How to Actually Get Good at Game Dev (The 3-Phase System That Works)

Most people trying to learn game dev are doing it wrong. Not because they're lazy. Not because they're dumb. Because nobody gave them a real system to follow.

I built one. I use it myself and I teach it inside Tenth Legion Games. It's based on how John Carmack and John Romero became two of the greatest game developers who ever lived. Three phases. No fluff. Real games, real skill, real progress.

Let's get into it.

Why Most Advice Fails You

When I started my game dev journey I did what everyone does. Bought the courses. Followed the YouTube tutorials. Listened to the experts.

Nothing stuck.

So instead of quitting, I asked myself a different question. Not "what am I doing wrong" but "what system am I missing?" I went back to how I actually learned software engineering and became a professional. Projects. Repetition. Deadlines. I rebuilt my entire approach around that and this roadmap is what came out of it.

Phase 1: Beginner (Month 1 Through Month 12)

The First Month: Learn the Process, Not the Code

Your first four weeks have one goal. Learn how game development actually works from start to finish.

Follow four tutorials. Build four games. Publish all four. One per week.

This is not about learning to code. It's not about art or level design or 3D modeling. It's about understanding the process. How do you start a project? How do you fill it with content? How do you get it out the door and into someone's hands? That foundation is what almost everyone skips and it's the reason so many people spin their wheels for months without making anything real.

Months 2 and 3: Your First Original Projects

Now you build your own games. One month each.

The main thing you're learning here isn't coding or art. It's scope. Scope is the number one thing that kills game dev dreams and most people don't even realize it.

I ran into this myself. I started a red panda game called Eden Bound early in my journey. I followed all the advice. Color theory, story, assets. Spent a ton of time on it. In the end it wasn't fun and the scope was way beyond what I could handle alone. I shelved it.

That experience taught me something important. You need to know what you're actually capable of before you can build something worth finishing. That's what these two months are for. Set a deadline. Stick to it. Learn where your limits are so you can push them.

Months 4 Through 6: A 3-Month Project

Time to expand. Multiple levels, more complex enemies, actual menu systems, more polish. You're not trying to become an expert at any one thing. You're learning how to manage a bigger project, how to stay focused longer, and how to ship something with more depth than you've built before.

Months 7 Through 12: A 6-Month Project

This is the biggest test of phase one. A six month project will push your limits further than anything you've done so far. Scope, discipline, burnout resistance. Can you stay focused on one thing for six months without jumping to something new and shiny? Can you keep adding features without losing control of the project?

You're going to learn a lot about yourself in this stretch. That's the point.

By the end of year one you'll have built 8 games from scratch and published every single one of them. Most game devs never get there. You will.

This roadmap is directly inspired by John Romero in the early days. Back in the 80s he was shipping game after game, sending them to magazines, getting them in front of players. He didn't care if they were groundbreaking. He cared about building skill by building games. That philosophy is the foundation of everything here.

Phase 2: Intermediate (Stack Reps)

After year one you have the foundation. Now you do one thing: stack reps.

Build games over and over again. Never go longer than three months on a single project. Four solid games a year. Brand new project each time, no going back to add features to old ones.

This sounds simple. It is. It's also something almost nobody does consistently and it's exactly why so many developers plateau.

Back in the early 90s Romero and Carmack were working at a company called Softdisk. They pushed out 13 games in a year. Thirteen. Nobody does that today. And yet somehow we're surprised that games take longer to make now even though the tools are infinitely better. We have free engines, free assets, tutorials for everything. They built their engines from scratch. Something got lost along the way and I think it's this: not enough developers put in the reps to actually get good.

So this phase is simple. Make games. Publish them. Repeat.

If you have extra time, start reading. Game dev books, game theory, engine documentation. Reading shouldn't replace making games but it will sharpen how you think about them.

And if your goal is to release a commercial game on Steam, you don't have to wait until the end of any particular phase. The more skills you build before you take that on the better your chances of success. But it's your call.

Phase 3: Good (Learn Deeply)

I was going to call this the expert phase. But expert feels like an endpoint and this is anything but that.

This is where you start going deep. Not just making games but truly understanding how everything works underneath. How engines are built. How renderers work. Low level concepts. Game theory at its core. What makes a game fun at a fundamental level, not just by feel but by understanding.

This is what Carmack and Romero did. They didn't just use engines. They built them. They understood the processor, memory allocation, every low level detail of their tools. That's why they could do things nobody else could do. That's why their engines were revolutionary.

You don't need to build your own engine to ship great games. But if you want to reach the highest level of skill you're capable of, this is the path. Learn deeply. Understand your tools from the inside out.

One More Thing: Code Matters More Than People Admit

After reading a lot about Carmack and Romero I keep coming back to one thing. They were master programmers first. Not master artists. Not master storytellers. Master coders.

Here's a simple truth: you can make a game if all you know is how to code. You cannot make a game if all you know is how to draw or write a story. Code is the foundation that everything else sits on.

That doesn't mean you need to be a master programmer to make good games. But you should learn enough to understand what's happening in your project. If you're an artist who can read code and follow what it's doing you're going to be a far more valuable collaborator than one who can't. If you're a solo dev even a basic understanding of programming will open doors that stay closed otherwise.

I'm personally going all in on coding. C# for Unity and learning C++ to understand how engines work at a deeper level. Not because I want to build my own engine to ship games with but because I want to understand the craft as deeply as I can.

Whatever stage you're at, start there.

The Short Version

Phase 1 (Beginner): Follow the 12-month roadmap. 8 games. Learn the process, learn scope, build discipline.

Phase 2 (Intermediate): Stack reps. Game after game after game. Never longer than 3 months each. Publish everything.

Phase 3 (Good): Go deep. Understand your tools at the lowest level you can. Study the craft seriously.

That's it. That's the whole system.

If you want to work through this with a community behind you and direct access to me, come join Tenth Legion Games. We have weekly group calls, Q&A, and a group of developers all building and pushing each other forward.

Join Tenth Legion Games on Skool

Want to hear me walk through this in full? Watch the original video on my YouTube channel.