TLDR: I'm a programming noob looking to learn game dev. Which games should I clone to get there?

Hey people on itch!

I worked in the industry for 3 years now as an environment artist. Worked on indie, double-A and triple-A projects, but now, I want to learn more.

I'm a total noob when it comes to programming, and since it's one of the most important pillars of game dev, I want to learn how this stuff works.

I came up with a crazy plan - I'll clone 10 of the most influencial games out there, in order of complexity. 

And since my dream as a kid was to make a game like Skyrim, a Skyrim clone would be the end game!

Here are the games I would clone:

DONE!   1 - Pong (gotta start easy...)

DONE!   2 - Super Mario Bros (1985)

DONE!    3 - Doom (1993)

IN PROGRESS 4 - The Elder Scrolls Arena

5 - Diablo 2

6 - The Elder Scrolls Morrowind

7 - Shadow of the Colossus

8 - Call of Duty WAW zombies mode

9 - Dark Souls

10 - Skyrim

The rules:

I must implant all the core mechanics that defined the game's identity;

The quality must be similar in all aspects (art, animation, sound, etc.);

The lenght can be shorter, eg. only one level, only a vertical slice, etc.

So, for now, I've made the pong clone. Yeee! 1/10.

Next up, Super Mario.

Do you guys have any suggestions of games that you think would be better than those, and, is anyone here interested to get updates on this? Eg. how long each clone took to make, what were the hurdles, what ressources I used to learn, etc etc.

Aaaand does anyone have tips to help along this journey? Curious to hear your thoughts!

Edit: Added TES Arena and Morrowind to the list. It would be smart indeed to make simpler versions of The Elder Scrolls first, before making something insane like Skyrim.

24-09-08 Edit: The Super Mario Clone is now ready! More info in replies.

Here is Super Capy Baras. :

Next up, Doom.!

25-01-18 Edit: The retro shooter is done! More info in replies.

Here is Dungeon Sugar :

Next up, The Elder Scrolls Arena !

Hey there!

It's been a month, and the second clone - Super Capy Baras - is now ready!

Here's a few snipets (GIFs are really compressed and poo-poo, the art looks better in game, I swear) :

First, a few takeaways I learnt, for those curious:

-No, you won't remember later.

It’s important to note and document things, especially in the blueprints, to understand what's going on when coming back to it later. By the end, I had a lot of stuff in the CapyBara pawn BP, and having to fix bugs 2-3 weeks later once everything was implemented was a bit painfull. And that's for a small, barebone game...

-Reuse your code, don't duplicate.

That was really hard for me. I couldn't find a tutorial making a full mario clone, only snippets there and there. So, I was experimenting a lot, going off multiple different youtube tutorials to create all the mechanics. In the end, a lot of stuff was duplicated, and then worked on appart. I also didn't really create reusable functions. This all became very apparent towards the end, when I had to fix the same bugs mutliple times over and over in mutliple similar blueprints (eg. the different blocks BPs, the different ennemy BPs, etc.), or accross mutliple similar custom events in the same BP. 

-To be a good level designer, you need to play A LOT of games.

I barely play platformers, and so, I saw myself struggling to come up with ideas when designing the levels. It’s difficult to understand the fun when you don’t play a lot of games in the genre. I resolved myself to watching gameplay videos, but it’s not the same as having the controller in your hands, feeling the jumps, the level of control in the air, or the tightness of the collisions.

-Unreal is not really built for unlit 2D projects. 

After searching online for countless hours, I couldn't find a way to play in Unlit mode. And without that, there's the tonemapper and default Unreal post-process (bloom and vignette) that mess up the 2D look and colors of your sprites - and not just by a small amount. People suggested using a custom post process and playing with those settings:

Turn off Bloom and Vignette, set Slope, Toe and Shoulder to 1.

