This Problem Changes Your Perspective On Game Dev

A new approach for thinking about game development and my very best advice for how to make a successful indie game.

0:00 - Design Is a Search Algorithm
1:33 - PROBLEM 1: Speed vs. Accuracy Tradeoff
3:26 - PROBLEM 2: Local Minimum
5:24 - PROBLEM 3: Infinite Search Space
7:52 - PROBLEM 4: Wrong Reward Function
15:58 - PROBLEM 5: Noisy Measurements
16:31 - PROBLEM 6: Exploration Costs
20:19 - PROBLEM 7: Multiple Captains
23:00 - Red Flags to Look Out For
25:27 - 3 Main Takeaways

My game Thronefall:

My last game "Will You Snail":

For the game developers among you:

Hope you enjoy. :)
#gamedev #indiedev

[snorlaxhd]

Instructions unclear: made a search algorithm while at sea

[GMTK]

This is one of the absolute best videos on game dev, Jonas. Completely true, packed with actionable ideas, and backed up with your own experience. Great stuff.

[katanalevygames]

That "prototype art and gameplay separately" is the single best piece of game dev advice I've ever heard! I always get stuck with half completed projects because one feels like it's put limits on the other. Thank you!

[JonasTyroller]

I have officially taught you everything I know.

[Bananenbauer]

This video should be played in every game design course. Period.

[lew.bow.studios]

This should be a GDC talk!

It was a remarkably interesting video to watch and I feel like so many things clicked for me as I was watching it.

Thank you, Jonas!

[theothetorch]

I like you, because you are a thinker.
Self reflection, analysis and then decision making are great skills of yours and it shows.
I love the clarity in your essays.

[Chamassa]

This applies very much to all creation. The good thing about building any kind of software is that the cost of experimentation is really low. Thanks for your insights!

[stefdevs]

Congratulations, this is now the most useful game design talk on YouTube.

[yayos]

The only bad thing of this video is that you didn't make it 2 years ago when I started my own indie game, stumbling on the way on all of these issues. It's the single best, most informative, complete and clear tool set on how to desing and make games. I will apply all of these for my next game. Thank you very much for making it!!!

[Kaiwala]

In a weird way this video was incredibly motivating for me. I'm at a point in my game dev journey where I've spent the last 7 years developing random odd projects and maps for Roblox. My last creation, which I finished a "final update" of in the summer of last year, is my current magnum opus. And whilst that map, called Archipelago is my best ever creation and it took 3 years of iteration and slow improvements (even a basic form of that optimised search algorithm you show throughout this video) all I have to show for those 3 years is a handful of niche players saying it's the best map in the game.

I wasn't paid for any of it. I have barely anybody who knows about it to show for it. The youtube video I made talking about it barely scraped by 1.7k views. And, it feels like that's not something I could throw onto a portfolio and wow anybody with.

All my game design and developer knowledge is so entrenched in Roblox, a platform that feels like a dying corporate hellscape, that I'm too afraid to jump across the trench and try and start from scratch learning how to make a real game in engines like godot because of the sunk cost fallacy of having 7 years of Roblox game dev knowledge.

But honestly, I've found myself coming up with 2 pretty cool ideas for a videogame that I would like to make. And they are ideas that just would not be feasible in the Roblox game engine. So I have a tough choice to make between 3 options in front of me-

1. Finally bite the bullet and start learning a new game engine. It's going to be tough, I have no experience with coding, and I know I will get frustrated many times and it will take a long time before I ever come to creating the really cool ideas I have.

2. Resign myself to the grip of my sunk cost fallacy and try emulate the games I wanna make in Roblox. This is far easier to do, but I don't think I will even be happy or motivated with the results.

3. Let another set of cool ideas I have die and take the easiest option which is to not act upon any of this.

I'm 22 years old, and part of the rising anxiety that comes with adulthood is feeling like all the free time and experimentation I had is gone. Like now I just gotta knuckle down, continue doing a real job, and have no time or energy to work on my passion projects. But I realise that that's all bull, I do still have time the only friction I feel is my own willingness to try.

This video has described many of the problems I bumped into over the years working as a team or even as a solo in my time on Roblox. I realise now that the spent time on Roblox isn't wasted, that those years of failure can be reframed to be my (very inefficient) scouting boats. I've done many of the stuff you said here before, the 2 captains problem such as me and another project lead butting heads, developers spending more time discussing the game rather than prototyping it, and spending far too much time on a prototype and making it look fancy before realising that is all wasted work.

I've learned these lessons already. But now this video has helped reframe all that into just another step of the process. Sure, my childhood is over. But my adulthood has just begun. And I got 80 more years of time left.

I think it's time I start learning how to make a game. Properly, this time.

Thank you Jonas.

[rubberface]

Brilliant! I would say that this also applies to other creative processes, like filmmaking and writing. We should always take a moment to reflect on the process and see what can be improved and what worked. I'll definitely come back to this when troubleshooting in the future. Thank you!

[iceluke]

You’ve just summarized everything I’ve learned in 2 years of my game design bachelor and more in 26 min.

Keep up your insanely good work!

[altafcreator]

Finally, a new upload! I'm still on the second minute at this point, but I already wanted to thank you so much. You were the person who ignited my game dev journey. Before I watched your videos a few years back, I already tried Unity but gave up. But then at one random night I found your Unity tutorial video, and the next day I decided to watch it, and now it's my new hobby. I've always liked programming since I was a kid, but now my favourite thing to do (in the programming space) is making games, and it was because of you. So once again, thank you. Thank you so much for making these amazing videos, Jonas.

[mathijsfrank]

Very insightful video. Honestly that tip on switching prototypes when 2 people are both happy with theirs is absolutely genius. Never heard of that before, but it just makes so much sense.

[Nikant]

Honestly, just thank you:
- You and your channel (Funny/Educational videos + some parts of your personality I guess?)
- Wowie game jams (One of the stepping stones on my gamedev journey)
- Discord community (I'm not a native English speaker, yet was reviewing games in room with a lot of people)
- WYS arg (I've learned CUDA programming and other stuff, met new people and still having a great time)

I've learned a few things just from a gamedev process presented in a funny/interesting way (devlogs). Now you're giving us compacted knowledge in an easy to understand metaphore. I don't even know what else to ask for.
Keep up the good work :D

[uheartbeast]

This is put so clearly, it's insane. Thanks Jonas.

[dwadthechad]

I've been waiting for a new video and I'm already so excited 10 seconds in. Honestly you rekindle my love for game dev and I can't thank you enough.

[NeverSink]

This is one of the best game dev videos I've seen. Explores the problem space very well, provided great examples and solutions. We're working on our first steam game right now and I found it super helpful, thank you!

[brunokaiser]

Absolutely amazing video. This has given me a completely different perspective on game dev. I'm suprised this topic (the search for the game) isn't more talked about and explored. Loved the addition of your own formulas.