Hey all, with the announcement of House Marque ditching their current game engine to use Unreal Engine , http://www.playstationlifestyle.net/2017/12/05/housemarques-next-game-will-use-unreal-engine-4/, I wanted to ask how everyone feels of the current state of commercial and custom game engines in the AAA and Indie Space. It seems like more and more studios are starting to switch to commercial engines, either Unity or UE4. Square Enix is creating Kingdom Hearts , Dragon Quest, and FF Remake in Unreal Engine, Rare is using UE4 for Sea of Thieves, Bend is using Unreal Engine for Days Gone, Insomniac Games used Unity for their smaller titles, and the list continues.

As a graduate student that is about to graduate, I just wanted to see if the concept of fully creating a game from scratch is becoming a dying breed for engine developers, and I should focus more on creating my projects using these more popular engines, especially from a portfolio perspective. Thank you.

Engines are great. The engine allows small teams power they wouldn't have and the developers of the engine is like having a dedicated team focused on only the engine.

It's much more cost effective to use a engine. You can focus on development with a engine. You don't need to learn so much with a engine.

But it isn't all rainbows and sunshine. A lot of developers who learn by starting with engines start too big because engines allow so much power.

There is also the huge redundancies that you get from these engines. They are build to be as flexible as possible and for this reason you get a lot of extra's that you don't need. Unity's 2D physics that is on by default or Unreal's graphics that is set too high and glossy are good examples.

3 hours ago, AxeGuywithanAxe said:

I just wanted to see if the concept of fully creating a game from scratch is becoming a dying breed for engine developers

It will die when games die and games will die when all intelligent life dies. There will always be a need and benefit for making games from scratch, just as there will always be a need and benefit for making games using already designed tools.

As for your portfolio think of what you want to do. If the companies you want to work for is known for using a engine then so should you, if you plan on being a developer known for being able to work on custom made engines, then you need to make games from scratch.

1 hour ago, Scouting Ninja said:

It will die when games die and games will die when all intelligent life dies. There will always be a need and benefit for making games from scratch, just as there will always be a need and benefit for making games using already designed tools.

I guess I should have said, will it realistically come to the point where Unreal Engine 4 and Unity take up so much of the market share, that the concept of having a proprietary engine will only be for those running on a decade + of tech.

1 hour ago, Scouting Ninja said:

t's much more cost effective to use a engine. You can focus on development with a engine. You don't need to learn so much with a engine.

Yeah, and I'm definitely feeling that now. I feel like I'm never making "real" progress with starting from scratch. I'm currently integrating physics into my engine, and I just constantly think about how much time could be saved from using an engine with decades of testing and 100's of titles. Part of me likes designing an engine and knowing exactly how something works, and not having to "rely" on another person and a general purpose method.

1 hour ago, Scouting Ninja said:

As for your portfolio think of what you want to do. If the companies you want to work for is known for using a engine then so should you, if you plan on being a developer known for being able to work on custom made engines, then you need to make games from scratch.

I think that's the problem I have, it seems like more and more companies are ditching their engines for UE4. From a portfolio perspective, what I can show with a proprietary engine is a lot more visually interesting, but it also relies on others tech. If I start from scratch, I might not be able to have as much to show, but I can explain exactly why I designed my renderer , game layer , reflection layer, etc the way I chose to. I'm just trying to find a proper balance.

32 minutes ago, Hodgman said:

At a AAA studio, if they're using a custom / in-house engine, it will likely be just as big and complex as Unreal... If you get a job at a place like that, you'll still be making games on top of a massive engine. Even if you get a job there as an engine developer, it will be much closer to fiddling with the UE4 code-base, than your student-home-made-engine codebase.

That makes quite a lot of sense, and I guess I didn't think of it from that aspect. 

34 minutes ago, Hodgman said:

Doing fancy tech stuff from scratch is also important, if you want to work in a very technical role... but making a custom renderer from scratch when you're applying for a gameplay role isn't the best use of your time. In that case, a gameplay system built on an existing engine would be a better portfolio piece

I think my current problem is that I'm not quite sure what my focus is. I wanted to start out as a graphics programmer, but after a series of projects and outside influences, I've sort of burned out on the graphics end (at least when it comes to shaders). I think I'm more suited for either gameplay/general/tools because I have yet to burn out on them, and I like the challenge of designing the architecture , instead of implementing mathematical theories and etc to make things "look pretty".

I've always had the notion that the best way to show my work, and to create a game on the side, was to build everything from scratch, but I've started to find a diminishing returns when doing it, I think I constantly fall into the trap of looking how vast other engines are and forgetting that they have been worked on by 100s of people over a decade or more. But I also like the feeling I get when I accomplish something, and like knowing that all the code is my own.

That seems to have been a tangent, but from your perspective when looking at an applicant, would you rather see someone who created a full game , leveraging an existing engine, or someone who might have a partial game with a custom built one?

