Malevolent's Pencil Sketch
Building a virtual world on a budget
If you search around on the internet, you will discover that there are several ways to build a MMOG. The most popular form is being part of an
already established "cool crowd" that draws in venture capital. Let's call that the Empirical method. The second most popular form is to
leverage your existing network of game content contractors, favorable technology that you've worked on here and there, and bootstrap yourself to a
point that you can sell a MMOG component technology while bobbing afloating your fairly successful game. Third, you can try licensing everything you
need on some "on some damn fool idealistic crusade".
We pretty much went swinging our lightsabers in with the latter.
Let me state that at the beginning of this project, everyone involved pretty much had zero experience building online games. We all had some sort of
familiarity with the game process, mostly as players, and a few of us had a bit more exposure to the realm of game making. However, that all said, it
was an amazingly huge learning curve to figure out how a MMOG works. We're still wrestling with a lot of the design patterns, but we do slowly knock
them down one by one.
The purpose behind this particular piece of literature is to inform would be indie developers of not only the true, hands-down, minimum costs to get a
basic world up and going that has a bit more than stale, rehashed content, and to provide a frank window into some of the skill-time-needs analysis
that we've hit and currently stagger along with. In other words, I'm going to answer why so many folks end up making technology instead of MMOGs.
Let's look at the dollars, first.
Basic cost analysis
Estimated cost of development to date: $86, 697.90
Basic costs less payroll: $13, 651.50
Note: Costs do not include business licensure, registration trademarks, marketing, office, or supply expenses.
Yikes!
How in the hell did that happen? Almost 15k out the window in less than 12 months? Well, that's what excel came up with based on a crude
analysis of my spending pattern in the last 12 months to support the project. Let's break it down a bit.
In my analysis, I created three "layers". The first layer deals with estimated payroll costs, which cover the estimated billable hours of folks working
on the title in the last six months. It is significant to note that everyone working on this title is doing so on a part-time basis, on a basis whose
contractual details are bound by a NDA. For practical purposes, because your labor section is probably different, let's just take our labor.
Without labor and office, legal and tax situations, you probably fall in the camp of "a group of folks with an idea". Someone in the group is the
primary venture backer. That person wants an idea of what they're going to have to spend to get the title to a sort-of-working early-early-beta
position. Right?
Layer 1: Core hardware and software
How many 3D modeling programs do you need to make your MMOG? We use five. Only one is free. Got hardware? Hint: live server, SVN repository, chat,
dev server, and something to back stuff onto: the minimum requirement. Oh, you probably need a domain and hosting environment. Don't try hosting it
yourself, in midway of development you'll discover you don't have time to admin your site anymore.
Bandwidth $2,220.00
Hosting $508.00
Servers $1,456.00
Software CALs $4,755.00
Layer 2: Models and supporting technologies
So, your primary modelers haven't learned how to build a basic humanoid yet. Not a problem, say you, I'll just buy one or two and use that as a
basis for moving forward. Well, you're going to need a mix of models to get an idea of how things are pieced together. Midway in development your
primary technology provider releases new frameworks that provide a huge boost to your capabilities. You probably want to pick that technology up too.
Pretty soon you end up with something that looks like:
TGE 1.5.X $150.00
AFX Combo Pack $74.95
Modern Soldier $34.95
Skeleton Pack $39.95
Adam Pack $39.95
Ava Pack $40.00
Dungeon Guardians $40.00
Torque in Motion $39.95
TA Combo #2 $99.95
Medieval Weap $34.95
bt Bundle $74.95
Ambient Audio $19.95
S. ACID Loops $64.95
turbosquid models $80.00
cubix $16.95
3d-diggers $110.00
fps Environment $59.95
rts Environment $39.95
D/D MMORPG Kit $99.95
D/D Gold Mem (1mo) $10.00
PG/GG MMOK $450 (150*3)
So, with a meek investment of approximately $15k, you too can have a MMOG with more than just basic content and a reasonable feature set. If you
already own the modeling software, the development technologies, and simply don't have need for the models, you can clearly save a good chunk of money.
What would make for a good write-up some other day is how to run the development of your artsy Studio parallel with the development of your game.
There's an abstraction point P up until you can do so. At point P, you can't. Anyway, that's for another day.
The crazy difficult business of making a MMOG.
Did you know that, by and large, all the model packs and technologies I mentioned in the previous list require you to break them apart and to rebuild
them to get them to work in your title? Or that some of them, in their present state, do not work at all with others without severe hackages?
