audio + UI ────────────────────┘
Then:
Developer 1
AI Systems ─────────────────────┐
│
Developer 2 ↓
Track Production ─────────→ Playable Race
↑
Developer 3 │
AI Audio + UI ──────────────────┘
This prevents the common small-team problem where:
Programmer finishes system → waits for artist → artist finishes asset → waits for programmer → nobody progresses.
38. Vertical Slice Strategy
The project should aim for a vertical slice as early as possible.
The first real target should be:
One car + one track + one pistol + four rivals + complete race.
It does not need:
- Career mode.
- Multiple cars.
- Ten weapons.
- Multiplayer.
- Complex garage.
- Advanced progression.
If that one race is not fun, adding more content will not solve the problem.
The vertical slice should answer five questions:
- Is driving fun?
- Is manually controlling the cockpit fun?
- Is shooting while driving fun?
- Are AI opponents fun?
- Does winning a race feel satisfying?
If the answer to all five is yes, the project has a strong foundation.
39. Tutorial Design
Do not begin the game by dumping a conventional control manual on the player.
Teach through situations.
Tutorial 1 — Drive
"Start the engine."
Teach:
- Throttle.
- Brake.
- Steering.
Tutorial 2 — Shift
Force the player to use gears.
Tutorial 3 — Shoot
Place a stationary target.
Tutorial 4 — Duck
Have an enemy shoot at the player.
Tutorial 5 — Race
Combine everything.
Tutorial 6 — Survive
Introduce multiple enemies.
The player should understand the game's fundamental tension within approximately 10 minutes.
40. Accessibility
Because the physical cockpit system is intentionally unusual, accessibility options are important.
Possible assists:
- Steering assist.
- Automatic gear shifting.
- Simplified reload.
- Weapon quick-reload.
- Reduced recoil.
- Increased driver health.
- Reduced incoming damage.
- Hold/toggle ducking.
- Alternative mouse controls.
- Controller support.
- Aim assist.
- Reduced camera shake.
These should not undermine the default experience.
They simply allow more players to enjoy the game.
41. Difficulty Modes
Street
- Generous driver health.
- Lower AI accuracy.
- More ammunition.
- Reduced damage.
Standard
The intended experience.
Bloodline
- Aggressive AI.
- Scarce ammunition.
- Increased damage.
- Less forgiving vehicle handling.
Nightmare
For experienced players.
- Extremely aggressive AI.
- Limited repairs.
- Severe damage.
- Minimal assists.
42. Ideas for Dynamic Race Events
To keep races unpredictable:
- Rival crashes ahead.
- AI vehicle catches fire.
- Police roadblock.
- Bridge opens.
- Train crosses track.
- Construction machinery moves.
- Shortcut becomes blocked.
- Rival deliberately rams player.
- Heavy vehicle enters track.
- Weather changes.
- Debris falls onto road.
These should be used sparingly.
The player should feel:
"That was insane."
Not:
"The game randomly screwed me."
43. Vehicle Personality
Cars should have different cockpit experiences.
A tiny hatchback:
- Small steering wheel.
- Narrow windshield.
- Cheap plastic dashboard.
Muscle car:
- Huge steering wheel.
- Wide dashboard.
- Vibrating engine.
- Large gear lever.
Rally car:
- Roll cage.
- Sparse dashboard.
- Racing gauges.
Luxury sedan:
- Large windshield.
- Soft suspension.
- Heavy steering.
This makes choosing a car an experiential decision rather than merely a stat decision.
44. Environmental Storytelling
The world should imply that SpeedBleed takes place in a strange underground racing culture.
Examples:
- Illegal race posters.
- Hand-painted sponsor logos.
- Driver wanted posters.
- Abandoned race cars.
- Graffiti.
- Crude advertisements.
- Trackside spectators.
- Mechanics.
- Rival photos.
- Trophy walls.
- Newspaper clippings.
The player should gradually learn about the world without requiring long exposition.
45. Branding Ideas
Possible SpeedBleed visual motifs:
- Red warning labels.
- Black/white/red colour scheme.
- Industrial typography.
- Fake motorsport sponsors.
- Skull/helmet imagery.
- Cheap racing stickers.
- Number plates.
- "NO BRAKES / NO MERCY" style slogans.
- Fake safety warnings.
Potential fictional sponsors:
- Bloodline Motors
- Deadbolt Tyres
- Redline Fuels
- Blackout Racing
- Graveyard Performance
- Hellcat Lubricants
- CrashKing
- Last Lap Industries
These should be fictional rather than directly referencing real brands.
46. Potential Unlockable Cosmetics
Cosmetics can be deliberately silly.
Car:
- Dice hanging from mirror.
- Fuzzy dice.
- Air fresheners.
- Stickers.
- Dashboard toys.
- Seat covers.
- Racing numbers.
- Custom licence plates.
- Spray-painted panels.
- Rust variants.
- Neon.
- Headlight colours.
Driver:
- Gloves.
- Jackets.
- Helmets.
- Masks.
- Sunglasses.
Weapon:
- Engraving.
- Tape.
- Paint.
- Scratched finishes.
- Different grips.