Meh, it was closer for sure, but not quite the same, especially very bright and dark colors. On top of that, I later found that certain post-processes crash the whole game in HTML5 (play in browser). In the end, I realized that the tonemapper is turned off by default in HTML5 anyways, so, I was good to go in this case. However, for desktop builds, this seems like a pain in ars. Something to think about when making a full commercial 2D game in Unreal, I think.

Alright, now, onto the next clone: Doom !

Hello there ! An other update, an other game ?

Yes, the Doom clone is here!

Well, more like Quake. Kinda...

This halloween, I entered the Itch 2024 Scream Jam and challenged myself to make our 3rd game on the list - the Doom clone.

After an intensive week of sleepless work, it was finally here: 

Dungeon Sugar, a halloween wave based retro shooter !

This time around, to save time, I started from the Unreal engine fps base template. I also tried replicating the PS1 style but with UE5's added global illumination:

You can check the game if you are curious:

https://mariannes-itch-games.itch.io/dungeon-sugar

It was so much fun participating in this jam (even solo). Over 500 games were submitted!! I tried to play as many as I could, there was a lot of good stuff.

Turns out, making a game during a jam seems to be a very good way to get ''free'' eyeballs on a project. Lots of people played Dungeon Sugar - it's my most successfull game so far, by a lot!

Don't get me wrong, it's not good. But we're slowly getting there.

My takeaways from this project:

-The first release is the launch of your game.

The game launched in a bit of a rough shape (it was working alright but had some bugs and balancing issues). After all, I only had a week to make it. I thought, it's fine, I'll just fix it all in the coming days. But then people started playing it, and some people even made videos about it! It was really cool, but all the bugs and weird moments were now immortalized in time for all to see...no matter how much it's fixed later.

So yeah, make sure you release in a good shape, because as soon as it's out, that's what people will see and remember.

-Yes, players are gonna do that.

 When something isn’t 100% implemented or doesn’t quite work as intended and you tell yourself ‘’so what, no one is gonna do that.’’

Yes. Yes, they’re gonna do that.

In my game, the sword was a bit buggy. When swinging multiple times really fast, it would sometimes re-equip the gun while swinging, looking really weird and doing no damage. I knew about the bug, but thought, no one is going to spam it that fast anyways. Boy I was wrong. Some people really, really wanted to use only the sword, and because it sometimes didn’t work, they had to spam it even faster, making the bug worse.

If you know about a bug in your game, assume right away your players are going to find it (they’ll find some you don’t even know about too).

-Balancing is really hard, and a lot of work.

I’m not sure why, I assumed that using values that sounded right would just work. It didn’t.

There wasn’t that much to balance: Size, damage and speed for two ennemies, ammo, health, the amount of damage my two different weapons make and how long they take to shoot. On top of that, the difficulty needs to go up wave after wave.

Sounds simple enough. Turns out, it was a lot of work to balance all of this and I underestimated how much I would have to play, replay, tweak and replay again.

I remember hearing in a gdc talk that if you have to tweak something, tweak it by at least 20%. Lower than that, you won’t notice the difference enough and you’ll be tweaking and testing forever. That is very true and one of the best tips I could give.

-Playtest your stuff asap – and playtest LIVE (and shut up).

Pretty self-explanatory. Don't wait. Your mechanics are in with basic UI ? Start playtesting! Even if it's just 2 or 3 people, they'll find a bunch of general issues already. 

Also, live testing in Discord is the way to go.

For this game, I got a few friends to playtest before the end of the jam. Some playtested on their side and sent written feedback via discord. I’ll be honest, that wasn’t super helpful. I think they were trying to be nice and it wasn't useful feedback.

I got 4 people to playtest live on Discord and that was really, really insightful!

I could see every single little pain point, every thing that didn’t quite work. I could also see how the game felt to play in general. Are they having fun ? Are they bored ? Are they annoyed ? Are they laughing? You can only know that through a live call.

Ok, next up, we are getting serious. An Elder Scrolls Arena clone !

I can clone the soundtrack, and also the visual of my hair