Seriously.
Having some background in MUDs, I figured that building a 3D MUD was mostly in the technology-programming aspect.
Oops! I was wrong!
Instead, the difficulty of making a MMOG comes in areas folks probably won't believe me. Now, this could be quite long as it has the potential to dip
into a vast sea of knowledge and experience. To keep it short, let's just cover three areas of MMOG development and why they are Craaaazee-hard.
Point 1: Artwork is hard.
I vastly under-estimated the amount of art-work required in a 3D MUD or MMOG. The amount of artwork in a basic MMOG is huge. In the client build that
we've shipped out to folks, we have 3370 textures and 735 models. In the dev-build of the game, there is a factor of three more than that. Most of the
stuff we haven't rolled out for a reason I'll get to in a moment. Let's pause that thought and dive a bit deeper into this point about artwork, shall
we?
Not only is artwork hard, but as your game grows the skill levels required grow. Let me provide an example of what I mean by this.
When you start out modeling, you've got to learn the 3D space. Once you've learned the 3D space, then you've got to learn how to assemble a model in
the editor of choice. Once you figure that out, then you need to learn how to export the model successfully.
Successfully exporting your model to your native game format, in the langauge of programmers, is NON TRIVIAL. There's always something.
Concave or convex? Brush winding. Invalid faces. Missing faces. Missing collision. And of course, my favorite: exporter crashes and mystery problems
with no documented fixes.
Once you've learned to export the model successfully, then you need to learn how to properly texture that structure. This means you need to have some
command, good command, over photoshop or gimp.
Once you're able to texture the building, then you need to know how to place lighting and ambience. I used to joke, back in my ignorant days, that
there is no reason for a 3D company to employ a person to focus only on lighting.
Ooops! Wrong there, too!
Lighting makes or breaks your environment and your models. Lighting can make your models done for $0.50 look like a million bucks or visa
versa.
Now, once you've had this already busy person build this one model, you've got to go back to them and tell them that it just isn't working. Sometimes,
an entire fleet of models don't work. Others get locked up in export hell. And some are just waiting on other areas of the game to wrap up development.
In an ideal world, it'd be a simple deal to pickup photoshop, interior modeling, and lighting. However, these are very unique skills that take months,
if not years, to perfect. Oh, and then toss in the fact that you need to learn at *least* four different modeling applications to make any give scene
within the game.
Yeah, that's what I think, too.
Anyhow, in an effort to keep this short ... let me just say that this is a critical point for the indie MMOG community, where I see so many
projects rely on free labor to a greater or lesser degree. The software you need folks to learn is not like learning microsoft word. It's more akin to learning
how to do real modeling and that is a process that takes a lot of time, which in our time-starved world is not always available. What can you do? I'm
not sure. Hire some people? It's not an easy question. Maybe if public education would force everyone to take 3D modeling classes....
Typical pipeline for structural modeling:
1. Concept
2. draft model
3. draft texturing
4. basic lighting effects
5. testing of model in environment
6. model updates
7. new textures
8. new lights
9. re-testing
10. go back to 5
Point 2. Making Items and Quests
I enjoy writing. I always have. I've written articles for major labels and I've even published a book, with ambitions to get more done once I
straighten out a few things personally. Nothing prepared me for the writing challenge involved in creating a MMOG. What does it take to write a quest,
anyway? Just make a quick fedexer and be done with it. Right?
Why making quests and items is not "trivial"
We have approximately (over) 2,000 items in the game right now. Did you know these are all done by hand (with a wee bit of assistance from a bit of
propietary software I built)? The reason why every item ends up being done by hand is because of game balance issues. So, no matter how clever you are
in making procedural content, at the end of the day someone with human eyes must look at the data and fix the item set. It also means someone has to
test each item that procedural content generator makes.
At beta launch, we had a significant problem with folks getting the uber longsword set and slashing through mobs at 250pts per whack. That got fixed.
What about quests?
Quests need not only be sensible, but they have to convey some sort of interconnectedness (points for the long word?) with the game universe --which is
still evolving owing to technical limitations, new ideas, and "how things are panning out". In the beginning, those miniature golems in the Outskirts
zone used to have personalities. The creation of the quests and story for your MMOG is a *fluid* exercise. Anyone that attempts to compel a static 250k
narrative upon a MMOG, I think, is going to have a real hard time at it.
