Hi all,

One of the interesting things that people talk about concerning video games, this gen, is the rampant price increase of HD-based games for the X360 and Playstation 3. Some aruge that costs have increased from $5 million dollars for a 'typical' Playstation 2 game, up to $10-20 million for a PS3/X360 game. Because of that, we argue a lot about how HD is hurting the industry due to an 'arms race'.

A major fallacy of HD gaming is that it always costs more to make a game for the X360/PS3. The fact is, it does not. It's a developer's choice to spend as much, or as little money on a game. Take the Shenmue series - it cost around $70 million to develop, with approximately $29.5 million USD for the first game (or 3 billion yen, at the time). Given inflation, that was a massive chunk of cash in the mid-90s for a game.

So why are budgets so high? HD games *can* cost more, because more *can* go in them. An 8-bit game was much cheaper because you didn't have to pay as many developers/animators to produce a good game. Studios have to spend hundreds of developer-hours to make a game. And with each developer costing around $100,000 a year to support...You can do the math and understand why it costs a lot of money to make a game like MGS4, or Killzone 2.

At any rate; I've always been a big proponent of middleware engines as a reasonable solution to the rising costs of development. When you work with a new system, you have to learn to utilize it's power in such a way that the engine can work with it....Which can be not only tricky, but expensive. In-house engines may be very flexible, but they're not always guarenteed to give good results (think Two Worlds on 360), and do take tons of time to build (Killzone 2, anyone?)

So then there's the most popular one out there - Unreal Engine 3, which was used on games such as Gears of War, Mass Effect, Lost Odyssey, Rainbow Six Vegas, The Last Remnant, and a littany of 3-4 dozen other games.

Why use it? It's a pretty good engine, as it was used in Gears of War 2 which won quite a few positive nods from critics, and game reviewers as well. Gears (visually) looks on-par with the best games of today. So what's the advantage? Gears 2 cost a fraction of comparible high-end games such as Metal Gear Solid ($50 million), weighing in at under $20 million to complete (it sold more than MGS4 did on it's opening week, BTW).

Unfortunately, despite the fact that Gears of War has had great success with using Unreal Engine 3 (the first Gears cost $10 million to develop - which is pretty small given some other games' budgets) is that few companies have seen the kind of technical success with the UE3 engine. Mass Effect had plenty of frame & technical issues, The Last Remnant has been butchered by reviewers for loading/tech issues, Lost Odyssey was good, but had more problems than Blue Dragon (which had an in-house engine) and so on.

But the thing I've come to realize about Unreal Engine 3 is....

One of the biggest arguments that non-Gears fans would lobby at it is that Gears' budget had the fortunate advantage of being developed by the engine maker, Epic....Which led to reduced costs, and increased familiarity with the engine. Even Halo 3 wound up with a budget 3 times the size of Gears of War 1 (approximately $30 million USD).

So guess what the tie that binds every technical-suckfest UE3 developed game, that Gears 1 & 2 never have had? Only 2 developers have made more than 1 game using Unreal Engine 3. Of the 50+ games that use UE3, only Epic and Ubisoft Montreal have used the engine more than once.

And lets look at the list of what they made with UE3:

- Gears of War 1

- Gears of War 2

- Rainbow Six Vegas 1

- Rainbow Six Vegas 2

Guess what else? The lowest score for any of the 4 games is still above 85% (RSV2) and had very few, if any, technical issues.

I find this very interesting. Maybe it's just me, but could it be that studios aren't giving UE3 enough time to work with - instead churning out 1 product and assuming that since it's visually OK, that it means UE3 sucks? I think that could be a very interesting argument to lobby against a lot of the UE3 developers. Just like consoles, the more time you work on them, the better you get at making a game better for the said console - Even Gears 1 & 2 had major differences in visual quality, because Epic learned to utilize what the X360 had available. So why can't we assume the same for game engines 2?

I think the major test of this theory (that the more a developer uses UE3, the more awesome the games get) is going to be when Mass Effect 2 releases, hopefully in 2009. It will only be the 3rd franchise to have it's series driven by UE3. If my theory holds true (and it looks like it is, since ME2's development time is much shorter than ME1's), then Mass Effect 2 will have major visual upgrades - not only eye candy, but in the stability department.

Going forward, I hope developers realize that engines such as Source, Torque, UE3, Crystal Tools, and others' should become staples of how they make games, and not just one-off projects. Oblivion/Fallout 3 follow the same logical line - as Fallout 3 used a modified Oblivion engine. And guess what? The development time for FO3 was only a fraction of Morrowind or Oblivion. That meant less cost for Bethsada, and a better experience than if they used a different engine.

Thoughs? Comments? I thought it was a good arugment to make for UE3.

when you consider that the license to USE unreal 3 is over a million dollars...

and that just gets you the engine and some artwork-which chances are you will not use the artwork..........

it is not the engine that costs the most, it is the artists making the art assets for the game that costs the most, in time and money.

unreal3 is a great engine, but a Great engine does not make the game Great, that is the job of the designers, the engineers and the artists that translate their vision into a game. it is the engineer's that are responsible for how the game Runs, the artists on how it Looks and the Designers on how it plays, if one of the 3 fails, then the game will suck-especially if it is the engineers that screw up

"Only 2 developers have made more than 1 game using Unreal Engine 3."

