Hi All,

How viable is it in the current state of the market to develop a triple-A capable game engine for story driven action RPG type of games? Would a senior programming team of 5 people working on it be sufficient to make the core and toolsets, capable of starting game development after 1,5 years while working on further development and refinement of the engine during a 3,5 years game development cycle?

Reason I’m asking is that I’m currently in the middle of starting up a cartoon animation studio using UA4 as a render engine but want to expand into gaming (my dream for the last 25 years) closer to the end of the year as well. Team size of the cartoon team will be ramping up to 15 before year end and we are currently busy with purchasing a professional grade inhouse mocap facility which can also be used for our gaming team from the start.

As I said, the cartoon animation team will work with UE4 from the start as a render engine but the aim would be to switch to our own engine when capable. Staff working in the 3D cartoon animation team (artists, storywriter, animators, etc) can support the game programming team in the initial development stages where required before ramping up with dedicated game development staff later in the development cycle.

I’ve done quite a bit of research on game engine matters but don’t know anyone in my network at the moment with senior game engine knowledge. Advice generally seems to be geared towards using UA4 or Unity instead of developing our own engine from scratch but I’m curious if anyone here has enough experience to give some indication of man years required to build such an engine from scratch if we do want to do this. 

Thanks for your time in advance!
 

19 minutes ago, Arjanvp said:

How viable is it in the current state of the market to develop a triple-A capable game engine for story driven action RPG type of games? Would a senior programming team of 5 people working on it be sufficient to make the core and toolsets, capable of starting game development after 1,5 years while working on further development and refinement of the engine during a 3,5 years game development cycle?

Anything is possible, given enough time and money. Likelihood is an entirely different thing from what is possible. 

In other words, "sure." With the right superstar experienced 5 people, yes - you can have an engine after 1.5 years. The trick is going to be to get the right 5 people.

5 programmers may be able to make a game engine in 1.5 years, depending on what you want. Just loading and rendering the world and assets can be done pretty fast, the hard part is all the postprocessing (lighting, shadows, etc.), you need very skilled people who have a deep understanding of that. I started from scratch as a beginner and have made a game (with procedural generation) in about 5 years almost alone, so I guess 1.5 years for 5 people is indeed not unrealistic.

Not sure how long it takes to make a level editor, but you could use pre-existing tools to build levels.

You should consider if it's worth investing about $1,000,000 into making an engine. If you expect to make $20m (5% of that is the UE4 fee - $1m) with the games you make with that engine, it is worth it.

40 minutes ago, Magogan said:

 I started from scratch as a beginner and have made a game (with procedural generation) in about 5 years almost alone

What do you mean by procedural generation. 

And from scratch you mean all things like  low level opengl api loading,own OS based event handling, physics too and all that stuff. 

I want to know this because,  i am also trying to make one but not as i mentioned above. I am using glfw,glew,bullet,assimp. So may I call it too from scratch. 

59 minutes ago, Arjanvp said:

Would a senior programming team of 5 people working on it be sufficient to make the core and toolsets, capable of starting game development after 1,5 years while working on further development and refinement of the engine during a 3,5 years game development cycle?

It took me 3 years to grasp how things goes under the hood of a game and it will take me five more to get my one running out in the market(sure it will happen - this is how I give myself confidence). 

mentioning I am student right now and i am learning . As you say the 5 will be experienced at their job,  so from my perspective it's gonna be ok?. 

Note: sorry if I just jumped in the conversation which is out of my league. 

3 minutes ago, Uttam Kushwah said:

What do you mean by procedural generation. 

A block world like in Minecraft, but better.

In my case, "from scratch" means I use the Direct3D 11 API and other Windows APIs. For sound I use SFML, for file loading and saving several libraries (assimp, libpng, fbx sdk, ...), and OpenSSL for crypto stuff. Apart from that I barely use any libraries. I also wrote a server for the game.

I started while studying, too, but I've finished my Bachelor's degree almost 2 years ago.

Glad to here your work story. May i have link to your game portfolio. 

I just have this one game:

The project is not up-to-date though and the trailer is pretty old.