How long does it take to write a full quest? Long. This isn't NWN --even that takes time to do.
Let's assume something generous and say that it only takes you an hour to write 512 quality words, which is the average word count for a single page of
paper and, for our thought experiment, the size of one very small quest. Ignoring actually testing the quest, putting in the items, putting in custom
artwork, and assorted quest related needs ... how long will it take to populate your world with quests? I'm told by our beta players WoW has 150 quests
for the early levels, which is what they expect from us. So, doing the math ... 150 * 1= 150 hours. 150 * 512 = 76,800 words.
That kind of productivity is possible if all you're doing is writing. But in truth, in the indie market, you're wearing many hats. One day you're
modeling and the next you're writing a faction code fix, the next you're flatting terrain and the day after that -- you swear you will-- is the day
you've set aside to work on quests. Only something keeps popping up. A hole in the map here, a model boundary glitch there.
What's a fix?
I'm not sure. You don't just wave a wand and turn a non-writer into a writer over-night. Nor can software compel good quests or quality items. It takes
talented folks and, thanks to this project, I am now ever in awe of the skill and ability of the fine folks I see in the world doing similar type
things.
Point 3. The strange time paradox & oranges to apples
I can pretty clearly chart the progress of Wraith according to a particular timeline that, I think, holds true for most titles. What surprised the hell
out of me is what happens when you do the beta launch. The full cascade of challenges is well beyond the length-limitation of this article, but if
you'll give me a chance I will cover two points. The time requirement increases, not decreases, and your costs triple.
When working on my first book, all of the self-help books about writing a book spoke to this same tenant that the real work of writing a book happens
once the book is done. "What the hell," I would say, "are these people smoking? They say, 'Writing the book is the hard part.' They've got to be on
some sort of pill. I'll write my book and the earth will be shattered by my brilliance!"
D'oh! Writing the book *was* the easy part. They were right. And you know something? I think the building of a MMOG is the same way. So let me say it:
building the MMOG up to beta is the easy part.
Maybe a good rule of thumb is to say this. Your workload at beta will triple if not quadruple when you hit that magic day of beta. Your costs will
easily double. And you will look forward to a good night's sleep after about the third week. (So if any of the above has seemed like it was written in
the wee hour of the morning -- you'd be right!:))
The last item to be of benefit to folks just getting started or who are planning on releasing a beta soon. When I was up at IMGDC, Brian Green sternly
cautioned me about avoiding monthly subscriptions as a means for getting the bills paid because, according to his experience, folks would immediately
draw comparisons of your title to other titles in the runup. It made a lot of sense and he nailed it.
Ignoring the pricing dilemma, the situation we've found is that a tremendous amount of our feedback falls along these lines:
"Can you make X more like EQ?"
"I really liked the way WoW handled Y."
"In DAoC you could do Z ..."
I'm running out of time to explore this and the related issues, so I'll just raise the indie-dev questions:
Your players experienced the functionality of the AAA titles and expect it from you, no matter what you say. What next?
Your will have folks complain about your graphics quality compared to the AAA titles. As this rift continues to grow, what impact will it have on
your title?
Think real carefully: how would your game handle comparison to other titles and folks saying that they won't play until those similar features are
in place?
Enough hard questions!
Why folks make technology
It's easier.
OK, so all that I wrote before for a three word answer feels unfair. So let me expand on it some before I collapse from weary exhaustion.
It's also a bit more complex than "it's easier", being that a huge component is "I'm more likely to get money" making components than I am making a
game. The odds of an indie MMOG making it onto the frontpage of a list with Runescape are so bewildering that you might have better luck playing the
lottery. (Still, it *is* a fun lottery to play --this making MMOGs....) The odds of you making a technology component that turns over some dollars is
fairly high. You've experienced the problem space, you know that x component would solve people's problems, which if you set the right price ($5-150)
means a definite number of sales over the long term. So, do you sell a component or try the riskier MMOG path? On the other hand, you make a MMOG
that's missing functionality from the big AAA titles and suddenly you find players storming your office asking WTF is going on. Hmmm.
I'm more convinced than ever of three important things with regards to making a MMOG in today's market. First, you do it because you just *have* to
have your own virtual world. Second, you realize that in the process of building your own virtual world that it won't be long before everyone has one
(damn you, Bartle, for being right!). Third, you don't build the MMOG because of something silly like logic.
You build it for the love of the game.
-Eric "Malevolent" Rhea
Back to Newsletter Main Page