The cosmetics should reinforce the late-90s underground racing aesthetic.
47. Long-Term Ideas
If the core game succeeds, potential future systems include:
Career Mode
Follow one driver from nobody to underground champion.
Rival System
Rivals remember:
- Wins.
- Losses.
- Kills.
- Humiliations.
- Vehicle destruction.
Crew System
Recruit:
- Mechanic.
- Spotter.
- Gun specialist.
- Driver.
- Fixer.
Garage Expansion
Start with a terrible workshop and gradually turn it into a professional racing facility.
Dynamic Championship
Different tracks and rival lineups each season.
Challenge Events
Examples:
- Kill three rivals with a revolver.
- Win with one tyre damaged.
- Finish without braking.
- Win using a stock car.
- Win while carrying no spare ammunition.
- Kill the race leader before lap two.
48. Technical Verification Plan
Vehicle Physics Tests
Verify:
- Gear ratios.
- Torque.
- Acceleration.
- Braking.
- Slip.
- Suspension.
- Weight transfer.
- Differential behaviour.
Hand State Tests
Verify impossible states cannot occur.
Examples:
Right Hand + Gun
↓
Cannot simultaneously grab Gear Stick
Left Hand + Steering Wheel
↓
Can steer
Both Hands + Wheel
↓
Maximum steering stability
Weapon Tests
Verify:
- Reload.
- Magazine state.
- Chamber state.
- Ammunition.
- Hit detection.
- Damage.
- Weapon drop.
- Weapon switching.
Race Tests
Verify:
- Checkpoints.
- Lap counting.
- Position.
- Wrong-way detection.
- Finish conditions.
- Elimination conditions.
- Pit lane.
- Race restart.
49. Manual Gameplay Testing
Every major milestone should be manually tested.
Feel Test
Can the player comfortably:
- Start.
- Accelerate.
- Brake.
- Turn.
- Shift.
- Drift.
- Recover from a mistake?
Combat Test
Can the player:
- Draw.
- Aim.
- Shoot.
- Reload.
- Steer.
- Duck?
- Recover from being attacked?
Race Test
Can the player:
- Complete a lap?
- Overtake?
- Be overtaken?
- Recover from a crash?
- Win through racing?
- Win through elimination?
PS1 Authenticity Test
Check:
- Texture filtering.
- Vertex wobble.
- Polygon density.
- Fog.
- Dithering.
- Resolution.
- Colour palette.
- Animation style.
50. Performance Targets
Initial targets should be:
- Stable 60 FPS on modest PC hardware.
- Fast loading.
- Low memory usage.
- Low-poly collision meshes.
- Efficient AI.
- Minimal unnecessary physics calculations.
Because the game intentionally uses low-poly assets, performance should primarily be protected from:
- Excessive physics bodies.
- Too many raycasts.
- Poor AI pathfinding.
- Overly expensive shaders.
- Unnecessary high-resolution effects.
51. Development Priority Rules
When deciding whether to add a feature, ask:
Priority 1 — Does it make driving better?
If yes, prioritise.
Priority 2 — Does it make combat better?
If yes, prioritise.
Priority 3 — Does it make the cockpit interaction better?
If yes, prioritise.
Priority 4 — Does it reinforce the game's identity?
If yes, consider it.
Priority 5 — Is it merely content?
Only build it after the systems underneath it are stable.
52. Features That Should Be Deliberately Delayed
Do not start with:
- Multiplayer.
- Career mode.
- Dozens of cars.
- Huge weapon roster.
- Advanced economy.
- Complex cosmetics.
- Massive open-world environments.
- Online progression.
- Procedural tracks.
These are all potential features, but none prove that SpeedBleed is fun.
The game lives or dies on:
Driving + cockpit interaction + combat.
53. Final Recommended Production Order
The clean production sequence is therefore:
PHASE 0
Planning
↓
PHASE 1
PS1 Rendering + Placeholder Assets
↓
PHASE 2
Vehicle Physics
↓
PHASE 3
Cockpit + Two-Hand Interaction
↓
PHASE 4
Manual Driving + Ducking
↓
PHASE 5
First Weapon
↓
PHASE 6
Combat Racing Vertical Slice
↓
PHASE 7
Track + Race + Pit Systems
↓
PHASE 8
Rival AI
↓
PHASE 9
Multiple Cars + Tuning
↓
PHASE 10
Multiple Weapons + Garage
↓
PHASE 11
Progression + Rivals + Game Modes
↓
PHASE 12
Full Content
↓
PHASE 13
Polish + Balance + QA
↓
PHASE 14
Release Candidate
↓
FUTURE
Multiplayer + Career Expansion
This ordering gives the team a playable game very early and ensures that every later phase builds on something already proven.
54. The Most Important Development Rule
SpeedBleed should be developed around one question:
"Is this making the moment-to-moment experience of driving, fighting and desperately trying to survive more fun?"
If yes, build it.
If no, postpone it.
The game's strongest potential differentiator is not its PS1 graphics, its number of cars, or its number of guns.
It is the combination of:
Manual cockpit control + high-speed racing + physical weapon handling + tactical ducking + vehicle damage + non-linear tuning.