How we create our game

Here are the distinct phases I nearly always follow when I create a game.


Phase 0: Destruction (1% of the time)
I start form one of my old game and I delete everything that won't be useful.
And I correct the thousands of bugs ^^.
It's like launching a tornado in sim city!

- The game launch, no crash at all.
- We can navigate in the menu (even if they aren't final)
- When the game is launched, we see a least a map.

In this part, I must rethink of all the existing code and improve it.
It allows me to have a strong base code.


Phase 1: Basics (9% of the time)
I add the basic gameplay elements of the game.
(Like the inventory in an RPG or troop selection in an RTS)

- Building a strong core for the game, with all the needed elements.
- Having a basic AI able to handle the basic actions.
- Heavily test everything to avoid later problem.
In order to test everything, I add random AI, lots of them.
Even in stupid situation that may not occur in the final game.

It's a part I like because we can imagine our game and we can see his birth.
It's also the part where most of my games are stopped, here’s why:
- We clearly see what the game can be, so if we don't see how to make it fun, it's better to stop.
- We see that the game we want to make is too big for us.
- We see that a game is not as original as we think.


Phase 2: Beta (80% of the time)
Most of the game design elements are added and working.

The aim is:
- Having a HUD allowing to do all the implemented elements.
- Having a first version of tutorial / story.
- To be able to play a full game.
- To show the game.
Placeholder graphics and audio remain if they don't block the game.
In this part, we see our dreamed game be true.
It's really an exciting part, even if our design evolve from our dreamed game.


Phase 3: Polish (90% of the time)
Everything is improved.

- A perfect game that can be sale
- Final graphic and audio
- easy to handle HUD
- Final tutorial and history
- The game is equilibrate
In this part, I ask lots of people to test the game.
Use feedback to improve the game and then ask to someone else.
(Asking several times the same people give less result)

It's the only part I don't like.
The game is nearly finished, and I can see that my game is great.
But I'm the only one... And this phase take lots of times.

Now that the game is 180% done I can start to sell it ^^.
(Polish phase is as long as the first 90% of the game)

Age OF Fantasy 0.5

The new version of my RTS engine, named "Age of Fantasy"
here's the engine:
here's the source Code:

If you are interest, you can use it freely, and even sell game with.
- add in the credit "original engine from Laurent Goethals (Andreil Game)"
- add me in the graphic part if you use some graphics.
- tell me what you've done!

What the engine have:
- map generator + Mini map (several size and map options)
- fog of war
- Peasant AI (fight + gathering)
- Evolution of the peasant (bowmen / warrior / barbarian)
- Building and evolution of a building.
It's all for now. (No campaign for now)

I'm not sure when I will make a new version, not before months ^^.
As it's a big project, it’s better for me to make a lot of pause.
Like that each time I rework, I improve the engine a little.

Playing with a 'non-player'

To play with a non-player, we must carefully choose the game and the way to play.
There is two ways to play:
- Competitive: We must crush the opponent
- Cooperation: We must help the other to win
A lot of game have the two aspect (Team Fortress 2, Battlefield 1942, Age of Empire).

These two ways unlock distinct way of playing:
- Competitive, you must be the best => new players are slaughtered => not fun for a non-player (and maybe hate for this game)
- Cooperation, the alliance must survive => Good player help the new one, and even train them => non-player level increase.
In short, if you want to play with a non-player, choose a cooperation game.

How to make a cooperation and competitive game more interesting for a non-player?
Example with ‘Battlefield 1942’
- Be in the same alliance
- Increase the bot number => less pressure on the non-player
- Play with him and protect him, even if you lose the game.
It’s also a new way to play for the experienced player, it can be really fun

I try to make my games playable for everybody (you, your mother, your girlfriend…) so they all have:
- A cooperation mode (or alliance setting)
- all public graphics (not bloody, not abstract)
- Easy to take in hand.
I’m really proud when I see players and non-player playing together one of my games.
(Done with ‘Pioneerz’ and ‘Dark Seal’)


Sound design is an important aspect of a game, but it’s often done in the rush.
When it’s correctly done, it’s possible to make a good and original audio with little work for the programmer.
Example with one my student project:



‘Aube et crépuscule’ is a strategy game where light and shadow fight. The game is composed of 8 music, there are can be played together. In the game, 4 music are always played.
Without details, the music and their volume are chose based on our type and on the plants surrounding us.

It take me 2 days (out of 6 months) to make it work and to make the sound designer happy.
The result is a procedural music that is never repetitive with a great immersion.
(The sound designer was really good, so the result is great)

If one day I must choose between a great graphic and a great sound designer, I’ll choose the last. As great sound design is rare, the game will more easily be noticed.
Here my youtube channel with some of my favorite video game music.

And you, what’s your favorite game music?

Do a PacMan

Creating a game is a common dream, but few people are doing it.
A MMO, a RPG, a RTS, it can’t be done like that.
(My first project was never finished, it was a bubble bobble. It was too hard for me)

So I’ve search for the perfect project to begin: Pac-Man
I know it’s not sexy, so why doing it?

1) Pac-Man is easy to do:
- It’s based on: moving, eating dots chase/flee the ghosts.
- The interacting elements are simple: Ghost, walls and dots.
- The comportment of the elements is easy to understand.
- It can be done alone

2) Pac-Man show most of video game problems :
- Data management (I eat the dots but so does the ghosts)
- Drawing (I’ve forget the transparency, we can’t see anything)
- Controls (Pac-man is so hard to move that I keep hurting the wall)
- AI (The ghosts are so stupid that the game isn’t fun)
- Music and sound

3)  Pac-Man is easy to extend, you can add features of your dream’s game :
- Multi-player game via network
- Addition of dialog and cinematic
- Better graphics
- 3D game
- Random map generation
- …

4)  When it’s done, you’ll have:
- A better understanding of your limit
- A better credibility for searching co-worker.
- More confidence in your capacity.

In short, everyone must do a Pac-man, be proud of it and show it.
Only later you should talk about your dream project.
(This advice is also true for professional who just start as an indie)
If you’ve made a Pac-man, show it in the comments!


Make a free website with emyspot - Signaler un contenu illicite sur ce site