Picking a type of game for my first portfolio project

[fading night]

Hey Everyone,

I’m in the process of switching from web development to computer graphics and have wrapped up Learn OpenGL and a few intro series for linear algebra. This puts me in a spot where I think I know enough of the basics to pivot away from tutorial-style content and start working on my first real portfolio project. I plan on doing two before applying to graphics jobs: one will be a game (with simple gameplay mechanics so I can focus on the realtime rasterizer aspect) and the other will be an offline pathtracer.

I was going to start on the game first, but I’m having trouble thinking about what kind of game to make. I know that the gameplay should be simple since the focus should be on the graphics code, but besides that I’m looking for some insights on what type of game would force me to solve the types of problems and implement the types of features that potential employers/interviewers highly value.

For example, would something like an evolution of the classic snake game (make it 3D with some fancy weather, terrain, and new dynamic gameplay mechanics like obstacles that fall into the playing field so there’s a bit more than just avoiding running into your body) be too simple? Do I want to chose something that will have me rendering enough stuff to stress-test my frame rate/gpu?

Just trying to brainstorm these questions before I kick off the implementation planning and development. Any insights and suggestions would be greatly appreciated!

[cursive jackal]

A game might not be a suitable portfolio piece if you want to get into graphics.

I suggest you work on something that showcases that you have a good understanding of the GPU, API and modern graphics techniques. Perhaps start with a Deferred PBR renderer and then you can build things on top afterwards.

[calm arrow]

If he's coming from web dev he probably needs a lot of practice in general resource-conscious and native programming though

On the contrary I would pick something ambitious, not simple

Make something you find wicked cool

Otherwise it's going to be hard to stay motivated enough to make progress

Rather than repeat the proselytizing that I am well known for around here, read the following hour of this thread starting here: #general message

[fading night]

Well making a game would require making a renderer, so both of your suggestions aren't necessarily mutually exclusive, right? The deferred pbr renderer could be used as the graphics engine for the game?

I read through that general thread, too, and I'm also pretty interested in snow (and weather / particle stuff generally)--One idea I had was to do a driving game but keep it pretty simple with 3 versions of the same downhill "track" on a mountain. One could be a snowy version with a snowmobile, the other could be rocky with rain and puddles (mountain bike maybe), and the third could maybe be a sand dune with a sandstorm (atv as the vehicle). Gameplay would be barebones -- maybe just a timer and obstacles along the way to not crash into. Goal could be to get to the end as fast as possible without crashing. Then just focus everything else of the graphics engine and what not--really the "game" is just there to give the renderer something to work with.

I'm pretty committed to executing on this, but I've got a set timeline I'm working within so I don't want to pick a project so enormous that I can't feasibly get something together that I can put on my resume for applications/interviews. I've roughly allocated the next ~16 months to getting two solid portfolio projects in, so about 8 months per project (real time renderer / game first then offline path tracer) working ~10 hours each week so I don't burn out alongside my full-time job. Given those (self-imposed) time constraints, does the plan sound feasible? Or am I totally assessing this wrong?

[calm arrow]

I didn't really use any techniques beyond the first 3 chapters of LearnOpenGL to build that flight sim

Certainly not deferred rendering

My suggestion is to pick a project that requires you to improvise

Copying someone's deferred/PBR implementation isn't a bad use of time it just won't turn you into a problem solving powerhouse imo

Those hours are a little restrictive for that though

Not the end of the world though, it might just force some scope reduction and focusing on the graphics aspects

That flight sim project was the result of about 1500 hours of work but most of that was not rendering, I'd say maybe 300 hours were spent on the rendering for the terrain and MFD GUI

So you could certainly trim the project down to fit within your requirements I think