Not many people know, but EverQuest was initially a software-rendered game. We first started working on EQ back in 1996 using Pentium 133mhz machines with Matrox cards. SISA had licensed the Pyrotechnics engine and had people already working with it on Tanarus and a Spawn PC game (later cancelled). We borrowed the engine, network code, and a level editor. It wasn?t very long before we had something running and online, albeit very primitive.
As time passed, we bought faster PCs, updated the level editor, added a simple interface, etc. Then news about the Voodoo 1 3D card began to float around. We hired John Buckley, who had worked on the engine at Pyrotechnics, and he started modifying it to work with 3D hardware. At that point we were planning on releasing both a software renderer version of the game and also support for the Voodoo 1.
A year or so passed and I remember distinctly being in a meeting with the EQ leads and Smed. We were discussing min specs and hardware requirements. After a while, Smed starting pushing the idea of going hardware only, 3dfx Voodoo 1 required. I remember being wary of the idea, but also intrigued by it. Such a decision back then was a big one and we knew there would be very few games out when we launched that would also be hardware-only. But when we considered all the more we?d be able to do with the game world, we fairly quickly agreed with the idea and the decision was made.
By the time we released we were running two Voodoo 2 cards in tandem and thought the game both looked and performed well. And when we released and the game caught on like wildfire, we were pretty sure we?d made the right call. Ironically, about the only thing we didn?t pull off technically was a seamless world (something else Smed had really wanted to see).
Fast-forward a few years and I was at Sigil talking with the Founders about the tech level we wanted to achieve with Vanguard. Many of us, coming from EQ, wanted to make another MMO using state of the art graphics. We felt it would make the world more immersive, and, quite honestly, we were graphics tech-heads and very excited about what hardware and Direct X would be pulling off in the years to come.
With EQ we were 3D and hardware accelerated, while our biggest competition at the time was UO, a 2d tile-based game. And we?d trounced UO pretty well. So looking at what might be Vanguard?s competition, we felt we had to visually trounce them as well. Any concerns about performance were mitigated by the belief that by the time the game came out, there would be plenty of PC horsepower readily available. And Vanguard players, if they didn?t already have the necessary horsepower, would undoubtedly upgrade their PCs. Heck, a LOT of people had bought Voodoo cards just to play EQ.
Of course, looking back now with 20/20 hindsight, we were very wrong. Over 80% of the people who bought Vanguard and tried to play it quit by level 2 or 3. What could be the reason? Well, given how fast it was to level the first couple of times, what could be so horrible that people would quit so quickly? My bet is crappy framerate (due to rendering too many polygons and too many and too big textures) and bad hitching (being the result of the world being seamless and having huge art assets). We also released early and didn?t have a chance to optimize the code (but this is a subject I?ve already addressed in my blogs and not at all the only reason we had issues).
Why didn?t our techno-geek approach work with Vanguard when it had worked with EQ? Probably there are a few reasons, but the big one I think is the perceived difference between the games and their competition. First, EQ didn?t have a lot of competition and many people were new to MMOs. This wasn?t true with Vanguard ? there were many released MMOs when it came out. Also just about anyone, techno-geek or not, perceived a big difference between a 2d tile-based UO and a truly 3d game like EverQuest. It was like going from cassette tapes to CDs ? just about everyone perceived a huge difference.
EQ was also a fairly hard-core game (although it?s mellowed through the years). EQ?s players were into the game big time, but also into the technology. They were willing to upgrade their machines to play this amazing looking (at the time) and amazing playing game. EQ topped out around 500k players, and I think it?s safe to say that the majority of them were at least fairly hard-core.
Now compare Vanguard and WoW. Vanguard, technically, is far more advanced than WoW. But perception-wise? It?s not like going from a cassette to a CD; rather, it?s like going from a DVD to a Blu-ray disc. Videophile that I am, I totally prefer 1080p to 480p. But my wife? She shrugs at the difference and gets on my case about buying expensive blu-rays all of the time, upgrading my movie collection whenever a new disc is released.
Then you have more casual MMO gamers. Not only to them is the graphics technology not a big deal, but they?re also far less willing to upgrade their PCs with expensive new CPUs and GPUs. Blizzard was brilliant ? they created a mass-market MMO that could be played by the mass-market. Where their game lacks in technology, they make up for it with the quality of art and overall polish. And, as a result, they have millions and millions of players, not 500k (yes, I know there are other reasons as well, and I?ll undoubtedly cover many of them in future blogs).
Now for those of you who read my pre-Vanguard launch posts, you know I?d already considered some of this. We knew that we were making a more hard-core game, and we also knew if we could even get 500k players that the situation would be very profitable. So why didn?t the more hard-core gamers upgrade their machines, like they had with EQ?
As mentioned, the perceived extra quality and extra immersion was not nearly as great as with EQ vs. UO. Also, even if you had a pretty buff machine, Vanguard still ran poorly in many situations. EQ had some performance problems, even on dual Voodoo 2s, but not nearly as great. Also, a 3D world was a relatively new experience in 1999, especially an online 3D world. In 2007, 3D wasn?t a novelty but rather a standard. And if you didn?t like one MMO, you had others to try-out. In 1999, you didn?t really have that luxury.
So where did we screw up? Were we just victims of changing technology and changing standards? No, I don?t think so. I think we were blinded by where we saw technology going and all of the cool things we could achieve by harnessing that technology. We used the EQ experience to bolster our confidence that people would upgrade for a great game. We looked into the future, using Moore?s law and the like, confident that while the game in development was a dog, that by release graphic cards would be both powerful enough and cheap enough that system reqs wouldn?t be a big issue. This, obviously, was not the case. Lastly, I certainly encouraged my graphics programmers to attempt more and more. They?d integrate something new, say high dynamic range lighting, and I?d go ?ooh? and ?ahh?, patting them on their shoulders and full of encouragement.
Looking back, and then looking into the future, I hope that I wouldn?t make these same mistakes again. Whether I work on a more mass-market game or something targeted and niche, I still need to remember that forcing new technology on people had better offer the player something night and day better than what they are used to. I need to remember not to get caught up in the allure of 3d graphics technology and what new cards and new versions of Direct X promise. High dynamic range lighting, a 50 mile clipping plane, and a massive seamless world do help with immersion, but not at the cost of performance and playability. Immersion has more to do with a pleasing quality of art and polish while simultaneously achieving a playable framerate. And the lessons learned with EQ, while invaluable, are not necessary the be-all and end-all of lessons to be learned. Having the EQ experience behind me is a huge advantage when it comes to building an MMO, but I need to remember that EQ was 1999 (and, for that matter, Vanguard was 2007). Developing a 2014 MMO will rely on lessons learned in the past, but also on understanding the audience and the market at that future time. And while I disagree with those who assert EQ?s success was all about timing, I do have to agree that timing was a big part of it. Times have and will continue to change.