Wow, that's pretty surprising. I would have assumed it was way more than that. So given that, I think you make a good point. Epic obviously know how to use their own engine, and what they did with it for Gears 2 was nothing short of amazing. I'm interested interested to see what Square-Enix, for example, can get out of UE3 in the future. Practice does make perfect, after all.

I think it was a pretty well laid out assessment :P Unreal 3 engine has obviously showcased some very promising capabilities, and I think dev's these day's just aren't learning the tools as aggressively as they should be, after all, the technical leap from last gen to this one was MASSIVE. The second iterations of all the UE3 titles should be far more promising on a technical level.

Yeah, One problem I have theorized with is that the DEV PCs and the console are entirely different monsters. I think these bigger games rely on their testing dept. to assure that the game runs as smoothly on the box as it does in development and at times, they just aren't up to the task.

As far as the engine goes, I think that inherently you will have better results when the people working on the game undertand the engine better because they partook(?) in developing it.

I can't wait for ME 2, but I've gotta finish 1 first, of course, Im already Level 24 "Commando" (go figure), so I'm sure I'll beat it in time :^) So far I haven't had too many issues with runtime on the game just grainy surfaces here and there, shadow rendering can be horrid at times too. My main complaint are the times my charachter gets stuck in a slight ditch and can't get out, thus causing me to reload the game. I want to see that ironed out for the next game. Oh and I want Rocket Launchers@!

I still think Sony needs to work with the UE team and optimize it more for SPU intensive work. I dunno, maybe do they're part to make it easier for developers xD

What about bioshock, the graphics are amazing, almost as good as gears, better in some ways, like water, if the keep this work they could make graphics as good as we have seen in gears. Its a very good point of view, if the develpers would make more than 1 game with the ue3, graphics could be even better without a doubt.

yea, but familiarity with art tools increases efficiency.  Having the engine allows the devs to have strong tools (that they need to learn) right from from the start.  if they start a new game with the same tools, then their efficiency clearly skyorckets as that learning period is removed.

It also allows the developer to tailor the game to their studio and game.  Epic has the advantage of being able to make the engine to support their games, but with a first game for anyone else that's nearly impossible.  The second game truly is the test.

The biggest reason to not use the Unreal engine is that it hurts your bottom line actually.

It's rumored to cost over a million PLUS you have to pay epic a percentage of the games intake.

Unreal should only be used for your midstream/games you expect to be ok or are worried about sales.... as it will absolutley kill you if you plan to make a big profit off it.

Hence why few teams have probably used it more then once.

If your going to make a bunch of the same series... your way better off making your own engine.

The problem is, that Unreal Engine 3 doesn't just 'get you the engine and some artwork'. Unreal Engine 3 utilizes the following sub-tools inside their package:

The third generation Unreal Engine was designed for DirectX 9/10 PCs, the Xbox 360 and PlayStation 3.^[1] Its renderer supports many advanced techniques including HDRR, per-pixel lighting, and dynamic shadows, and builds upon the tools available in previous versions of the engine. Unreal Engine 3 IPP (Integrated Partners Program) includes:

And I think that's a very important list. You have physics, facial expressions, video decoders, AI, UI (and UI logic), lip-synching, animation, motion capturing, graphics rendering, and animation solutions all inside one package.

That's very significant, because instead of having to custom-build said solutions, you have only need to learn the coding and use the interfaces in UE3 to get it working - So there's a much lower development time creating a solution for a specific in-game element. The director of The Last Remnant was pretty quick to talk about how great UE3 was to work with. I hope that the team gets to use it again to show how they've learned to create a second-gen UE3 game, and show how much better it can be used for RPGs. The same can be said of Feel Plus and Last Odyssey (could N3-2 use UE3? I hope so).

Kasz216 - I understand the argument your trying to make of UE3's major licensing costs...I wonder if we'll ever get a 'solid' figure on how much it costs to license - Could it be that multiple games using UE3 would cost a lot less? It's a good question to ask.

Furthermore, my next question would be 'how many games made in the past 5 years use the SAME engines for sequels?'

Because I honestly cannot name a ton that have. Yes, a few games like Rock Band I am sure did, but I really wonder how often studios use the same engine. Will Konami ever use MGS4's engine for a new game? What about the Halo 3 engine?

And please note: I'm trying to make the argument that developers DO need to stick to specific engine solutions for the budget crisis' they are facing....Regardless if it's UE3, Torque, Oblivion, or whatever. I can't imagine the logic behind making a $100 million dollar game like Grand Theft Auto IV, and then never using the GTAIV engine for any other games (yes, they're using it for TL&TD expansion...But could you imagine how much cash they'd save if GTAV used the same engine?).

Oh, and PS: I do work for a developer that makes an engine for typical gaming solutions: Torque. It's amazing how well they've utilized the engine, and made it quite versatile. It's one of the cheaper engines, as licenses start at $100 per workstation.