Acacia Research Group
have been releasing their annual Middleware for Interactive Entertainment
report for a number of years now, and have just released the 2006 edition
. The study includes analysis and forecasts of the middleware market for the next five years, with Acacia expecting 11.9% growth over that time, growing the market “from $718.4 million to nearly $1.3 billion”.
Principal/Senior Analyst at Acacia, Michael Arrington, notes that these numbers may seem larger than last year’s estimates, but puts this down to a more inclusive study, which he says includes the group’s efforts at better “pinning down mobile”.
Gamasutra spoke to Arrington via email about the study, the growth of the middleware market and where this will take the market in the next five years.
How has the way you have conducted your survey changed since last year's, and what was the methodology of this year’s report?
Not much has changed, but we get better data and more refinement with each iteration. This is a model we've been working with for some time, so during the course of a year we speak with as many providers of middleware, developers, and platform makers as possible and adjust our thinking as necessary.
We have been pretty successful getting more developers opening up to us about the tools and middleware they're using, what they'd like to see going forward, how many people in-house are doing what task, etc. This has allowed us to refine our numbers for spending on third-party solutions but, more importantly, has allowed us to nail down what we think are pretty good estimates on in-house development of solutions as well. That, and firming up mobile and interactive TV estimates, is one of the reasons our numbers look larger than last year. We've always had a pretty good picture of spending on third-party solutions - now we think we're about right with the big picture as well.
The report mentions that "customers who were previously and otherwise would have continued to use Renderware began to seek other engine-level vendors" in the past year - what other options are out there, and what kinds of market share do they hold?
Well, there are a lot of options but the big ones are Epic Games (Unreal
), id Software (Quake
), and Emergent (Gamebryo). Epic has been focusing more on its middleware business lately and, though there's always been a demand for its technology, I think it is seeing the fruits of that.
On the other side of the coin, Emergent has continued to refine Gamebryo (and is introducing some products that will help it expand beyond the engine business, while enhancing the value for Gamebryo users at the same time) and it has had a couple of notable hits (Oblivion
this year, Civilization IV last year). Those things help put Gamebryo into play with a larger number of potential customers.
id continues to be id. It licenses its technology but doesn't put much (well, any) marketing muscle behind it. People come to id because the tech is awesome and because being able to say you're using an id engine makes you “1337”.
There are others, like BigWorld for MMO focused developers, GarageGames for the ama-indy (mostly) folks, etc. but those three are the big ones.
What kind of market share does in-house software hold in comparison to third party software?
Third-party solutions are about 26% of total spending - $308.3 million in 2006.
What categories do you break third party software into?
Engine, Physics, Network, AI, other low-level [defined as “a wide range of products
from memory management and file access to audio, video, and graphics rendering”
] and other high-level [defined as an “emerging area containing specialty systems”
What kinds of developers are more likely to use third party middleware?
Hard question. Almost all of them use some kind of middleware - such as the stuff RAD Game Tools provides. As the solution becomes higher-level you'll find fewer developers that are interested in it. So, physics is an easy sell to most (anyone for whom physics is not a key component of the game, e.g. not hard-core racing games, flight simulators, etc.). AI is a little harder to sell, because a lot of developers want to create their own solutions, especially in terms of higher-level behaviours beyond pathfinding and such.
Game engines are the hardest sell, unless they come with instant market recognition - "We used Unreal 3!!!1!!". "Other high-level", where we include stuff like IDV's SpeedTree are case-by-case. SpeedTree is successful because it's a good solution and it's inexpensive (relatively). If you need trees and don't have a botany-crazy egoist developer on staff you'll probably use SpeedTree.
Where do you see physics middleware going in the five years?
Another hard question. Physics could go either way. In general we think more and more developers will continue to move to third-party solutions. The question for physics solutions providers is whether they will be getting paid for them or not. AGEIA's solution, while it may not be for the mass market right now, has pushed a button that will eventually result in physics running on hardware.
Whose hardware? Not sure. Maybe custom stuff though that's really a hard sell beyond hardcore gamers. We think it's more likely that graphics chip/card guys are going to be integrating physics capability as a means of continuing to differentiate their offerings. So, yeah, great future for physics middleware but will it still be Havok and AGEIA in five years? Probably but they're both going to have to continue to innovate.
Where do you see AI middleware heading in that time-frame?
We think you'll see providers continue to offer higher and higher level solutions in their products. We may see some stuff move onto hardware as well; the simple, repetitive stuff. More developers will embrace third-party solutions, but not en masse. At the high levels, especially as it involves characters moving around and interacting, there is sure to be some 'conflict' (or, taken another way, synergy) between physics and AI solution providers - you can't make decisions without taking the world into consideration and you can't properly reflect the world without some consideration of intent.
What kind of effect on the market has Sony's PlayStation 3 Tools & Middleware program had so far?
Sony is Sony. Middleware providers want their stuff used by PS3 developers so they'll do whatever they need to, to get on board. Sony, for its part, has been (so it seems to us) less active promoting its T&M program than it was with PS2. That's okay though. I think that developers are smart people and they don't really need Sony to educate them all that much.
Do you think we'll see more in terms of that kind of program?
All platform makers have something similar. I think they'll continue to evolve but, actually, I think they won't be as important to middleware vendors as the PS2 program was. It's more of a "okay everyone with a solution get in line and we'll approve it for use on our platform" and then developers will decide whether to use or ignore the solutions that are available.
Why do you believe the "issue of complexity" remains one of the strongest drivers behind demand for third-party middleware?
Anything that is, in developer-speak, "trivial" will be taken care of in-house. Anything that is really interesting or cool to developers will be taken care of in-house, to a point. As the platforms themselves and the gameplay and features required by the market become more complex, developers must spend more and more time building the solutions they need and want. At some point this costs much more money than it is worth - especially if there's someone else offering a solution for much less than you can build it in-house.
This is why middleware is a business at all. If it was fun and interesting to build a video playback system then RAD wouldn't have much business. It's no fun, so developers buy RAD's solutions. In an ideal world, developers would just buy a technology package that takes care of everything they need and just build content and gameplay. It's not an ideal world and won't ever be, but we'll continue along that path - because the developers that get closest to that pinnacle will be able to make better games with less money spent on stuff that isn't related to the content of the game. Nobody in Hollywood makes their own cameras or lights anymore (most technology development is in the area of VFX, stuff that makes their film different from other studios) - everything is spent on actors, scripts, sets, etc.
How has the acquisition of Criterion by EA changed the market for the major players in engine-level middleware?
Criterion going into EA simply means that most non-EA studios that might have used them are looking for other solutions. That's primarily good for Emergent. If someone was looking to Renderware to solve their problems they probably aren't going to go with Epic or id.
How will the next-gen consoles released this November effect the middleware market?
More complexity. More focus on networking. Both represent opportunities for middleware providers. And, people have already moved into the development cycle for these platforms - in order to have launch titles in place.
Why do believe the middleware market is growing, and at such a fast pace?
Well, for all of the reasons above. Providers would like to see it growing even more quickly. The real inhibitors to growth are:
The market is pretty fixed, people buy x games a year and more people making more games just isn't supported at retail.
Developers have really long development cycles (1 to 2 years) so they pretty much only come back to the middleware store every few years
Developers are a tough market to crack and, as your solution goes higher up the value chain, the more difficult it is to gain their trust and acceptance.
Now, we can turn that around and say that middleware is successful and growing because of 1 and 2 as well. By using middleware, developers can cut their development time on stuff someone has already figured out and use that time and money for technology that will differentiate their game or on content and gameplay - which will make their games better and, hopefully, help put their titles near the top of that heap of games that gets bought each year.