Game Engine Usage Cost?

Started by
17 comments, last by ranakor 11 years, 2 months ago

So this was sort of a shock to me, as I am coming from a long time programming background that is not in the game space, but maybe its seen as normal to you with more experiences. Either way, I wanted to ask about this and get the thoughts of everybody, because I figure maybe one of you has some insight I lack or can otherwise provide a reason for this. And I can not learn if I do not ask, either way.

So I started t.o research different game engines and tools with an idea to work on a project, and I was amazed by the terms and price points being quoted to me. One wanted 100k for only binaries, no support and no updates (So if I found a big bug, I was screwed). Another wanted a 30% revenue share, to effectively take 30 cents of every dollar I made (That would be in gross, not net!). Most refused to not provide support (one engine cited it as damage control, as apparently they get a lot of vocal people on the internet bad mouthing them who do not even know basic C++). Some quoted me a price, then stated that was per person allowed to touch the code. Some wouldn't quote me a price unless I signed an agreement that I wouldn't tell others the price I was given (I refused, so they wouldn't tell me the price). Most sent me an auto-reply if I was lucky, and must have simply decided it wasn't worth replying to because I got no response, Most of them dont seem to even want people to know the price.

Is this the standard in Game Development circles? Are these prices so high for any specific reason, given the sheer volume of the games they could be doing? And why are the rev-share percentages so high? The landscape doesn't seem to be very Indy friendly, and I'm having trouble understanding why this is given that anything that stops games from being successful or limits what people just starting out can do also limits their own possible market share for the things they are trying to sell, so I see such rev-share and initial costs as limiting to the engine sellers as well. After all if only a few can afford to use their engine, they will make a lot less money.


As ignorant as I am sure I sound, It seems very short sighted to me, and even accounting for support, dev time, etc, it seems excessive to me as the ignorant noob who just wants to make a great game and give people something they will enjoy, Am I wrong to think this way? What am I missing? And what are the common options?

Advertisement

The landscape doesn't seem to be very Indy friendly

Yep, and this is why Unity3d has managed to carve out a huge percentage of the indie developer marketplace -- it's only a few thousand.

And before that Torque was doing ok in the indie scene for only $100.

After all if only a few can afford to use their engine, they will make a lot less money

Whenever I see businesses doing strange things, I always remind myself:

* The first rule of capitalism is, if a business continues to exist, the it's probably making money.

* The second rule is that if it could be making more money, then a competitor will likely force it out of business.

...so there's usually a method behind any kind of business madness, otherwise they'd cease to exist.

In the case of companies that sell an SDK for $1,000,000... yeah, maybe they only get 10 customers, or $10M per year. They could choose to sell it for $1,000 instead, but then they'd need to find 10,000 customers, and set up the infrastructure to deal with all of them. Their brand would also be less prestigious because the market would be flooded with 10,000 low quality indie games associated with that engine.

Half the companies I've worked for have used licensed engines. One of them in particular cost over $1M, with full source and support included.

The other half have used their own game engines developed internally, usually by a quite large, dedicated engine team, made up of expensive programmers.

In my case, I've recently started a company to make a game, so we had to face the buy vs build dilemma. As you've covered, the big, quality commercial engines are very expensive. Alternatively, you could hire a team to build a custom one for you. if we say that the average games salary is $80,000, and that one average programmer can build an adequate engine in a year, then a custom engine will cost you $80,000. In practice, you probably want multiple programmers, and you want them to be well above average, so a game engine is worth quite a bit...

In our case, I'm a professional engine programmer, and because it's my company, I'm willing to work for free... so we chose to build one simply because with those factors it became a lot cheaper... but if you don't happen to have a professional engine programmer willing to work for free, then the ridiculous prices offered by the vendors aren't always that bad.

Not sure why people would badmouth because they get support. Asking questions on a forum isn't what I would pay for anyway.

Why haven't you considered any of the free engines? OGRE can be used to build a framework, T3D is MIT now, etc.

Hodgman covered most of it. The theory is that they employ 100s of people dedicated to the engine and support. Those are, in theory, people you would have to pay to make an engine of similar quality. The cost/benefit analysis breaks down when you figure out which engine features are usable by your game, as they have to support features for all games. Also remember when you buy an engine, you have an engine NOW instead of a year from now.

It is my experience that a lot of internal engines tend to suffer either from code bloat or incomplete stuff(ie. they implement only what is needed and often only for what it's needed for without a mind to what might be needed in the future). That certainly isn't the case everywhere, but in my experience that seems more likely with internal engines. That said, external engines might not have a small part of implementation that would take a long time to wait for implementation unless you pay a lot of money for source.

