Creating a Game in a Week Using AI tools. Day One

In the next 7 articles, I will create a basic prototype of a game with offers, gameplay, and everything I've been talking about in this blog for the past year. In the end, you will have a complete understanding of where to start your path in game development.

Facts right away:

• Budget: $0
• Development on my own and without programming
• All graphics using AI tools and stocks

In this part, I will make a plan, and choose the genre, and format of the game. The rest of the parts will soon be in my Telegram channel.

Game format

For indie developers, there are essentially two options: casual and hyper-casual formats:

In short:

• Hyper-casual game — the simplest type of game that includes basic gameplay and monetization through a bunch of annoying ads.
• Casual game — still a simple type of game aimed at a wide audience but with deeper gameplay (e.g., like in match-3) and monetization through in-game purchases.
• Hyper-casual games really irritate me. As far as I'm concerned, they are the bottom of game development, the essence of which is to show as many ads as possible until the player deletes this crap. 

Therefore, I choose the casual format, even though it's tens of times more complicated.

What should be in the game:

• Main mechanics
• Meta-mechanics
• Shop with basic offers
• Interest curve — difficulty levels will be automatically adjusted for each player, thereby rocking them emotionally.

What is needed for this:

Graphics and sounds

The main part will be generated using AI tools, and some will be taken from image stocks:

Stocks:

• Freepik
• Shutterstock
• Envato Elements — you can also find all the sounds for the game here.

Stocks are needed primarily to assemble the game menu and all UI elements; the rest will be done by AI tools. 

All services have free subscriptions, which should be enough.

AI tools:

Here it would be logical to mention Stable Diffusion, and that would be enough, but it requires a powerful graphics card that most people don't have. Therefore, to make the experiment more objective for everyone, I will use analogs:

• Bing images — an AI tool from the Bing search engine that generates quality images to suit any taste right in the browser.
• Photoshop beta — that very Photoshop with an AI. I will use it primarily to correct images from Bing because Photoshop generates mediocre images from scratch.

Game code

The game will be on Construct 3 — probably the simplest engine in the world, with which you can create a game in just half an hour of clicking. And if you encounter difficulties, you can turn to Jarvis.

Let's try the advice in practice:

And it works.

Of course, it won't write the entire project for you, but it will help you find out local points, and that's awesome!

Main gameplay mechanics

First, let me briefly explain what mechanics are from the point of view of body chemistry:

1. The player doesn't care what the game is about, whether it's flying on dragons or match-3 puzzles. All that matters to him is getting bursts of dopamine, and the more of them per unit of time, the better. 
2. To make the bursts significant and the game not boring, you need to take away dopamine. It's like with food; you eat while you're hungry, but if you're full, you don't care about food. To prevent the player from getting bored with the game, you need to take away dopamine and leave them hungry for as long as possible.

In short, to make the mechanics interesting, you need to alternate between two micro-cycles:

• Negative: Any micro-loss — loss of dopamine.
• Positive: Any micro-win — dopamine reward.

Negative cycle:

Positive cycle:

The result of these micro-cycles will be the victory at the level, which will give the maximum dopamine, or the defeat, which will take away the maximum. 

In essence, any game works according to this principle:

• Anticipation;
• Action;
• Receiving or losing.

But the phenomenon of match-3 games is that the density of these cycles is one of the highest in the entire gaming industry, which is why millions of people get hooked on these games.

In short, it's like in toxic relationships, when they mess with your head with both good and bad behavior, you get hooked on it like a drug addict.

I will choose a ready-made mechanics in which these principles are already considered.

Choosing the main mechanics

Let me just say that the main gameplay will be as simple and basic as possible to fit within the confines of 7 days of pure time.

The puzzle category is the most accessible 2D game option that can be created even alone, and on top of that, it is the most profitable genre in all of game development.

The category includes about 20 sub-genres:

Ideally, combining two puzzle mechanics into one and creating something unique, but it will take at least a couple of weeks, so I will choose the Blast mechanic, which is self-sufficient in itself:

The Blast mechanic looks like this:

The principle is almost the same as in match-3, but:

• You need to combine 2 items, not three;
• Items burst on click, not swipe;
• The competition is 3.5 times less than in match-3;
• The top games in this genre are old games that thrive due to their age, but have quality an order of magnitude lower than the same match-3, which is a huge plus because there is potential to make it cooler;
• The top 10 games in this genre earn over a million dollars a month;
• More than 10 million players come to the genre every month, and the more traffic in the genre, the more people Google can potentially recommend your game to.

In short, it's perfect.

By the way, one of the misconceptions in game development is that organics are only keys and category searches. In reality, 90% of the traffic is recommendations to people playing similar games, so there's no need to be afraid of competitive niches.

Final plan

• The game will be in portrait orientation;
• Genre: Casual Blast mechanics;
• Meta-gameplay — I'll add construction or restoration of buildings for earned stars;
• Basic shop;
• And most importantly — balance and the interest curve.
• Articles on progress will be released 2-3 times a week, one day for the game and one day for formatting the article.

Contact me in my Telegram if you have any game development questions, I'll answer some in private and reveal some in the next articles.