Bestl 2D game engine for starters? (that isn't Unity or Godot) [clearer question than my last post]

Catomax26 said:
I have seen people saying me “don't use game engines”, and that is very prejudicial because i am not aiming for a small game, i am aiming for a huge project.

Yes, using a game-engine for a large 2d project is definately the way to go. Especially if you don't yet know programming, but even if you did - the tooling you get will make producing a content-heavy project so much simpler.

However, there is not that much choice left. I'd personally recommend Unity regardless - the “asking for money” part, they did go back on, And it wouldn't even have mattered for a project like yours, even if it did go through. Unity is far from perfect, I personally even dislike it based on problems and some backwards-ass workflows. But it's the best engine for beginners, and would have all the features that you need.

Unreal is a choice, but it'll require eigther working entirely in Blueprint (visual scripting), or learning their version of C++. That is not even much lower level than C# that Unity has, but Unreal has some oddities in how they use C++. Blueprints can actually be a good tool to code game-logic, to an extend, but it's also very horribly slow in terms of performance. I have no experience with Paper2D in Unreal, but do not that people will say “its crap” about pretty much anything. I'm saying that Unity is crap, but that shouldn't deter you from trying it.

Other than that, there's not much left. You have things like Rpg-Maker, which is designed for very specifiy kinds of 2D games, and pretty limited (doing anything that doesn't fit perfectly causes headaches non-stop; and their ability to customize tooling is non-existant). At least in the XP-era it was, Rpg-Maker MV seems to be a bit more flexible, but I still wouldn't recommend it.
You have GameMaker, which is also more specialized towards 2D-games from what I've seen; but I don't have any experience with it.

That's pretty much all the more common game-engines out there. Maybe someone else knows something more arcane, but in reality, I'd just try Unity, Unreal and maybe GameMaker for a week or so to see if you like it. You can make the type of game you want to make in pretty much eigther of those, guaranteed. It's just a matter of different workflows, and some things being simpler than others/different pain-points.

Catomax26 said:
I have seen people saying me “don't use game engines”, and that is very prejudicial because i am not aiming for a small game, i am aiming for a huge project.

Was that me? I only said using no engine is an option you might consider, not that you should do it.

However - what do you mean with ‘huge’ project?
As a beginner you should start small. Very small. Or you will almost certainly fail.
But this does not mean your game should have only 3 levels. I talk primarily about the complexity of the program, not about content.

So let's list some games illustrating increasing complexity…

Pong: Really simple, but you already need to know and use some math, like collision detection of simple shapes, controlling constant velocity motion of objects, and reflecting their direction.
(Game engines may make things like collision detection easier, but actually figuring out how to use their tools is just as hard, i'm sure.)

PacMan: Still simple constant velocity motion and collisions, but there is some AI and behavior.
(Again, game engine does not give a noticeable benefit to reduce complexity yet. Imo.)

Super Mario: That's already a rich and complex game. Motion involves acceleration and there is physics simulation. Interactive objects. AI is simple but varied across enemy types.
Starting with this as a beginner, chances you'll succeed are small. I recommend to do some simpler games as listed above first for learning purposes.
(Engine can help a lot to manage complexity i guess.)

Modern 3D action games: Still pretty much the same, but math, physics, and character animation becomes very hard. Mentioned only to complete the list. Using a game engine becomes a requirement to most people. Also, doing it all as a single person is very ambitious.

Assuming you want something like 2D Metroid with pixel art, i think that's a realisitc goal, but only if you have talent and invest maybe two years but likely more, even if you do use an engine like Unity.

If you distrust Godot (why btw?) then a nice 3D engine is Defold (https://defold.com/).​ I used it for two smaller projects​ but I'm no expert in it, though. It uses Lua as scripting language, which is used in a lot of games for modding etc.

Catomax26 said:
making a 2D game with almost realistic lighting

What do you mean with ‘almost’ ?