We should get back to topic though.

Knowing more about the game the OP wants to make would help estimating how long it would take to make such a game. And what is the budget?

Specifically to answer your question regarding AAA quality rather than Indie games, which is a totally different ballgame..  (this may not be what you want to hear if you are posting this in beginners):

• If you are only moving into the games business from animation I personally would strongly recommend you stay with an established engine for at least 5-10 years before considering the investment in making your own. First step should be to hire a tech lead with boatloads of experience who will be able to help you with these kinds of decisions, as you learn the games business.  
• It isn't immediately clear from your post what your reasons are for wanting an in-house engine, you need to justify this in terms of the cost etc. You haven't stated for which platforms, whether it will support multiplayer etc.

Caveat I've been out of the AAA business for a while now, so others will give you better idea as things have only got more complex, and writing an in-house engine has become less viable so these are very conservative estimates, but to give some kind of rough guess:

If an existing company decides for good reasons to invest in making their own AAA multi-platform engine / toolset specific to their games (forget making a generalist engine), it could easily take 5 years+, preferably being built by a separate team than those working on the actual games (so that the games can still keep you in business in the meantime), and you'd need to be able to attract and pay for the right staff (which is not a given). As a rule of thumb, 10 top guys will produce better results than 100 mediocre guys, with different guys specializing in different areas, and you'd probably need a few others working on toolchain stuff as well.

1 hour ago, Arjanvp said:

How viable is it in the current state of the market to develop a triple-A capable game engine for story driven action RPG type of games? Would a senior programming team of 5 people working on it be sufficient to make the core and toolsets, capable of starting game development after 1,5 years while working on further development and refinement of the engine during a 3,5 years game development cycle?

Most AAA titles are built using Unity or Unreal these days anyway. Most commercial studios understand that repeating the efforts of others is poor use of time if you're intending on a viable commercial project.

But let's look at the credits for a couple of recent triple A titles: Breath of the Wild lists 50 programmers, plus similar numbers of artists and twenty or so designers plus many, many more in testing and regionalisation. Red Dead Redemption 2 lists so many that I can't face counting them. Breath of the wild was in development for five years, RDR2 was announced two years before release, which means it'd probably been in development for at least year before that. In both cases, the companies were likely using a fair amount of existing in-house code and tools as a base for the their development. Every year, the teams and budgets on triple A games get bigger.

The people making these games aren't fools, and their staff are not inept. It's simply extremely unlikely that a small team can produce something similar in any reasonable timeframe and, honestly, I think it's a mistake to even try. Making massive, detailed, AAA games is what the big studios excel at; smaller teams are better to play to their strengths rather than facing off on the same ground and you only have to look at the triumph of games like Slay the Spire, FTL, Hollow Knight, and the like to see what can be achieved.

40 minutes ago, Irusan, son of Arusan said:

The people making these games aren't fools, and their staff are not inept. It's simply extremely unlikely that a small team can produce something similar in any reasonable timeframe and, honestly, I think it's a mistake to even try. Making massive, detailed, AAA games is what the big studios excel at; smaller teams are better to play to their strengths rather than facing off on the same ground and you only have to look at the triumph of games like Slay the Spire, FTL, Hollow Knight, and the like to see what can be achieved.

I have to disagree with one thing - and it is common among game devs (especially here on gamedev.net), that there aren't just AAA games and low budget pixel art games. You have plenty of games that are not AAA, albeit having huge technical quality and often even custom game engines. Such examples being: Risen, Witcher 3 or Kingdom Come: Deliverance (which used CryEngine). Some having larger teams (like Witcher), some smaller (KC:D or Risen - which had tiny team compared to AAA games).

Their quality is high, their development takes also few years, yet these aren't AAA titles.

Making a high quality engine is not that big problem, given enough time and resources - it all boils down to what your target product needs. Far worse problem which I'm facing is making actual tools for the engine (this process is actually more time consuming than making an engine itself). Side fact though - I'm making toolset for my own engine in my free time (unlike engine core and its actual applications, which were made for actual, non game, products).