There are nowdays still studios like Deck13 that are going to make there own engines (in the Deck13 case they called it Fledge and wrote a serious of articles in a popular games magazine) so there will always be people making there own engines especially when fees play a role and you have the man-power to do so. Some studios as Guerrilla Games working on titles like Horizon (build on there Decima Engine or Naughty Dog roll there own ones too especially when going just for a single platform only.

The benefits of using third-party engines is primarily a costing factor while I agree on that, I strictly disagree on the

5 hours ago, AxeGuywithanAxe said:

using an engine with decades of testing and 100's of titles

part because even those will have massively problems and cause huge troubleshooting and development time consumptions due to unfinished updates and features that do not work (I'm looking at you Unity 5 <.< ) so one could say it is primarily the pricing. I have worked at a lot and talked with a lot more of so called indie studios and they all calculated there budgets in a way that development of an own engine would be no point to them; instead all of them used Unity and some got to Unreal. Unity on its own has also the benefits of comming with C# that is, looking on Unreals code base, a lot more beginner friendly while those who have made a few games with Unity tend to go to Unreal because it has due to the C++ code base even more capabilities to customize what you need. It also feels like people tend to make too huge plans because those systems offer them to be easy as breathing. We have never had so much unfinished games on Steam and disappointed players/backers on Kickstarter as ever before so this is one of the downsides of this business.

On the other hand it seems like a big task to make an own game engine but after 5 years now from starting to get my first renderings with C# and a custom OpenGL.Net library and the old Nehe tutorials, several revisions and versions later and switching from C# to C++ completely 3 years up to build a huge code and tools base state of today, it was and will be a wonderfull and ongoing task doing this in my rare spare time while working with Unity and Unreal at business. The technology curve is high but you will benefit from that knowledge regardless of the position you might enter later. I have had my hands inside Unreals base code and build pipeline but also huge AI projects on Unity making that multitasking to solve logic tasks for massive AI and could always benefit from what I do after work. I have learned to live without the "convinience" stuff that we got from the STL in C++ 11/14/17 most of the time (and some people might now tend to jump to the keyboard and quote this to tell me how wrong my opinion is ) and ported some of the C# "convinience" into C++ while I mostly learned how good it feels to have the freedom to know your system regardless how huge it has grown. But at all this feeling of beeing home in my own system, I also have had situations where it was better to use one of the third-party ones Unity/Unreal in cases where mobile support was important for example.

So it seems to me less a question of making an own engine/games from scratch or not rather than could you pay for it while I think it is totally ok to use the ready ones at business but I also sit on the roll-your-own-whenever-possible train and will prefer to do things using my own code/tools base as hobby and partially in business.

(Pleaso note, this is as always my personal opinion and no! my opinion isnt wrong )

11 hours ago, Shaarigan said:

On the other hand it seems like a big task to make an own game engine but after 5 years now from starting to get my first renderings with C# and a custom OpenGL.Net library and the old Nehe tutorials, several revisions and versions later and switching from C# to C++ completely 3 years up to build a huge code and tools base state of today, it was and will be a wonderfull and ongoing task doing this in my rare spare time while working with Unity and Unreal at business. The technology curve is high but you will benefit from that knowledge regardless of the position you might enter later. I have had my hands inside Unreals base code and build pipeline but also huge AI projects on Unity making that multitasking to solve logic tasks for massive AI and could always benefit from what I do after work. I have learned to live without the "convinience" stuff that we got from the STL in C++ 11/14/17 most of the time (and some people might now tend to jump to the keyboard and quote this to tell me how wrong my opinion is ) and ported some of the C# "convinience" into C++ while I mostly learned how good it feels to have the freedom to know your system regardless how huge it has grown. But at all this feeling of beeing home in my own system, I also have had situations where it was better to use one of the third-party ones Unity/Unreal in cases where mobile support was important for example.

So it seems to me less a question of making an own engine/games from scratch or not rather than could you pay for it while I think it is totally ok to use the ready ones at business but I also sit on the roll-your-own-whenever-possible train and will prefer to do things using my own code/tools base as hobby and partially in business.

Thank you for your perspective! I think one of my main issues I'm having is forgetting just how long it takes to get the entire project done and might be rushing it to much. I like the idea of creating my code , but it seems like a lot of work, and I don't necessarily know if it outweighs the benefits. I've learned quite a lot building my own engine, but I still don't have a full finished project.

Ive made the experience that rolling your own is less coding rather than getting all the informations together you need for the systems and (in my case) for code refactoring.

I'm just at revision 6 restructuring the project for the third time as it growed a lot in the past 3 years from new knowledge and experiences made with it working 10 to 15 hours a week in avarage in parallel to my business job is a complete development time of arround 1 1/2 to 2 years in full time on a 40 hours per week base doing anything from scratch (I've started with nothing than an empty Visual Studio project and up for today use nothing except zlib, tinyEcc and the OS APIs). Sounds much but in the end it competes to development time of a game