I personally like Gdevelop. It has some lighting features, and shaders too. But depending on what you think with almost, it might or not be a good fit for you.

MonoGame is pretty neat. It's a kind of modern open source version of Microsoft's XNA framework, which is what Stardew Valley was written in

Was that me? I only said using no engine is an option you might consider, not that you should do it.

Yeah, it's definitely you! :D

Just thinking about it kind of resonated in my head, but it felt wrong, considering I'd have to learn 9000 different commands and spend like 20+ years in making a custom game engine for a project like mine. I think it only works for small-scale projects or hobby/just for fun projects. I am an anti-sucker for complex programming; i don't think it's necessary to make 5 different courses for C++, Java or Python programming, as they teach you the basics of it, but not the full extent indie developers are searching for. Sure, they may make C++ or Java game programming courses, but those are meant for XL and XXL (btw these are meant to be in bold writing but meh) scale projects, such as open-world games or games just too complex.

As a beginner you should start small. Very small. Or you will almost certainly fail.

Well, i have LOTS of games planned, but i can't even reach the idea of making a small side-project. My brother has told me countless time to do it, but i still don't manage to even start making something crummy and crappy such as a simple platformer for practice. For some reason my mind is only single-targeted and it's telling me “OoOoO Cato you have to start with your big project first oOoOo”. I have plans of making 3D and open world games, but these seem to be temporarily shelved, as i want to finish my 2D projects first, That way, i can somewhat milk out UE5 for what its original purpose is.

Assuming you want something like 2D Metroid with pixel art, i think that's a realisitc goal, but only if you have talent and invest maybe two years but likely more, even if you do use an engine like Unity.

I do know my game will take around 5+ years to make. I am not aiming and messing with time limits either, by the way. I'm just seeing if i can pinpoint a specific game engine which is aimed for Metroidvanias alike, and doesn't have a problem with complex key combos or is too 3D aimed.

grbrg said:
If you distrust Godot (why btw?) then a nice 3D engine is Defold (https://defold.com/).​ I used it for two smaller projects​ but I'm no expert in it, though. It uses Lua as scripting language, which is used in a lot of games for modding etc.

Sorry, i am only interested in 2D game engines. Sure, i may use this, but i am definitely not aiming for small 3D projects.

I just don't like Godot (for reasons unknown), and because EVERYONE is saying “Dude if you want to be a gamedev you HAVE to use Godot because uhhh Unity is dead and everyone is using Unity and uhh Godot is a replacement for Unity.

_Silence_ said:
What do you mean with ‘almost’ ? I personally like Gdevelop. It has some lighting features, and shaders too. But depending on what you think with almost, it might or not be a good fit for you.

I mean it would be having some sort of realistic property to it (such as almost-realistic shadows, sunrays, etc.)

Does it work for things more than small, hobby-like games, or actually long games? (Sorry if i sound like a broken vinyl player but my game has just too many modules and stuff to work with)

@shortround From what i have experienced, XNA-based games are very hard to mod, as you have to decompile it all to even start. It also seems to require 3rd-party programs to decompile/mod it together.

Does it really work for platformers? >w<

Catomax26 said:

@shortround From what i have experienced, XNA-based games are very hard to mod, as you have to decompile it all to even start. It also seems to require 3rd-party programs to decompile/mod it together.

Does it really work for platformers? >w<

You would have to decompile any C or C++ or java based engine in order to hack it. If you want your users to be able to mod a game, you can always provide a modding API to them. If you want them to be able to re-write core parts of the engine, you can just release the source code on github

@shortround If it means making a <> by scratch, then that means a solid no. It says it's meant for programming veterans, so please no. As crazy as it may seem, i feel some sort of repulsion with learning programming, because they only teach you the system work and engineering part of it, not the passionate and project-based part of it. Nowadays, it's so HARD to find veterans who like programming because it lets you do more than income and working, less actual game devs who LIKE learning full programming at least.

@Catomax26 game maker