Using an existing engine saves time as it allows you to start developing the game straight away. However there is often a large up-front cost (at least for the triple A engines like Unreal) and it may also limit your game design as the engine may not support a particular feature you wanted.

Given that you will be a small indie developer (and as such self funded) I guess you won't have the $350,000+ for a triple A engine so it would be a choice between something like Torque or rolling your own. If an existing engine will do what you want then it is probably best to use one - it is one less thing to worry about.

Quote:Original post by bronxbomber92

Quote:Original post by Obscure
Using an existing engine saves time as it allows you to start developing the game straight away. However there is often a large up-front cost (at least for the triple A engines like Unreal) and it may also limit your game design as the engine may not support a particular feature you wanted.

Given that you will be a small indie developer (and as such self funded) I guess you won't have the $350,000+ for a triple A engine so it would be a choice between something like Torque or rolling your own. If an existing engine will do what you want then it is probably best to use one - it is one less thing to worry about.

Can you (or someone else) list all the "levels" such as AAA, and describe them? And, if anyone would know, of what level would the Unity (http://www.unit3d.com)engine be considered, AAA perhaps?

TBH there aren't 3 discrete 'levels' that you can easily pidgeon-hole a middleware engine into. Scoring is a better way of thinking about it. The more boxes it ticks, the more chance it has of being considered AAA (i.e. top-10% on the score sheet means AAA).

Additionally, different projects/developers/publishers have a different set of boxes they want ticked (e.g. engine-X has strong online, engine-Y has strong physics, X>Y for an MMO developer, but Y>X for a racing-sim developer).

Engines that support normal mapping, have terrain systems, etc are 10 a penny - practically all on the market tick the same (easy) 'features' boxes. The things that tend to make the biggest difference are practical criteria:

- which [commercially shipped] games has this engine been used on before? (i.e. is it finished? and has someone else taken the risk and proven that this engine works for a full game rather than a tech demo and doesn't have any nasty surprises in store?).

- what's the product support like? (i.e. when we hit a problem at 3am on the day of our beta milestone is there going to be anyone who can help?, even anyone to come on-site? Are there enough engineers to support all the games using this middleware at the same time?).

- how good and complete are the content tools? - SDKs and tick boxes don't do much for me. A renderer for example is there so that the artists can make their art look good and make it usable in an interactive way when combined with the gameplay code. It's no good if the only way to make full use of parallax mapping is for a programmer to write some code. Seriously, it's all about the tools that come with the engine!

- how stable and established are the company behind this engine? (i.e. will their phone lines go dead at a critical time, will bugs get fixed, etc).

- does it support (and are they already licensed middleware providers for) all the platforms we want to release the game on? (e.g. EA probably doesn't care so much about Linux or Mac, they do care about PS3 and Xbox360).

- what genre of game was this engine originally developed for, and how well does that map to our game? (completely genre neutral engines are rare - or suffer due to being too generic...).

- does this engine work/integrate with the other middleware we're planning on buying? (using a one-engine-does-it-all is easier in some ways, but is much riskier if your game differs too much from the one it was developed for).

- do we get the source code so we can modify it ourselves or finish it off if they go out of business?

- given that we'll be paying a per-game license for this engine, how does the cost compare to developing in-house (of course in-house development cost also includes down-time when no game can be developed because the engine isn't finished).

Quote:Original post by bronxbomber92
How are the games themself classified as either AAA or some other level (A and B?)?

The same way you rate a girl as a 10 or an 11. Can you describe that process for us? Then change "girl" to "game" and "11" to "triple-A" and "10" to, um, "not triple-A." There is no system - a game is either triple-A or it's not. (There is no "double-A" or "single-A" or "B" or "C" - there's no system.)

Like Tom says, there is no system. Publishers make a lot of big budget games but only some of them hit big. Those are triple A games - anything else isn't.

Kind of like being number one film at the box office. You are either a big hit or your not.

Quote:Original post by bronxbomber92
Thanks for clearing up my queries about game engines! How are the games themself classified as either AAA or some other level (A and B?)?

The term "triple-A" is something that crept out of the world of marketing on a dark, stormy night while nobody was looking. It was created by the constant hyping of games in development. Magazines were constantly being told that "Game X" would be the Next Big Thing. Instead of saying "It's on our 'A' list of games!" like all their rivals, some marketing type would would invent a mythical "AA" list for their own game. This naturally evolved into a "AAA" list once other marketing teams cottoned-on to this cheap trick.

It's corporate willy-waving gone mad.

It is also a fundamentally meaningless term. Every game in development is a "AAA" title in potentia. However, you only ever find out if it's a "AAA" title in reality after it hits the shelves and your accountant sees the returns on your investment.

In any market, regardless of industry sector, there is only one question worth asking: "How much profit will it make me?"

Chris Sawyer's "Rollercoaster Tycoon" games -- which were arguably rare examples of "indie", one-man-band games breaking into the mainstream market -- were developed for a fraction of the budgets of the "AAA" titles they competed against, yet sat in the Top Ten games charts for months. And since Chris had far, far lower overheads, he naturally made a lot of money out of them. Far more than he would have done had he approached their development using conventional, big-budget techniques.

The trick, as always, is to sell what your customers want to buy. If you can do so on a low budget and keep your costs down, you're far more likely to be around in ten years' time than if you decide to mortgage yourself to the eyeballs and risk it all on a single attempt at winning the gamedev lottery.

Pick your business battles with care.