Anyway, it's expensive and restrictive, but more often than not it's for good (read: logical in context) reasons.

Actually the very rise of engines like unity make it illogical in context, as those engines are part of the context now and cost between 0 to 1500 per dev, royalty free.

I am coming from a long time programming background

me as the ignorant noob who just wants to make a great game

Kind of contradictory.

Anyways, to answer your question, the prices are high because on top of the fact that they are streamlined pieces of software, they also come with many, many tools for artists and designers that become extremely helpful. For example, UE3's Kismet visual shading language. If you are a seasoned programmer like you initially claimed, than I think that you could in time make your own software (albeit much less flashy) to run a 3D game. Its mostly the content you should be worried about, however. Torque3D is now MIT'd open source software, so that is a great open source C++ engine, and you can modify it any which way you please or not at all if you so choose. If I didn't prefer C over C++, I'd have already been using Torque to make something I have been working on.

C dominates the world of linear procedural computing, which won't advance. The future lies in MASSIVE parallelism.

To explain: I'm longtime dev in other things but due to different issues I didn't have time for graphics programming before this.

To paraphrase my 3d artist other half "Looking back, Its really frustrating how some really great games that just had [cruddy] graphics did very badly, while games that sucked did well due to epic shiny graphics".

That is why we are looking into an engine; because my other half is a Lead 3d artist for an MMO company and she has her ideas on what an asset pipeline should be - shes also the most experienced of us - and so has "requested" that it be as nice as possible. I just cant afford Granny and as of this morning she now "has her heart set" on a specific engine that I can not afford so as a result, I am looking for options that will make her happy,. and the Torque and Unity3d engines all fail her requirements due to various issues.

1. Unity3d: Is actually very restrictive on what you can do with it, and 'm also reading a lot of information that suggests it has poor performance from a networking and "open world" perspective that makes (from an art perspective) crowds, instancing, impostors, shadows, and instances particles an issue. I'm also reading a lot of item on unity3d that suggest core functionality in its base objects for even simple things like getters and setters just doesn't work, due to a really bad lack of QA, and while I do not know how credible such claims I am reading are, the idea of paying almost 2 grand for something that doesn't work really scares me and a lot of people seem to run into the problem.

2. T3D: I have no personal experience with it, but my Artist took one look at a youtube video on its feature set and said "no"..

Maybe some of you with more experiences with it can dispel these myths or provide greater detail to explain each of the engines? I have a feeling I just have the classic artist verses programmer problem that is a part of the game industry anyway ;)

Hell no, unity 3D is NOT restrictive at all, and it comes with the asset store allowing for VERY VERY cheap VERY VERY good add ons, so basically toss in an additional 1000$ toward the asset store and you get a ton of AAA features back (nice networking, GPGPU particles, Real weather simulation, Real Time radiosity, virtual texturing supporting up to 256000X256000 textures! etc, some of those soon released, most already on the store).

Unity is really nice for artists too so tell her to download it & try it!


As someone who's spent around 10k$ on the unity asset store i can tell you that it's absolutely amazing and IMHO, above and beyond most of the triple A engines IF you include the asset store (and at 10K, i own most everything on there, which is quite far from a 300K licence for some engines . . .)

Also download this if you have a DX11 card and want to see what some add ons can do for unity, all the particles in this demo are done with TC particles, which is going to be sold on the asset store probably as soon as next week, with a price estimate of 85$ for indies :

http://forum.unity3d.com/threads/166715-Museum-of-the-Microstar-A-DX11-contest-entry-by-RUST-LTD-and-co

T3D (Torque 3D, in case there was confusion) would be your best choice, IMHO, if you are using C++. Tell you wife, girlfriend, or trans-friend (or whoever you were referencing) that the choice is either to start from scratch or take T3D. I really don't see how that isn't an easy decision as you can add whatever features you want given some time. Not to be unkind, but tell your "other" that until she learns how to program modern 3D game software she really can't make good decisions on the platform as she really isn't basing her decision off of anything but ease of pipeline, which can be easily changed, and for free. I mean, although she might be experienced, this is a project on a completely different magnitude. You probably aren't trying to make a AAA game, so you won't have a AAA pipeline, nor should you.

C dominates the world of linear procedural computing, which won't advance. The future lies in MASSIVE parallelism.

This topic is closed to new replies.

Advertisement