The reason game development feels difficult isn't because you lack talent [AI Game Development Record #4]

You are a legend.

Thank you for opening this article again today. This is #4 of a 7-part series.

Back issues so far↓↓↓
#1

#2

#3

If you have read this far, you might already have a vague realization: that game development is not as distant a world as you once thought.

If you use AI, the code will appear. Things that actually work take shape faster than you imagined. 'Wait, it runs this easily?' I was quite moved the moment it actually worked.

But at the same time, some of you might be thinking, 'Even so, isn't making a game still difficult?'

I thought so too at first. But in reality, bringing a game to life was much easier than I had expected.

So, why does game development feel difficult? It is because you try to complete everything all at once.

The reason it is difficult is not because you lack talent.

The reason people stumble in game development is, in many cases, simply because they don't know 'what order to build things in.'

Trying to create a beautiful screen from the start. Trying to think of the entire interesting scenario from the start. Trying to include music, effects, and rankings from the start. If you do that, it naturally becomes exhausting. There is just too much to do.

Games have many elements. Rules. Characters. Controls. Screens. Sound. Effects. Difficulty. Sense of accomplishment. Mechanisms that make you want to play again. If you try to do all of these from the start, beginners will almost certainly stop.

If I had tried to do everything from the start, I probably wouldn't have kept going.

That is why 'order' is more important than talent. First, decide on a small scale what kind of game you want to make. Next, shape it into a playable state. After that, refine the feel and the visuals. Just by thinking in this order, the difficulty changes significantly.

The first thing to decide is the 'core of the gameplay'.

In my case, when I make a game, I first decide on the 'core of the gameplay.' You don't need to overthink it. Just decide one thing: 'What is fun to do in this game?'

For a tissue character test game, it's 'writing characters without tearing the tissue.' For a falling-block puzzle, it's 'lining up four of the same animals to clear them.' For a shooter, it's 'shooting while dodging enemies.' For now, just this is enough.

You don't need to think about the world-building, the story, or the visuals perfectly from the start. In fact, if you try to think about everything from the beginning, you will get stuck.

Decide only this: 'What makes it fun?' This is the first step.

Next, make it work in its simplest form.

Once the core of the gameplay is decided, the next step is to make it work in its simplest form. The important thing here is not to be greedy. Do not try to create a finished product from the start. Title screens, elaborate effects, rankings, and sound should all be put off for later.

First, confirm whether the core of the gameplay holds up. If it's a tissue-writing test game, first check if you can write on the screen. Can you create a gauge that simulates pen pressure? If the gauge goes up, can you create an effect where it tears? If it works to this extent, it is sufficient as a first form.

If it's a falling-block puzzle, first check if the blocks fall. Can you move them left and right? Do they disappear when the same ones line up? This is enough.

Whether it looks pretty can be thought about later. Whether it's fun can also be thought about after you've made it move.

You cannot improve something that doesn't move. But you can improve something that does. This is very important.

Finally, add the sense of satisfaction.

Once it works in its simplest form, you can gradually add a sense of satisfaction from there.

Add a satisfying sound when you succeed. Add an effect that makes you want to try again when you fail. Tidy up the visuals a little. Display the score. Adjust the difficulty slightly. Add a mechanism that makes you want to show it to someone after playing. These are the parts you add later.

You don't have to try to put everything in from the start. Make it move first. Then tidy it up. This is the order.

Learning to ride a bicycle is probably the same. No one can turn, stop, and ride long distances perfectly from the start. First, you take one pedal stroke. Once you can do that, you take two. You gradually increase the number of things you 'can do'.

Game development is surely the same. The character moved. Objects fell. Collision detection was added. The score increased. Even if each of those things is small, when they pile up, they properly become a game. I realized this many times while making games with AI.

AI makes it easier to follow this order.

With AI, it becomes much easier to proceed with this flow.

'I want to make a game like this,' 'First, please make it in its simplest form,' 'Next, please fix this part,' 'Please make it more fun to play.' You can consult in stages like this.

You don't have to try to complete it all at once. In fact, if you try to complete it all at once, your instructions to the AI will become complicated. As a result, it will be harder to succeed.

Whether you are asking an AI or thinking as a human, the order is important. Decide the core of the gameplay. Make it move in its simplest form. Add the sense of satisfaction. If you proceed in this order, it becomes much easier for beginners to create.

Game development should be open to more people.

Game development should be open to more people. It doesn't have to be for special people only. I want it to be a world where people who love games can properly give shape to the feeling of 'I want to try making one too.' That is what I think right now.

And AI is significantly widening that gateway. Of course, using AI doesn't mean everything will easily become a masterpiece. However, it has become much easier to take that first step. There are times when you can overcome obstacles that would have stopped you in the past by working with AI. Things you thought you couldn't create alone are gradually taking shape. That experience is quite fascinating.

In #5, I will reveal the process of how I actually use AI to turn a game into a 'working form.' I'll show you what kind of language I use to ask the AI, how I run the code that comes back, and how I make adjustments from there. I will trace the production flow exactly as it happened, while showing you the prompts I actually typed in.

Finally, the thing I wanted to convey most today: game development is more about the order of operations than talent. First, decide on the 'core of the gameplay.' Next, get it running in its smallest form. Finally, add the finishing touches to make it feel good. Thinking in this order significantly lowers the hurdle of game development. AI is a tool that helps you take that first step.

If you have ever thought, 'I want to try making one someday,' please start by creating something small that works.

#5 is scheduled to be released tonight.

See you in #5!

Thank you so much for reading until the end!!!!